Re: [time-nuts] Anybody have suggestions for time related science fair projects?
I don't know what sort of scientific level this contest is geared for, but would guess that for middle-school level, extreme numbers-oriented analysis of esoteric, time-nutty things may not dazzle, but bore the participants, judges, and audience. It may be best to relate to more hands-on, everyday experience and observations of "normal" people. I like the suggestions about GPS and stroboscopic and lasery stuff, where one can maybe appreciate how modern everyday things work (like GPS, or how it's possible to talk to or send a picture to anyone in the world on your cell phone, and how these could not happen without precise time), or something visual and physical. Some of the props should be "ordinary" things, like the a cell phone or GPS receiver, for example. Lasers are always good as long as there's a direct visual component to the observation. Strobe type stuff is particularly easy, because it's doable with mechanical and acoustical props, and signal measurement times are in reach of common lab equipment like generators, scopes, and counters, and of course there's a big visual experience component. Small power visible lasers are common nowadays, so easy to use. Strobe lights are fairly common too, but maybe not so much as the other items. You can build (or buy) quite a nice strobe light nowadays using high-powered LEDs - the kind used for replacing incandescent and other illumination. This is quite easy and much safer than dealing with flash tubes, and is much more versatile. In fact, maybe this could even be a science fair project. The time element is in the stroboscopic effects and ability to slow or freeze apparent motion - almost everyone has observed this and can relate. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] TCVCXO Adjustment
If it's just to set for the initial setup or aging, just do it the old-fashioned way, with a trimmer pot to run the Vt - simple, easy to program, and it remembers the setting. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] ULN regulator with more current capability than LT3042?
Before adding complexity of parallel devices or external passing, why not just try the old parallel-resistor trick? If the load takes a certain minimum current under all conditions, provide less than that via a resistor from the raw source to the regulated output, likewise under all conditions. The main regulator will still do its thing, but not have to supply all the current. You're just looking to get a little more margin between the spec current and actual load. The short-circuit current would be increased too, so take that into account. Now some will say, "what about the ripple voltage from the raw supply causing ripple current going right to the output?" Depending on the regulator's characteristics, I think it should be able to handle it just fine - it's easy enough to hook it up and see how it works. It would be good have OVP on the output regardless of the regulation scheme to protect the load if it's sensitive to OV. Obviously, the parallel resistor one could allow the voltage to go high if some load is lost, but the same can happen with a regulator failure too. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Machining some aluminum help!
One thing I forgot to add - if you must use the 4-40 screw size for some reason, you can save a lot of grief by using a slightly larger bit than the standard tap drill. You're not really too concerned with optimal fit and strength here - it's more about being able to make a whole lot of usable screw holes without losing too many bits and taps, or reworking. You can experiment and see what size makes the best tradeoffs. It's amazing sometimes how a little more root clearance can makes things so much nicer - a shallower cut, so less torque on that tiny tap, more room for chip clearing, and better lubing. You can also get deeper threads to compensate for less meat in the thread root. Even the drill bit will have a better chance of survival, since it will be slightly bigger and stronger. Of course, if you go too far, there won't be enough thread - but then you'll have reason to jump to the next size up. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Machining some aluminum help!
This is the first time I've looked at time-nuts in about a month, and I noticed the run about this project. I'm very late to the party, but have a few suggestions that may help - if it's not too late. I quickly scanned many of the posts, and agree with many of the ideas. Please forgive if my suggestions are redundant to what's already been said. First, I assume that the aluminum box is a simple extrusion, so it will likely be a soft alloy that will tend to gall horribly with machining - especially bad for anything that needs high precision. With these kinds of material, go big, starting with bigger fasteners. With 1/4" walls, you can easily up it to 6-32 or 8-32, as long as the holes are fairly shallow, and you can jig it up for good centering and plumbness. With 4-40 and blind holes, you're just asking for trouble - especially taking a chance 40 times. The tap drill will be quite skinny, and prone to deform and wander as it goes in, and can easily be snapped off when it stalls due to the galling - and that's just the drilling stage - the tapping will be worse. Bigger threads give you a chance to get it done with fewer fasteners and holes, and much less grief. The thread depth should allow for at least one pitch-diameter of penetration for strength, but preferably two or more, so you don't have to worry about finding exact right screw lengths that won't bottom out. Depending on the thickness of the end plates, you could get down to two or three fasteners per side to hold the small pressure needed. If you're using a drill press, punch mark the hole centers, then use a center drill to make the pilot holes for the tap drill. If you're freehand drilling, put the piece on the floor and drill downward, keeping it as plumb as possible. Definitely use an oil or other lubricant for all the drilling and tapping operations. For sealing, I'd recommend against fancy o-ring features and such - these are also harder to machine cleanly in soft aluminum, and add unnecessary complexity. If the end caps need regular remove and replace operations, then go with a pliable gasket, have more fasteners to get more uniform compression, and make the threads deeper so they'll last longer. If the sealing is one-time, or seldom needs to be broken, I'd recommend using a gasket sealing goop that will work fine with few fasteners and even rough surface finish. My favorite is Permatex #2 "Form-A-Gasket Sealant," which I've used for all sorts of stuff over fifty years (back then it was Radiator Specialties brand). Don't use a silicone goop unless you want to spend a lot of time scraping off the old stuff if it needs to be opened. If you do use a goop, it's a good idea to machine in features for prying the lids off, such as gasket-plane screwdriver slots, or extra tapped through-holes in line with the mounting holes on of the lid. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Sub-ps delay line
If you make a variable delay by adding external RC (varicap) circuits, the edges will be slower, and the amplitudes will be affected. This will tend to complicate the detection and reshaping of the clock by the LTC6957. Pay particular attention to page 24 and Fig 8 in the datasheet, regarding the strong effect of input overdrive on prop delay and symmetry. This should be considered in the design of anything added in front. It may be better to use this effect advantageously instead, to get variable delay by tweaking DC bias in the right places. Depending on which edges are ultimately used in the ADCs and DACs, you may be able to affect all sorts of beneficial timing control - or have all sorts of timing problems - depending on how you handle it. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Sub-ps delay line
Even presuming this system is for clocking within a single-board environment, that seems like a fairly tall order for setting resolution and long-term stability. It could be tough to keep these numbers with temperature and supply variations in all the circuitry involved, including the parts that ultimately receive the various clock signals. On the other hand, making it adjustable delay-wise should be fairly easy - almost anything you do to tweak the operating conditions of this essentially analog system will change the delays. The trick will be to compensate for the unwanted effects, and consistently control certain things to set the delay. I'd recommend using differential signals throughout, if possible, with coarse delays set by sections of on-board transmission line, and fine adjustment by tweaking DC bias at various points. I don't think you need to add any extra variable delays such as varicap circuits - you should be able to affect enough control and compensation just with the normal part characteristics. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] HP 10811 40 Hz low.
I had a reference problem on my Z3801A years ago, and ultimately found that the opamp that controls the 10811 oven temperature was bad. I think it turned out there was a bad batch of certain date codes. Replacing the IC with an equivalent type fixed it right up, with no other changes or adjustments. Also note that the the ovens are sequenced to avoid drawing too much power from cold-start. As I recall, the inner one has to get up near operating temperature to get things running ASAP, then its status signal enables the outer oven driver (a switched-mode converter on the Z3801A power supply board) to start - it takes longer to stabilize. I think the inner oven should take maybe 20-30 minutes tops, to reach the spec range, and the outer one gradually catches up over maybe an hour or two. The crystal is very sensitive to temperature until it gets to turnover, so the frequency can seem way out of whack compared to its normal operating point. The oven temperature doesn't have to be very far off to cause this, so a failure in the inner oven control that throws it out of range could cause problems. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Totally unrelated, but..
It would be helpful to give some specific info about the circuit and regulators used for these experiments, such as operating voltage, load current, and regulator type. I may have missed some points in the discussion on this, but I think the following things are the case: 1. The circuit exhibits excessive internal interference only when using certain regulator ICs, but works fine with others of the same type. 2. No apparent oscillation or excessive output noise can be found with scopes and SAs to explain it. 3. Various battery voltages have been tried in place of the regulator to run the circuit and seem OK. 4. The receiver carrier is 467 kHz (not MHz?). 5. The regulators in question are LM78XX, which use a band-gap reference. 6. Changing or adding various filtering caps seem to have no effect on a "bad" regulator. Without additional info, I would suspect that the bad parts have a low-level oscillation somewhere near (or harmonically near) the carrier, the LO, or the IF, that is too small to see above the PS noise floor, but big enough to cause problems. It is likely these frequencies are in the range of where a linear regulator could oscillate. I doubt that one could oscillate in the VHF or microwave region, out of reach of your SA's span and sensitivity (presumably, depending on what you have). If the regulator could oscillate up there, it could certainly leak through or around any typical near-band filtering and decoupling in the receiver, and cause problems. Looking in the time domain with a scope, there will be less sensitivity and dynamic range, so there could be something below the floor, and maybe below the SA's bottom frequency range too. The SA's low-end can be compromised by the need to safely AC-couple the PS voltage into the 50 ohm input. A reasonably-sized coupling cap may put it too far up to see a small but important signal in the kHz range. If it's a low frequency or in-band oscillation, you may want to look at the receiver circuits for internal susceptibility and PSRR - you may have discovered an unanticipated weakness, separate from the regulator issue. Another thing to consider is that there may be a specific supply voltage that causes the problem - like a marginal circuit in the receiver front or LO going unstable. The battery test, I presume, was at certain discrete voltages. It may be worth running it on a variable PS over a continuous range. It's possible that the bad regulators just happen to land at a "bad" voltage. With a fairly wide tolerance spec, they could be all over the place. One way to eliminate this is to measure as precisely as possible the output voltage of a bad regulator, then replicate it with a variable supply. If the receiver circuit works fine throughout its supply range, then the regulator is again the prime suspect. Also consider what the load current is versus its max rating - if it's anywhere close, it could be on the verge of current-limiting, and all sorts of strange things can happen. If this is the case, adding a helper resistor from input to output should get it back into the normal range. If you haven't already, before all kinds of experiments and analysis, try the good old heat and cool methods - blast a bad regulator with freeze spray or a heat gun and see what happens. And of course, do the same to a "good" one. And maybe also the LO. This is an interesting case, and I think we all would like some more info on the particulars. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] So what’s inside that Cs Beam Tube anyway?
Wow. Nice job of dissection, and good pictures - very informative. No wonder those things are so expensive. It's a shame that they're not built in such a way that just the wear-out parts could be replaced, and not wasting all the rest of the design and craftsmanship that's probably just fine. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for a low power very low noise DC/DC converter (100 - 200 ma 10VDC or 15VDC)
I should mention that the input supply filtering to the DC-DC converter should have good attenuation at the switching frequency, but not at low frequencies, so there's no need to get carried away with the size of the filtering at the converter input. Too much filtering, especially inductance, can be detrimental, depending on the type of converter used. If the converter is a simple chopper type without regulation, it will put out in proportion to input, so line rejection will be poor. If line ripple is an issue, then common solutions are brute force more filtering with capacitance, or linear regulation at the input or output. If a self-regulating type DC-DC converter is used, and uses PWM, it will usually have pretty good line rejection, so eases the low frequency filtering requirement. However, since its input resistance is negative at low frequencies, it affects the choice of input filtering. Avoid overdoing it with L-C input filters that cut at too low a frequency or have high Q - it can result in oscillation. Lots of C is usually OK, but not too much L. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for a low power very low noise DC/DC converter (100 - 200 ma 10VDC or 15VDC)
Yes, for best quietness, you definitely should "can it up" in a metal box, and use feed-through caps for all the I/O, including the commons or grounds. You have to figure out also where all the currents flow, and contain the loops. With sufficient L-C filtering on the input and output (all inside the can along with the converter), you should be able to get whatever degree of cleanliness is required. The in and out ports should include common-mode filters (also inside the can). When it's all said and done, there should be the input power port +/- via feed-throughs, and the output +/- supply port via feed-throughs. The can should not carry any current except that which flows through the common-mode filtering, so don't connect it internally to the either input or output "common," even though ultimately they are tied externally. This will help to keep the ripple and displacement current loops contained within the can. The can should be solidly connected (mounted) on the system ground chassis, and the feed-through caps will give the final degree of RF suppression. The I/O ports then can be connected as required. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 1PPS to 32.768 khz
I'm kind of late to the party on this one, but I think the simplest approach with the least disturbance to the operation of the original system would be to form a VCXO and PLL. Good old 4000 series CMOS stuff should be plenty fast enough. Two pieces should be sufficient. For example, a CMOS 4060 counter/oscillator plus a 32,768 Hz resonator, rigged as a VCXO could make the clock, and also divide it down to a convenient range for 1 PPS comparison. A 4046 PLL or some simple logic could then do the comparison and make the correction voltage to the oscillator. Another option may be to use the clock chip's own system to get a comparison frequency, if there is a definite and fixed relationship to some output signal, say a digit scan line, or punctuation (colons between HMS) flash signal. It's conceivable to then use the chip's XO, modified to make it voltage-tuned, along with some form of phase detector logic. This could be very simple to implement, but would take some figuring out, and risks hurting the clock chip if you screw up while experimenting - a definite disturbance. So, it's probably best to use an external circuit for all. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 1 PPS 50-ohm driver
For high-drive outputs, my favorite is the 74AC541 octal bus driver, which is very nice for paralleling outputs through small series Rs. If you look at the pinout, you'll see why. BTW with any of these high-drive circuits, it is essential to provide good bypassing of the supplies, and be sure of clean low impedance pathways for the signal output and return currents. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for Z3801 OCXO
You may have seen this before, but there is a nice writeup on how to disassemble the whole thing, with pictures and all. http://www.realhamradio.com/GPS-oven-journey.htm I had to take mine apart quite a few years ago to replace the opamp in the inner oven temperature regulator circuit. If your problem is just adjustment, you don't even have to open the inner oven. BTW as I recall, the ovens are sequenced, to reduce power at turn-on. I forget the order, but one (probably inner) has to be up and running close to setpoint, before the other is activated. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Mechanical clock sound pickup circuit
This may be totally ridiculous, but maybe there's another way to get a balance wheel signal. The X-band Doppler type microwave motion detectors can pick up various object signals in free air from quite a distance, so maybe up close there would be enough resolution and penetration of the metal parts of a timepiece to get a usable signal in and out. It would tend to accentuate the fastest part of any motion - the balance wheel in this case. I can picture setting one up with the horn pointed at the thinnest part, likely the watch face, from maybe a few inches away - or whatever it takes to not overload the detector. The audio detector signal (if sufficient) could then be processed in the same way as with a microphone sound signal. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] ACAM GP22 Chip
I don't want to muddy the water too much, but I'd second going all-digital, along the lines of what Bob L. recommended earlier. It seems like you just need to make some specific delay times, so various logic counting or shifting circuits should be just fine - and well-defined and understood. If you go with coaxial delays, the loss and dispersion may cause much analog grief to determine how much compensation and amplification is needed, and the thresholds for comparison for squaring it up. This is especially aggravated with long lines. In the hundreds of meters range, it may be better to go with optical fiber instead. The bandwidth can be huge, and losses will be tiny, even into km lengths, and all kinds of E/O and O/E conversion parts are available for the ends. There can be some timing jitter due to the noise of the conversions, and AGC issues, but probably less than the noise associated with adapting long coax cables to this task. Another option may be acoustic glass delay lines. I don't know if any are made nowadays, but they were common in TVs and VCRs, with time delay somewhere in the horizontal line duration range, and very compact. It could be that there are many more kinds for various applications (or maybe all obsolete - replaced by digital). Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for ECL divide by 3 with symmetry
Problem solved - one missing connection was fixed, and it now runs just fine. Symmetry looks good on a scope, and the toggle rate is plenty enough. It runs OK to beyond 120 MHz input. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for ECL divide by 3 with symmetry
After studying the various divide by 3 circuits, I decided to try designing one that would be simpler in terms of package count, using available ECL DIPs on-hand. Instead of the JK-FF version followed by duty cycle-fixing circuitry, I opted for two 10131 dual D-FFs to provide the state machine, with 50 percent duty cycle. A 10116 line receiver provides the input interface and two-phase clock. So, a circuit of three DIP packages does the whole works. A quick paper analysis showed that it should work. I gathered up the parts and I built it onto a small vector circuit board, but it did not work. I did it relatively quickly, so probably have a wiring error to figure out. To make sure I didn't miss something, I ran a check with a simple logic simulator that I found, and it proved out OK, design-wise, so I think it should be good to go once I figure out the proper wiring. A summary of the circuit and operational simulation is attached. It should be fully synchronous and glitchless up to the toggle limit. Also, starting it from the one disallowed state seems to be no problem - it quickly cycles to the proper sequence. Using D-FF decoding within the counter section has a little more prop delay than the gated versions. I didn't check the timing limitations yet, but I'm pretty sure it will be OK at the required 50 MHz toggle rate. Ed div3 sim1.rtf Description: MS-Word document ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for ECL divide by 3 with symmetry
I found I have some 10135 dual JK FFs on hand, so will try a version based on JK, that needs no extra gates, with a 10116 triple line receiver for I/O. If it turns out that symmetry is needed, I'll add a 10131 D FF, and the 10116 already provides both clock edges. Either way, it should be reasonable in simplicity and power. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for ECL divide by 3 with symmetry
Thanks guys - good info. Some of these circuits must be the ones I've seen before, and some are new to me, so I'll keep them too (until I lose them again). I'm not sure if I will need symmetry in the output, but it seems better to have it for cleaning up if necessary. The resulting signal will be the reference frequency for a phase-locked microwave oscillator - the old brick style. So, it ultimately drives a class-C 2N5109 stage that drives the SRD for the sampler, with plenty of harmonic content anyway. I just have to make sure the harmonic I need doesn't somehow get canceled out or land in a null. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Looking for ECL divide by 3 with symmetry
I need to build an ECL divide by 3 circuit to run at about 50 MHz input. I know there are lots of examples out there, but I vaguely recall years ago I stumbled upon one or more that also provided more of a symmetrical output nearly 50 percent duty factor, by using both input edges, or reclocking with another FF. I saved the info, but of course can't find it now that I need it - in my computers, papers, or online. Does anyone know of these tricks, and any example circuits - this would save me some rediscovery and design time. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Looking for ECL divide by 3 with symmetry
I have rediscovered what I need, so no problem anymore. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Z380XA The saga of the aging 10811
It could be that the inner oven temperature regulation is off a bit, or even failed. There was a report at one of the popular time-nuts oriented websites (I forget which one, but it's the one that had extensive coverage of the Z3801A) about inner oven problems a few years ago. I took mine apart and found that it had an IC of a particular date code range that was prone to failure. I can't recall whether I replaced it with the same type but different date, or an alternative, but it worked just fine after that, with no tweaking of coarse EFC needed. The IC was a dual opamp I think, that controlled the oven temperature according to the thermistor signal, and drove the heater transistor(s). It was a fairly high performance type, but not that unusual. I think it was a Linear Technology brand part. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Tektronix Sample Heads
Yes, I think those were from the era when service info became sparse - treated as modules or black boxes - unlike the products from the good old days, which had much more info available. I believe the SD26 is to fit in the Tek 11801B or other 11000 series sampling scope setups. A quick look in the Tek catalogs circa late 1980s to the 1990s should help to ID various models. Check the usual manual sites such as bama or ko4bb, or maybe even Tek has some info available. You may be able to get some more info if there are manuals available for these scopes Also, if you plan to use it as an independent sampler, it will not do much on its own. Remember it's just the sampling head - a modular front-end that determines the performance, but only within the proper support environment. The scopes have the circuitry and smarts to operate the heads to get useful results. It is extremely far from trivial. The simplest way to use it would be to get one of the scopes. As David suggested, the tekscopes group should be the best resource. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DeLorme Tripmate GPS receiver
One thing I want to clarify - it is not a Jupiter GPS module - it just uses the chipset, presumably hooked up as in the application info. The board is proprietary, and there seems to be no standard electrical or operational interface as would be expected in an OEM GPS module. So, the only way to figure it out is to go by the chip details to see how it's supposed to work. previous message: I peeled open the shield can without too much deformation, so it can be restored. I found that it's a single board, with the DSP on one side, and the RF section on the other. It is a Rockwell chipset, with 11577-11 DSP, and 6732-13 RF. On searching I found that this seems to be called their Jupiter GPS from circa late 1990s - I found quite a lot of info at the module level, but not for the actual ICs, like pinout data. The set includes all the usual GPS stuff including 1 PPS, and is capable of several levels of on-ness. So, if the uP that makes it a DeLorme merely sets some control lines to activate it, then I should be able to override them to force it always on - if I can figure them out. If instead the uP programs something internal to the DSP to control power states, then fuggetabout it - it will be junk. So, does anyone know of the Jupiter chipset, and where to find chip-level info for these parts? Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Simple AC mains zero-cross detector
Actually, the core saturation depends on how much voltage is applied at a given frequency. Most power transformers are run partly into saturation at rated line, to get the most from the copper and iron available, in exchange for heat and less efficiency. The magnetizing current and losses will occur even with no load. The resistive loss will go up more with load. But, I don't think this matters in this application anyway. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DeLorme Tripmate GPS receiver
I opened it up and found a 4 by AA cell battery holder for power, and a single module that looks pretty proprietary. One side of the module has a patch antenna, and the other has the brain. I assume there's an RF board in between, but cannot open it further without possible damage - the shielding can is soldered at the edges of the brain board. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Simple AC mains zero-cross detector
It seems to me that a low voltage secondary should be OK by using a fast comparator IC rather than a transistor to decide - the gain of the IC allows for much smaller detection levels, so the equivalent zero-crossing velocity could be the same. An IC tripping in a 10 mV band should provide the same effective ZC velocity at 12 V input as a transistor working around 100 mV with 120 V input. Or am I missing something? Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] DeLorme Tripmate GPS receiver
I peeled open the shield can without too much deformation, so it can be restored. I found that it's a single board, with the DSP on one side, and the RF section on the other. It is a Rockwell chipset, with 11577-11 DSP, and 6732-13 RF. On searching I found that this seems to be called their Jupiter GPS from circa late 1990s - I found quite a lot of info at the module level, but not for the actual ICs, like pinout data. The set includes all the usual GPS stuff including 1 PPS, and is capable of several levels of on-ness. So, if the uP that makes it a DeLorme merely sets some control lines to activate it, then I should be able to override them to force it always on - if I can figure them out. If instead the uP programs something internal to the DSP to control power states, then fuggetabout it - it will be junk. So, does anyone know of the Jupiter chipset, and where to find chip-level info for these parts? Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] DeLorme Tripmate GPS receiver
I just picked up an old DeLorme Tripmate GPS receiver for cheap, and am wondering if it can be used for getting a 1 PPS signal. I looked online a little and found it's pretty common, but didn't see anything about getting deep into the guts. Most hacks seemed to be about getting it powered up, fooling it into starting up (apparently the character string ASTRAL has to be sent to it), and running to output navigational info (NMEA). The interface is a DB-9 serial connector that I suppose went to a PC and nav software. I don't care about that, and I don't want to even talk to it - just whether it can be fired up and automatically running with an internal mod, to get a 1 PPS out, without ASTRAL or anything else. Does anyone know or have info about this aspect, or info about the guts - block diagrams, sub-modules, documents, schematics, etc? I haven't opened it up yet, but will soon see what's in there. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Frequency distribution isolation transformers YCL 20F001n arrived
I investigated those 10b-t isolation modules a while back, and have saved every module from every network card and router/hub/switch that I have junked out. The very old 10b-t stuff is the best for getting LPFs and individual per-channel (port) type parts. When they started making 10/100 Mb/sec, the 17 MHz filters were eliminated, and the parts got integrated to ever-higher levels, with multiple channels in each module. You have to be able to find the data sheets to be sure of what's in them - some are transformers only, and some also have LPFs and common-mode chokes in various combinations. The filter sections can also be cascaded for even sharper cutoff, but there's quite a bit of crosstalk, so a lot of higher frequency stuff gets through, especially above 100 MHz, so it's mostly effective from around 20-100 MHz. It has been mentioned before that very sharp filters will tend to have more phase noise (phase shift with temperature/component variations), but the negative effects depend on the application - I only care about frequency reference distribution to SAs and synthesizers, for example, so I don't worry about exact phase and timing between equipment. There is a nice variety of magnetic parts from all types of network devices, including DSL an ISDN. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Linear voltage regulator hints...
If the plan is to use a three-terminal regulator after all, I'd suggest not using a low-dropout (LDO) type if the raw input supply is noisy - the LDOs usually have PNP output transistors (for positive regulators), so may tend to have poorer HF input ripple rejection than equivalent ones with NPN passers. At low frequencies this is no problem since the regulator loop takes care of it, but as the loop rolls off, the PNP becomes a common-base amplifier, allowing more HF from the input to pass on through. I alluded to this in my previous post - from an input HF rejection perspective, it's usually best to use an NPN passer for positive supplies, and conversely a PNP for negative, working as an emitter-follower. If the raw input comes from a switching supply, there will tend to be a lot of HF ripple, so this could be a concern. If this is the case, another option is to have a two-stage regulation scheme with as much pre-regulation and filtering as possible. This of course eats into the overhead budget, so may not be practical in many situations. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Linear voltage regulator hints...
I was just looking at various modern LDOs, and I see that they are greatly improved wrt PSRR - I think the older style PNP passers have been supplanted by new topologies that also even include MOSFETs, so there should be plenty of choices out there. So, I'm changing my recommendation - to avoid using older type PNP output LDOs, while newer types should be OK - just be sure to consider the specs. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] up converting 10MHz to 20MHz
There are a number of ways to multiply-up the 10 MHz, but what about just changing things around to make the 20 MHz VCTXO phase-locked to the 10 MHz OCXO, and then discipline the 10 MHz. There will be phase noise issues, whether doubling one thing, or dividing the other by two, but overall, I think it's easier to divide or subtract with some type of mixing process, and then drive the VCTCXO from the result, forming a PLL. The overall process should be OK as long as the uP gets the right answer on average, over the necessary time frame. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Linear voltage regulator hints...
I'd recommend going with what Bob Stewart mentioned, using separate gates as buffers, operated from a better-grade reference, to shift from the noisier and driftier logic supplies, into the more critical circuits. It's simple, and can be powered from a modest reference circuit. If the logic circuits themselves need better supply noise and tempco performance, don't use any kind of three-terminal regulators - use a good opamp driving a pass transistor. Use a reference IC that has a buried zener for lowest noise - this eliminates all the low voltage references and three-terminal etc regulators that use band-gap references. The down side is that the good kind of reference ICs will need a higher (like 10V and up) operating voltage than may be available, so that complicates it. For a system using a conventional PC-style supply, with +5V and +12V available, an LM399, for example, could run from the +12V, along with the opamp circuitry, while the pass transistor could feed from the +5V, dropping to the +3.3V or whatever low logic supply is needed. For modest current requirement, use only an NPN pass transistor in emitter-follower mode. For higher currents, add another NPN emitter-follower in front of it for more drive - its collector can be supplied from the +12V via some limiting R, to ensure enough overhead. The opamp and associated network resistors, of course, should have performance commensurate with the reference, and sufficient for the application. Since there's also plenty of digital and PS noise around, a lot of bypassing in the right spots should help a lot. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] strange carrier
A signal like that coming from a dish makes some sense to me. I vaguely recall from about ten years ago investigating how the satellite receivers work, that a fairly strong control signal of around 20 kHz was used in some to select the various LNBs and their polarizations in more complicated systems. This was passed via the cables superimposed on the DC power along with the returning IF signals between the set-top box and the dish units. If the neighbor's setup has a bad connection in a cable end, there could be a pretty strong third harmonic of a 20 kHz-ish signal leaking out, with a good-sized antenna possibly formed by maybe 50-100 feet of partly-opened cable shield, depending on the possible ground loop paths. Another possibility is if the LNB power line from the STB has lots of 20 kHz-ish noise on it from a failure in the local SMPS. If the possible faults were large, you would think it would be noticed as a reception problem by the neighbor, but maybe a partial problem is enough for you to see interference. If the interference is from the control signal, it would likely be derived from a uP clock, so quite stable, while if it's from SMPS switching, it should not be very stable, and also loaded with line frequency sidebands. If that is the case, maybe you could inform the neighbor so that they can fix the problem (or you fix it for them), thus improving their reception and reliability, and eliminating the interference. I could be entirely wrong on this, but your last post rang a bell in my head as soon as I saw satellite dish. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] strange carrier
Maybe it's leakage from another time-nut's experiment in the neighborhood, or some commercial equipment. There's probably lots of stuff going on in that area. It's not necessarily a broadcast carrier, but just a frequency that happens to be generated somewhere and getting out big enough - maybe an EMC problem. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Electrolytic Capacitor Question
The Ta caps in old HP gear should last virtually forever, especially if they have already lasted for decades. What you are referring to as wet slug Ta caps are mostly dry solid ones in hermetically sealed cans. There can be some actual wet slug types, but only in certain spots where their unique characteristics are necessary. These can often be identified by the end seals - they are not hermetic, but elastomer sealed with rubbery material, and the anode lead is a solid Ta wire that's butt-welded to a steel wire that's solderable. You can see the weld where it goes from the bluish or brownish Ta color to the tinned lead, except in types that put a glob of epoxy over the end. As I recall, a lot of the ones used by HP are marked 109D type after the original Sprague (or Mallory?) product line. The dry types have a true hermetic glass to steel seal that's soldered to the can and the anode lead, which is steel or copper alloy all the way. This is the best kind of electrolytic cap for lifetime and durability, in my opinion. The wet slug ones can leak after many years, and the sulfuric acid electrolyte can damage things nearby. Whether they can be replaced readily depends on the application. Wet slug types have the lowest DC current leakage and highest temperature range of all electrolytic caps, very low ESR, and wide voltage range. If it's a low-leakage circuit requirement, and fairly low C (like up to a few tens of uF), the best bet may be to use plastic caps, even several paralleled to get the right amount - but the physical size will be much larger. For larger values, this is impractical, so the next closest thing is dry Ta caps, with more leakage, but may be OK for bench use at room temperature. One very common use in HP gear is to have a wet Ta switched across a YTO coil whenever the oscillator is in CW or narrowband mode, for extra noise filtering. Any DC leakage would cause errors in the tuning current. Don't use regular Al or OSCON caps for low leakage circuits. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Lightning arrestors for GPSDO antenna
Of all device types, I think gas tubes are the best for this sort of application - very low C, and high surge current rating. I'm picturing the kind that are used in power supplies and such for limiting line transients - about 1 cm dia and length with axial leads. I don't know what kind are used in lightning arrestors, if they are the same or scaled up in size. Whether you make it able to take a direct hit depends on how big of a hit, your budget, and the environment of the antenna and lines. If it's the tallest thing in a huge field in a lightning-prone area, then it could be a big issue, but I don't think most people have that situation. You may want to look at the US National Electrical Code (NEC) for ideas - I believe that subject is covered there. The main thing there would be safety against injuries and fire, even if the equipment is destroyed. I think what you would want is kind of a pi network - the lowest impedance path to ground at the antenna zone that can be practically realized, then a high common-mode impedance (or even fusible) line to carry the signal to the building, then another low impedance path to ground at the building. This means that in my opinion, you should not put the feedline in metal conduit unless it's essential for protection - or underground, which should improve the grounding. You want the antenna zone to absorb the brunt of any discharge, then use the higher line Zcm to hopefully give some degree of isolation from there to the building. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Wine cooler as temperature chamber
I have this nice little thermoelectric 12-bottle wine cooler (about one cubic foot inside) that I've fixed twice already, and it just crapped out again. It's always the same thing - bad caps in the switching power supply - they're just too small to take the necessary ripple current. So, I could replace them again and be good for a couple of more years, cram bigger caps in there and maybe have a permanent fix, or decommission it from beverage service and convert it to a chiller cabinet for the lab. I'm wondering if anyone has experimented with these things to see how low in temperature they can go. In normal service, the minimum setpoint is 50 deg F, so not all that cold, but I'm sure it can do better than that with a good supply and running full blast. There's about one inch of insulation on all sides, and the door is double-layered glass. There's a circulator fan on each side of the TEC. I would put in a bigger supply and new control system, but it wouldn't be worth it if it can't chill much better than original. I don't know yet if the TEC is accessible for possibly upping the size and rating. I have experimented with R-12 type mini-friges for this purpose - they can typically reach minus 40 deg running continuously, but will be oil-starved at the high vacuum, low flow conditions there, so may not last long compared to normal service. They're kind of awkward and ugly too - the best would be a nice small, glass-doored wine chiller, with a normal refrigeration system built in, but maybe a TEC type would be OK for some uses. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Help identifying a display board
It could be a combined time display with a channel number and measured value, from some kind of data logging instrument. LN was big in thermocouple measurements and the like. With those digits you could show temperature at three digits resolution, selected channel 00-99, and hours and minutes, with the LED for AM/PM. Or NOT. One clue would be to see if they used any of the Panaplex decimal points. None would be needed on time or channel readouts. If only the second DP of the three-digit one is connected, then it could be set to 1 or 0.1 degree resolution. If no DPs at all are used, then the temperature could show 1 degree resolution, and display -99 to 999 deg F or C, which is a pretty decent range for common T/Cs, depending on the application. It could even work for cryogenics, reading say 000-999 deg K. The board also appears to be missing one DM8880. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Any simple way to get 200 MHz from 10 MHz?
The best solution depends on the situation. Is this to build into the unit? Is it to be permanently applied? Is a DIY solution acceptable, and if so, at what level of complexity? If you happen to have a synthesized VHF generator, just run it from the same reference and set for 200 MHz - you'll already know the specs too. If you don't have one, or need one to permanently commit, you may be able to find a used one fairly cheaply, especially since you don't need high setting resolution. For example, a Wavetek 3000 or 3001 may be sufficient - I have a few of these units, and have never paid more than US$20 for one. If you end up wanting or having to go DIY, then I'd recommend building a VCXO with a 200 MHz crystal, then PLL or injection lock it to the reference. If you choose to multiply up instead, there will be a lot of RF filtering needed to get rid of the intermediate spurs, so I'd recommend having at least one crystal filter stage in there somewhere. Either way, a crystal should be involved. A good example of a 100 MHz PLXO is in the HP8566A/B - it is locked to the 10 MHz OCXO, and runs pretty much everything in the instrument. It wouldn't be too big of a stretch to make one for 200 MHz. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Mini Circuits RF TX question
I agree with Tim - a common 10.7 MHz IF can is close enough, and will provide a nice tuned circuit that you can rough-in with an extra cap, and also fine-tune with the coil slug. If you want a fairly wide BW, then an IF can may be too narrow as-is - but it can be broadened with a little extra loss. Either way, I think it would be superior to using a broadband transformer and cap combo. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Mini Circuits RF TX question
I agree with Tim - a common 10.7 MHz IF can is close enough, and will provide a nice tuned circuit that you can rough-in with an extra cap, and also fine-tune with the coil slug. If you want a fairly wide BW, then an IF can may be too narrow as-is - but it can be broadened with a little extra loss. Either way, I think it would be superior to using a broadband transformer and cap combo. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Efratom Rubidium SPTB-100/LN-001 info wanted
I have three of these units, and have been looking for the right manual for years. Thanks for this part, but I believe the SPTB is different from the FRK in the schematics and some of the construction detail - for example, since it's 5 MHz vs 10, that circuit is different, not just the crystal frequency. I have found other things in the PLL too, that probably relate to the LN performance. I have managed to use the *FRK manual, which is close enough to get by - the circuit operation is the same, just the design details differ. I also had to replace a capacitor in one unit that would not lock - it was a leaky ceramic one in the PLL. I had to do a bit of guesswork and shotgunning, since the circuit topology and parts layout were quite a bit different in that area. *I may be confusing it with another model - the M-100, I think, that is similar. This was a few years ago. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Need Fluke 6071A synthesizer info
I recently acquired a Fluke 6071A, and have found some pieces of the service and operating manuals, but not the schematics. Does anyone know where the schematics can be found? This unit looks pretty good except the FREQ and UNCAL indicators are flashing. I managed to decipher enough of the diagnostic error code info to find the sub-synthesizer loop is unlocked, and I hope that adjustment of its VCO will fix it. This section is of course buried inside the inner layers of the RF deck stack, so will be tricky to get at and rig for temporary running while opened up. The manual info is good for figuring this out, but the schematics would of course be a big help for this and future maintenance. Is there any kind of fluke-nuts group on febo.com? I think there's a fluke group at yahoo, but that stupid neo form of groups interface is so screwed up that I'm reluctant to join any more yahoo groups. I will probably have to try anyway for this one. I also vaguely recall that in recent months this model was noted as being pretty good in terms of phase noise for time-nuts purposes, but I couldn't find it in the discussions. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 10811 Outer oven controller schematic
Paul, If the 10811 is that far off, are you sure the oven is working? A few years ago mine had a failure of a particular date code range of the opamp that controls the oven, that were prone to failure.at high temperature. The symptom in the Z3801A was that the outer oven seemed bad - it did not turn on, but it was because it was waiting for the inner one to reach nominal temperature, but it never did. Once you get it all apart, replacement of the IC is no big deal, but what a PITA to get to it. I vaguely recall posting the whole story on that website that has big coverage of the Z3801A - I can't remember the name, since I haven't been there in a while, but it should be easy to find. The website had all kinds of Z3801A info, including a nice writeup on how to take the oven apart, which is where I started. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Lead acid battery noise levels
NiCd batteries should have the lowest noise for their size due to low resistance, but if you look at ever-lower frequency, the Hg should be superior since it has the most stable voltage with time and temperature. Drift (including self-discharge) and temperature variation response can appear as very low frequency noise independent of the other noise sources and operating conditions. Hg batteries are so stable that they were commonly used as voltage references or to power small circuits without any additional regulation needed. Ed Mike Feher wrote: A long time ago, when I was concerned about a phase noise issue, I found an old NBS article. It was on measuring phase noise and included a schematic of an ultra-low noise amplifier. In that amplifier they used Mercury batteries. I also glanced at the referenced article, stating NiCad is the lowest noise, and, NiCads were available for a long time, yet they used Mercury. Regards - Mike ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Lead acid battery noise levels
Second attempt at emailing again: NiCd batteries should have the lowest noise for their size due to low resistance, but if you look at ever-lower frequency, the Hg should be superior since it has the most stable voltage with time and temperature. Drift (including self-discharge) and temperature variation response can appear as very low frequency noise independent of the other noise sources and operating conditions. Hg batteries are so stable that they were commonly used as voltage references or to power small circuits without any additional regulation needed. Ed Mike Feher wrote: A long time ago, when I was concerned about a phase noise issue, I found an old NBS article. It was on measuring phase noise and included a schematic of an ultra-low noise amplifier. In that amplifier they used Mercury batteries. I also glanced at the referenced article, stating NiCad is the lowest noise, and, NiCads were available for a long time, yet they used Mercury. Regards - Mike ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Lead acid battery noise levels
Third attempt at emailing again: NiCd batteries should have the lowest noise for their size due to low resistance, but if you look at ever-lower frequency, the Hg should be superior since it has the most stable voltage with time and temperature. Drift (including self-discharge) and temperature variation response can appear as very low frequency noise independent of the other noise sources and operating conditions. Hg batteries are so stable that they were commonly used as voltage references or to power small circuits without any additional regulation needed. Ed Mike Feher wrote: A long time ago, when I was concerned about a phase noise issue, I found an old NBS article. It was on measuring phase noise and included a schematic of an ultra-low noise amplifier. In that amplifier they used Mercury batteries. I also glanced at the referenced article, stating NiCad is the lowest noise, and, NiCads were available for a long time, yet they used Mercury. Regards - Mike ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Frequency subtraction with D-flip flops
I have been experimenting with three mixing devices: an RF double-balanced mixer (MCL SRA-1B), a DFF (74AHCT74), and an EXOR (74ACT86). The mixer and EXOR give similar results - the sum and difference frequencies, and a slew of various other products that need to be filtered out. What was appealing about the DFF is that I thought it should give only the difference, so the upper sum sidebands should not be possible (except for the harmonics of the difference frequency). When I tried the DFF, this appeared to be the case, but it also contained some of the lower frequency spurs that the other mixers had too, but at much higher levels. I'm not sure if it's due to that metastability thing, or the asymmetry of the input frequency sources, due to their waveforms (one is nearly triangular) or the logic levels. In the time domain, maybe the edges (one or the other) correctly represent the difference, but it looks nasty in the frequency domain, and turned out very difficult to filter out, being below the desired output - I was hoping to only need low-pass filtering to clean it up. The best bet so far seems to be the EXOR, since it provides some conversion gain, compared to the RF mixer, and a reasonable amount of spurious content. I'll be trying various tricks with the DC bias, levels, and symmetry to see if it will clean up a little easier. This section has to provide two outputs - one digital for the PLL, which can be ugly as long as the edges are right to land at the right frequency, and one very clean sinewave to run the PLO. It looks like this will need lots of filtering regardless of the mixing method. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Speaking of Costas loops
Here we go again - the first send didn't seem to get through. This is the second attempt. This talk of Costas loops reminded me of something I wanted to investigate some day. I read somewhere a while back about carrier-phase measurements, and various methods for recovering the GPS carrier frequencies, including the Costas loop, and something with carrier-squaring. Nothing I found showed actual examples or detail of how this is done, only high-order mathematical descriptions. For my needs, I'm more of a frequency-nut - I usually don't care about getting time info, but I'd like perfect 10 MHz for reference. Can using only the carriers lead to simple ways to get the same (or better) frequency stability as a conventional GPSDO, but without the time and location info, or is it pointless to worry about it, and just go with full GPS decoding of everything? Or, is carrier-phase just an enhancement only if you already have the full GPS info? I know that the group could redesign the whole GPS system with tubes if necessary, considering recent philosophical discussions on that, so I think there's plenty of knowledge here about carrier-phase related stuff too. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Speaking of Costas loops
This talk of Costas loops reminded me of something I wanted to investigate some day. I read somewhere a while back about carrier-phase measurements, and various methods for recovering the GPS carrier frequencies, including the Costas loop, and something with carrier-squaring. Nothing I found showed actual examples or detail of how this is done, only high-order mathematical descriptions. For my needs, I'm more of a frequency-nut - I usually don't care about getting time info, but I'd like perfect 10 MHz for reference. Can using only the carriers lead to simple ways to get the same (or better) frequency stability as a conventional GPSDO, but without the time and location info, or is it pointless to worry about it, and just go with full GPS decoding of everything? Or, is carrier-phase just an enhancement only if you already have the full GPS info? I know that the group could redesign the whole GPS system with tubes if necessary, considering recent philosophical discussions on that, so I think there's plenty of knowledge here about carrier-phase related stuff too. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] wwvb 60 khz tuning fork crystals Some insights
Hal Murray said: They make 74xU04 for many values of x. The U is for Unbuffered. They have lower gain in the linear region. I thought they were intended to be used for things like this, but I don't understand that area. Can anybody give me a quick lesson or point me at a good URL? I always thought the unbuffered U versions were preferred for ring oscillators mostly to save power - you don't want the high-drive output stages to be cooking away in linear mode if not needed. The propagation delay can also be less since the U ones have only one stage instead of three (the building block is the totem-pole inverter stage), but they can't drive very much load anyway. I think that most MSI and LSI parts that have built-in ring/crystal oscillator sections use the U topology, but I don't think there's anything special about it - it's the simplest thing that works. I've made quite a few CD4000 and 74HC oscillators, and never worried too much about U versions or not, except for battery-run items where power is critical (or you can run the oscillator at lower voltage). Often they are made from inverting gates that are part of a shared package, where you wouldn't want puny drive capability in the other gates anyway. They are relative power hogs though, whenever linear biasing is needed. Except in the 4000 series, I don't know if U versions are available in anything but the '04 hex inverter, but I suppose it's possible. I think the Schmitt-trigger types like HC14 are necessarily buffered, so have three stages, since you need a non-inverted version of the signal for the positive feedback to the input. I've never tried making one in 74AC - I don't know if it's even possible to bias one up that way without it burning up. I'm working on some related circuits now, so maybe I'll set up an experiment to see how much current it would take for one inverter - I've often wondered about this. I read about this years ago in various CMOS application notes, so I may be missing some key points - there should be plenty of info online. The older generation (when CMOS was fairly new) info may provide more detail about the guts than that related to the newer, higher performance families. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Frequency subtraction with D-flip flops
I am revisiting that tracking generator reference I brought up here a while back, and trying to get my head around how a DFF can take the difference between two frequencies. I have studied and thought about the various topologies and conditions, and searched online for good explanations, but haven't found anything concise that applies to this simple case. During the last discussion on this matter, I learned that another DFF should follow the first, and be clocked the same, in order to reduce effects of metastability. For the actual application, I believe there are limits to the differencing process, so I'd like to check here to see if my thinking is right. First, let's call fd the frequency of the D-input, and fc the clock, which produce a signal fo at the Q output of the first DFF. 1. It seems to me that whenever fd and fc are within a factor of two of each other, either one can be viewed as sampling the other below the Nyquist rate, so an alias signal fo, the difference frequency, is produced. It doesn't matter which input frequency is higher. 2. It seems to me that whenever fc is much higher than fd (fcfd), then it's clearly sampling above Nyquist frequency, so no aliasing occurs - just a delayed (by a fraction of a cycle of fc) version of fd shows up at the Q output, and fo=fd. If so, then this should be true all the way down to fc2*fd. 3. It seems to me that whenever fd is much higher than fc (fdfc), then it's clearly undersampling, so aliasing will occur, producing fo=fd-n*fc, where n is the highest integer that allows fdn*fc. If so, then this should be true all the way down to fd2*fc. For example, if fc is 200 kHz, and fd is 15.883 MHz, then n=79, and fo=83. kHz. 79*0.2=15.8 MHz, so 15.88-15.8=.08333 MHz. If fc=5 MHz, then n=3, and fo=883. kHz. If fc=4 MHz, then n=3, and fo=3.88 MHz. Now onto the second DFF, which reduces the metastability effects of the first. Let's say that normally the Q output of the first goes to the D input of the second, the clocks of both are the same fc, and the Q output of the second is the cleaned up version of fo, delayed by a fraction of a cycle of fc. 4. It seems to me that whenever fd is much higher than fc (fdfc), that fd could be used instead to trigger the second DFF, which would reduce the metastability of the first DFF somewhat, and also synchronize the output signal closer to the edges of fd - but with some metastability from that too. 5. It seems to me that the fastest possible logic family should be used for minimum metastability, even if slower ones can clock easily at fc and fd. So, I'd prefer 74AC-type parts over HC, even at 15 MHz. So, do I get it, or am I missing something? Please be nice. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] wwvb 60 khz tuning fork crystals Some insights
I presume you used the regular 74HC04 or 74HCU04 inverter, not the 74HC14 Schmitt trigger input type?? If the '14 is actually used, that may explain the problems around setting the feedback biasing resistor value - you may be overriding the built-in hysteresis to get it in the linear region. Usually that R isn't very critical with regular crystals, but maybe tuning fork types need more gain, or, if it's actually a '14, the input impedance is probably lower than a regular gate, so it's loading the resonator. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Regulator choices
I believe the original problem was that the raw unregulated voltage may be marginally too high for a conventional three-terminal to take safely. I have often encountered this problem, which is due to the wide input range possible considering the worst-case line voltage tolerance, transformer regulation, transformer selection limits, and possible surge voltages. If you drop the voltage with extra stages or series devices, you may run out of headroom, but if you don't, then it may run dangerously close to the maximum input rating of the regulator. If adding to, or reusing existing power circuits, there's often already some degree of protection from MOVs or gas-tubes, but these are very coarse, so are unlikely to be effective. Whenever I run a three-terminal regulator from raw DC, I put in a series fuse or PTC, and an over-voltage clamp ahead of it to assure that the input rating will never be exceeded. Also, the load may need to be protected - it depends on its characteristics. If the raw voltage is too high, the first step is to add some series diodes that can drop it some - you just have to make sure there's enough remaining headroom at lowest line and maximum load, etc conditions. Be sure to bypass the regulator input with as much extra C as possible to stabilize it, and provide more filtering. In the other extreme, when high-line occurs, there should be a comfortable distance from the raw voltage to the input rating. Then, only transient protection should be needed - heavy zeners or transorb type devices should be enough. If facilities are available, it's best to do this design part empirically, with actual parts, variac, and curve tracer - you can put real conditions on the real stuff. I also always add the usual reverse-protection diode across (O-I) the regulator, even if this fault seems unlikely - I can't count the number of times I've accidentally shorted out the regulator input nodes during design and construction - this would normally take out the regulator if it has lots of output C (almost always). If the peak at high-line is too close to the input rating, then I use an amplified zener clamp, with a big bipolar transistor driven by a big zener of the proper voltage. Alternatively, a traditional SCR crowbar circuit could be used, but I prefer a more transparent, self-resetting solution. Depending on the particular application, my favorite, and the most durable version is the substrate style, where the supply is referenced to a chassis ground or board ground plane. For example, to protect a positive supply, a large PNP transistor is anchored directly to the chassis - I'm talking about using only power devices, like TO-220 and larger - with no isolation, to provide maximum cooling from the collector tab into the substrate (the collector lead should go to the supply common too). The emitter then goes to the +supply, and the zener (typically 1W size) goes from the base (K) to ground (A). A small resistance around a couple of hundred ohms from B-E will make sure it doesn't turn on unless needed. For a negative supply, an NPN would be used, etc. If the supply common is not the substrate, then the same circuit topology can be used, except that galvanic isolation from the transistor tab to the substrate will be needed. Cooling won't be quite as good, but since the circuit floats, either type transistor can be used for any polarity. This same type of circuit can of course be used to protect the load side - this is even easier since the voltage is usually fixed, and the regulator (even if failed) adds to the series impedance limiting the fault current. I also put a large reverse-protection clamp diode on any supply that is exposed to the outside world, or is within a multiple-supply environment, where a fault between any two is possible - during design or testing too, as above, even if unlikely in operation. The transistor SOA rating should be sufficient to trim the peaks during possible high-line and transient conditions, and clear the series fuse if necessary, depending on the situation. Bigger is better for this purpose. The degree of protection and complexity, of course, depend on the criticality of failure, and value of the load. Not much is needed for routine or low power circuits, but for very important stuff, these things can make it nearly indestructible from the powering perspective. Ed Bob Stewart wrote: I also wanted to reduce the amount of power wasted through passive devices. As it turned out, though, I had more tolerance for heat waste than I had thought. But, the general discussion this has become is also good. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] wwvb 60 khz tuning fork crystals Some insights
If you're talking about the email processing, I think something's going on. It sometimes take two or even three tries to get one through, while the most recent one I just sent (about regulators) went OK the first time. Let's see if this one does. B BTW how do you guys get the thread to continue on from the current message? I have only been able to copy and paste parts into the new email, and that loses some of the links and formatting in some way. I don't think this is the cause of the email problem above, since it seems uncorrelated with any editing or whether I send a brand new topic. I'm using an old version of Eudora (7.1.0.9), if that matters. Ed Ed strange no body and you sent it 3 days ago. On Thu, Jun 27, 2013 at 12:13 AM, ed breya https://www.febo.com/cgi-bin/mailman/listinfo/time-nutseb at telight.com wrote: I presume you used the regular 74HC04 or 74HCU04 inverter, not the 74HC14 Schmitt trigger input type?? If the '14 is actually used, that may explain the problems around setting the feedback biasing resistor value - you may be overriding the built-in hysteresis to get it in the linear region. Usually that R isn't very critical with regular crystals, but maybe tuning fork types need more gain, or, if it's actually a '14, the input impedance is probably lower than a regular gate, so it's loading the resonator. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] wwvb 60 khz tuning fork crystals Some insights
Still having email problems - here we go again. This is second try, please excuse if both show up. Hal Murray said: They make 74xU04 for many values of x. The U is for Unbuffered. They have lower gain in the linear region. I thought they were intended to be used for things like this, but I don't understand that area. Can anybody give me a quick lesson or point me at a good URL? I always thought the unbuffered U versions were preferred for ring oscillators mostly to save power - you don't want the high-drive output stages to be cooking away in linear mode if not needed. The propagation delay can also be less since the U ones have only one stage instead of three (the building block is the totem-pole inverter stage), but they can't drive very much load anyway. I think that most MSI and LSI parts that have built-in ring/crystal oscillator sections use the U topology, but I don't think there's anything special about it - it's the simplest thing that works. I've made quite a few CD4000 and 74HC oscillators, and never worried too much about U versions or not, except for battery-run items where power is critical (or you can run the oscillator at lower voltage). Often they are made from inverting gates that are part of a shared package, where you wouldn't want puny drive capability in the other gates anyway. They are relative power hogs though, whenever linear biasing is needed. Except in the 4000 series, I don't know if U versions are available in anything but the '04 hex inverter, but I suppose it's possible. I think the Schmitt-trigger types like HC14 are necessarily buffered, so have three stages, since you need a non-inverted version of the signal for the positive feedback to the input. I've never tried making one in 74AC - I don't know if it's even possible to bias one up that way without it burning up. I'm working on some related circuits now, so maybe I'll set up an experiment to see how much current it would take for one inverter - I've often wondered about this. I read about this years ago in various CMOS application notes, so I may be missing some key points - there should be plenty of info online. The older generation (when CMOS was fairly new) info may provide more detail about the guts than that related to the newer, higher performance families. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] wwvb 60 khz tuning fork crystals Some insights
Email still seems to not be getting through - it seems like sending twice works sometimes. Here this one goes again - sorry if it shows up twice: I presume you used the regular 74HC04 or 74HCU04 inverter, not the 74HC14 Schmitt trigger input type?? If the '14 is actually used, that may explain the problems around setting the feedback biasing resistor value - you may be overriding the built-in hysteresis to get it in the linear region. Usually that R isn't very critical with regular crystals, but maybe tuning fork types need more gain, or, if it's actually a '14, the input impedance is probably lower than a regular gate, so it's loading the resonator. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Frequency subtraction with D-flip flops
I still can't tell if my email is messed up. Last time the second attempt went right through. I'm trying again now. Please excuse if the redundant original shows up too. Original message: I am revisiting that tracking generator reference I brought up here a while back, and trying to get my head around how a DFF can take the difference between two frequencies. I have studied and thought about the various topologies and conditions, and searched online for good explanations, but haven't found anything concise that applies to this simple case. During the last discussion on this matter, I learned that another DFF should follow the first, and be clocked the same, in order to reduce effects of metastability. For the actual application, I believe there are limits to the differencing process, so I'd like to check here to see if my thinking is right. First, let's call fd the frequency of the D-input, and fc the clock, which produce a signal fo at the Q output of the first DFF. 1. It seems to me that whenever fd and fc are within a factor of two of each other, either one can be viewed as sampling the other below the Nyquist rate, so an alias signal fo, the difference frequency, is produced. It doesn't matter which input frequency is higher. 2. It seems to me that whenever fc is much higher than fd (fcfd), then it's clearly sampling above Nyquist frequency, so no aliasing occurs - just a delayed (by a fraction of a cycle of fc) version of fd shows up at the Q output, and fo=fd. If so, then this should be true all the way down to fc2*fd. 3. It seems to me that whenever fd is much higher than fc (fdfc), then it's clearly undersampling, so aliasing will occur, producing fo=fd-n*fc, where n is the highest integer that allows fdn*fc. If so, then this should be true all the way down to fd2*fc. For example, if fc is 200 kHz, and fd is 15.883 MHz, then n=79, and fo=83. kHz. 79*0.2=15.8 MHz, so 15.88-15.8=.08333 MHz. If fc=5 MHz, then n=3, and fo=883. kHz. If fc=4 MHz, then n=3, and fo=3.88 MHz. Now onto the second DFF, which reduces the metastability effects of the first. Let's say that normally the Q output of the first goes to the D input of the second, the clocks of both are the same fc, and the Q output of the second is the cleaned up version of fo, delayed by a fraction of a cycle of fc. 4. It seems to me that whenever fd is much higher than fc (fdfc), that fd could be used instead to trigger the second DFF, which would reduce the metastability of the first DFF somewhat, and also synchronize the output signal closer to the edges of fd - but with some metastability from that too. 5. It seems to me that the fastest possible logic family should be used for minimum metastability, even if slower ones can clock easily at fc and fd. So, I'd prefer 74AC-type parts over HC, even at 15 MHz. So, do I get it, or am I missing something? Please be nice. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] WWVB remodulator for the Spectracom 8170...
I sent this on the 18th, but it didn't show up, so here goes another try. I don't know if something's wrong with my email. Please excuse if the redundant original shows up too. Original message: At low frequency, the HC parts should take very little dynamic current - the big current draw is from biasing gates in the linear region for oscillators and low-level signal triggering, for example. Running them at minimum voltage should reduce the current a lot. Sometimes it's best to separate the linear-biased items and run at lower voltage, or from a slightly isolated version of the main supply, via a resistor (and a bypass cap), to limit the current in case anything gets stuck. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Have 10 MHz need 19.2 MHz
Since the actual crystal frequency appears to be 19.2 MHz rather than 19.5, my original suggestion of injection locking it would have to be modified. For example, you could divide the 10 MHz by 25 ( a single 74HC390 could do it) to make 400 kHz, so 19.2 MHz could sync to the 48th harmonic. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] have 10MHz need 19.5Mhz
Since short-term stability is no problem, and you already have the right frequency in the 19.5 MHz oscillator, you can try injection-locking it to the 10 MHz. Just divide the 10 MHz by 20, and feed the 500 kHz edges into the original oscillator section, and lock to its 39 th harmonic. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 8566B with Ovenaire 10Mhz oven.
Before you tear into the OCXO, have you checked to be sure it is properly connected and set up, and its working conditions are right? There must be a BNC jumper on the back to connect the 10 MHz, and the reference selector switch set to internal. Furthermore, the power to the XO section is not turned on until the OC is warmed up, so a quick checkout may wrongly point to OCXO failure. There is an indicator LED in the power supply circuitry that shows when the XO power is on. You have to leave the 8566B plugged in to the mains for at least ten to fifteen minutes to be sure it will go. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 8566B with Ovenaire 10Mhz oven.
When the unit is up and running, you won't be able to tell if the oven is working from touch - the fan blows right at it. If you open it up and run in standby for a while (an hour or so), it should feel a little warm. Also, of course, check the oven power supply, which should always be on whenever it has mains power, and the oscillator supply, if it comes up. I presume you have the manuals, which are available at agilent.com. You will definitely need them. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] What was 50.791095 MHz used for?
Does anyone know what 50.791095 MHz was used for? I assume it was some sort of telecom clocking rate - perhaps obsolete. I serendipitously discovered that I have some VCXO modules of this frequency, that will likely solve a problem on one of my projects. I'm now curious about what they may have been used for, and if that is a standard frequency that will be available for future use. Online searching seems to indicate it is very common, but doesn't say what for. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Downsizing dilemma, HP 3335A
You don't save these kinds of synthesizers for high frequency coverage, but for their 10 to 11 digit frequency resolution. If you anticipate needing that, then of course they should be kept and fixed. The long-obsolete telecom standard connectors and ranges are pretty much useless - sacrifice that one first if you need parts for the others. If you need to justify keeping them, you can use them for practical everyday applications. For example, each one can store a telephone number - as long as the power doesn't go out. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] aging/failure of un-powered xtal oscillators?
If there are plenty of them, I'd recommend dissecting one to see how well they're made, and what sort of components are inside. If the crystal is hermetically sealed, it's probably OK. If there are any aluminum electrolytic caps in there, they may be no good. Most other parts from that era should be OK. If possible, you should pretest a bunch of them to see if they seem up to snuff. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Trimble Thunderbolt, any easy way to create 500 MHz reference from one?
I assume the question is about going from 10 Mhz to 500 MHz. The possible solutions depend on how clean the result has to be. For counting grade use, this has been done in a number of test instruments like the HP5345A and HP5370A that I'm familiar with, and certainly others. If you look at the multiplier or PLL systems in equipment that does this, you can get an idea of what's involved. It takes quite a bit of circuitry to provide a really clean output. If you already have such an instrument in service, you can modify it to tap off the 500 MHz, buffer it up, and provide an external connection for it. This won't be enough if you need a dedicated or more compact built-in source, but if you can find a carcass of one of these instruments, or just the appropriate boards, you can build up a unit that should be quite good. Of course, any synthesizer that covers that frequency could do the job directly. One of my favorite oldies is the Wavetek 3000 or 3001 that reaches 520 MHz - I have acquired a couple over the years for very cheap. These are tall, half-rack width, so not very compact, but could be cheap enough to dedicate to an application. If it must be very small and modern, you should be able to find off-shelf a complete phase-locked fixed 500 MHz or programmable module, ready to go, but it will be quite expensive new. If you want to start from scratch, I'd recommend looking for an off-shelf SAW-resonator based VCSO from an RF vendor (expensive new, but maybe can be found surplus), and a PLL IC such as those available from Analog Devices and National (now p/o Texas Instruments). Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] hp5065b !!!
I'm interested in these improvements too, and hope some can be applied to other models. I don't have a 5065A, nor do I expect to ever get one, but I do have three identical old Efratom units. Whether it's worthwhile to modify any Rb units also depends on whether it's possible to rejuvenate their Rb lamps, as discussed here a number of times, so that they can be maintained virtually indefinitely. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Photodiodes for high frequency OPLL
I don't think that you can effectively directly mix two laser wavelengths in a semiconductor light detector and get a useable IF - it's hard enough just to get the tens of GHz modulation signals out above the noise floor, let alone a tiny difference signal between hundreds of THz. You need an optical interference or nonlinear device up front to do the mixing and get the wavelength discrimination, while the optical detector(s) serve as the first IF O-E transducer. My knowledge of this stuff isn't up to date - maybe nowadays there are detector devices and methods that take care of this directly, but I don't think so. Most really high speed diodes are optimized for the 1550 nm region where EDFAs work, but maybe they have usable response at other ranges. It depends on your particular application and wavelength. I think detectors are usually specified over the entire IR region, so datasheets may tell enough. Here's link to some good info, but not current state of the art: http://e-collection.library.ethz.ch/eserv/eth:28429/eth-28429-02.pdf There are various methods that use lower frequency modulation techniques so that regular detectors can be used directly. If you study up on related patents, you may find some ideas and leads to appropriate actual devices. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Photodiodes for high frequency OPLL
Ooops - never mind. I wrote before my memory was updated. My experience in E-O stuff was years ago using AM at relatively low frequency, and nowhere near the lasers and microwave/gigabit/sec stuff - I didn't think the detectors were fast enough to actually keep up with the optical carrier frequency. I was also picturing wavelength/spatial separation with interference in order to allow relatively slow detectors to see it, or mixing in nonlinear optical materials. Ed Bruce Griffiths wrote: A photodiode is in fact a nonlinear device for optical fields as it is essentially a linear optical power detector. The output is proportional to the incident optical power not the field amplitude. Photomixers are routinely used in wide range of diverse application such as translating the frequency fluctuations of the (Mie) scattered light due to Brownian motion of the colloidal particle sizes to baseband. The size of the scattering particles can be inferred from the shape of the resultant frequency spectrum. An interferometer of itself (without a detector) is a linear device that merely superimposes optical fields and will of itself produce no difference frequency output. Bruce ed breya wrote: I don't think that you can effectively directly mix two laser wavelengths in a semiconductor light detector and get a useable IF - it's hard enough just to get the tens of GHz modulation signals out above the noise floor, let alone a tiny difference signal between hundreds of THz. You need an optical interference or nonlinear device up front to do the mixing and get the wavelength discrimination, while the optical detector(s) serve as the first IF O-E transducer. My knowledge of this stuff isn't up to date - maybe nowadays there are detector devices and methods that take care of this directly, but I don't think so. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Need info on Trimble 4000S GPS Surveyor
I recently acquired a junker Trimble 4000S GPS surveying unit. It's mid-1980s technology, so very big, but nice to salvage various RF and signal processing goodies from. I have no plan to get it working, and no need for the function - it's just for parts/subsection use. I found that this unit is way beyond obsolete, and found no support info on-line. I'm just wondering if anyone has or knows of info on it - even a block diagram would help, or anything to fill in some blanks. I've saved the RF/PLL modules intact, and recorded the interconnections and supplies, so I can (eventually) figure out what everything does. This unit uses both the L1 and L2 carriers, and has eight canned function modules plus an OCXO to do the front-end work. Other than the power supplies, the entire interconnection appears to be just two IFs and two references going to the control/DSP board assembly (four huge boards), and a 16-bit parallel DAC drive from there to control the EFC line to the OCXO. The L1 1575.42 MHz chain uses a 16.368 MHz VCXO locked to the 10 MHz reference, running an LO of some integer multiple that results in a reference around 38.4 MHz labeled ECL 38.4 F0 on the main board, and an unlabeled signal IF called TTL LIMITER. Internal markings 768 and 384 may indicate PLL IFs of 76.8 and 38.4 (76.8/2) MHz. The L2 1227.6 MHz chain uses a 28.644 MHz VCXO locked to the 16.368 MHz reference, and LO that results in another unknown IF that runs through a similar TTL limiter. It appears that the LO is an integer multiple of the 28.644 MHz, with a PLL IF possibly around 59.2 MHz, marked 592 FO. Only the unknown signal IF from this section goes to the processing boards - no PLL IF seems to go beyond these modules. The unknown signal IF goes only to one of two apparently identical DSP boards, unlike the others that all go to the main board. The L1 downconverter appears to use quadrature mixing, but I can't tell what happens after that - the I-Q signals go into a bunch of baseband circuitry. The L2 one also has a quadrature mixer, but only one output goes into its baseband circuits - the other is just terminated. As I understand, the L2 is always encrypted, so useless for data, but its carrier can be used to enhance overall accuracy - I recall studying that a few years ago, but forgot the details. So, maybe the L2 portion is only for carrier recovery of some sort. So, here's what goes between the RF section and the rest: RF to main board: 1. Main reference 16.368 MHz 2. L1 DSP reference approx 38.4 MHz 3. L1 IF unknown frequency RF to one DSP board: L2 IF (or recovered carrier?) unknown frequency Main board to RF: 16 bits DAC EFC tuning for 10 MHz reference Antenna to RF: Input to diplexer, with +15 VDC feed for preamp I'd appreciate any info or ideas on deciphering the rest of the way - maybe the modules will be useful for something as a system, rather than just parts. I'm especially interested in GPS carrier recovery techniques for frequency only - not time. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Off topic project sort of heart rate monitor NEED BEATS PRE MINUTE TO ANALOG VOLTAGE
If you only need cardiac pulse rate, an optical pulse-oximeter type interface - but without the oximetry part - would be much simpler than an ECG system, and it doesn't need galvanic connection. If you need more than that, then ECG methods should suffice. Back in the 1970s, there was a lot of interest in biofeedback and TM techniques, using ECG and EEG. Common feedback signals were sound and light related to the detected physiological signals. A chart display is a good visual feedback with dynamics too, and it records the events for further study, so you can investigate the QRST details and whatever else is interesting. Detecting and processing the signals is nearly trivial now with modern electronics - it wasn't even that hard back then. If you study what's transpired over the last forty years, you'll get a pretty good idea of the whole history of the field, and the technologies involved. Nowadays of course, the PC can take care of most of the processing and feedback/indicator functions. For a DIY, I'd recommend for each channel run the high-gain instrumentation amplifier into a V-F or FM converter to translate the signal into a workable frequency range like a few hundred Hz, then use a sound card and SW to post-process. Copy front-end interface ideas from the real stuff - professional medical equipment - to make sure it's safe to hook up to your own hide. If you don't need absolute details, you can AC-couple the physiological signals into the V-F process with appropriate time constants, to avoid the drift, and average in SW. I doubt that interest in this area has waned over the decades, so it's likely there are all kinds of cooler, newer stuff available off-shelf, so you may be able to find it ready to go - and even way beyond what you had in mind. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] VCXO in a watch timing machine
How do you know that the frequency range is incorrect for the function? Assuming this is for calibration of wristwatches, maybe the idea is to set them to some nominal value at room temperature, then expect them to drift to the right frequency at skin temperature or so. There would have to then be some relationship between the approximately 18 MHz and the ideal 32,768 Hz watch crystal frequency. If the idea is to divide the 18 MHz down to an accurate version of 50/60 Hz or 1 Hz, then yes, it looks like something's wrong. But, does it need to be a VCXO, or just settable to the right frequency? I believe that 18.00 MHz is a standard frequency for VCXOs and TCXOs, so readily available from typical vendors, or you could build an oscillator as good as needed. The package dimensions seem like a standard 14 pin (only 4 are used) DIP style - very common. I think most of those modules use a limited tuning voltage range like 0-5 V if the supply is +5 V. If you haven't already, check to be sure the full tuning voltage range is covered at the module. You can usually get a little higher frequency by exceeding the tuning voltage - especially if there's nothing to lose anyway if it's damaged - but unfortunately it's hard to get lower frequency since the varicap diode will reach zero or forward bias if you go very far the other way. Also, if the supply voltage is off, it could have drastic effects on operation. Changing the supply voltage a little can affect the frequency and tuning range, so this is another option for slight adjustment - but may give unpredictable results. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] What cut do I have?
If you meant to say the frequency starts out too high when the oven is cold, then I think it's AT cut. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
I tried to send this message on Sunday, but for some reason it didn't go through, so here it is again. Please excuse any redundancy if the original shows up. I will have an update of the project soon. Hi Bob L., Your suggestion of the 300/953 scheme was inspiration for what hopefully will be the simplest solution of all - I've started building it. First, I should clarify more, that the original scheme actually has three phase-locked loops - a 10.7 MHz, a 10.0594 MHz, and the final one, 1207.133 MHz. The last one is a PLO brick that just multiplies any RF input by any n within reason to phase lock the microwave output (to nth harmonic of input). I wasn't counting that one, since it's more or less a fixed function, but it's a variable (arbitrary n) in the numbers game. So, when I was referring to getting rid of one PLL, I meant not needing to produce the intermediate 10.7 MHz, since the 953 gives a rational number solution directly from 1 or 10 MHz - this is the single PLL scenario. I tested the PLO and microwave section with 15.88333 MHz = (10 MHz/600)*953 from a synthesizer, and it worked just fine. The PLO is tuned near 1207 MHz, and uses whatever n lands it within lock range, so n=76 in this case. If you adjust the cavity, n can just as easily be 75 or 77, with different output frequencies, or a number of numbers that satisfy the bounds of operation. So, the trick is to produce that one correct frequency from the 10 MHz reference, cleanly enough to get the job done, and feed it to the PLO - the n value takes care of itself. The way it's partitioned now, I will have one can containing the 15.883 MHz VCXO (74HC86 and a 16 MHz ceramic resonator), two LAN LPFs, a 74HC4020 feedback divider (1/953), and a CD4046B PLL. A second can, which is needed anyway for handling the various external and internal 10 and 1 MHz references, will not only route and scale, but will also include the divider to make the 16.6 kHz (10 MHz/600) reference for the other box. So, the overall synthesis chain is (10 MHz/600)*953*76=1207.13 MHz. Pretty simple. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
Yes Robert, the 59. kHz is effectively added to the 10. MHz, but not by direct mixing. The 1 or 10 MHz reference drives a D-flop flop, which samples the 10.059444 MHz, leaving the difference frequency 59. kHz, the feedback signal in the second PLL. The direct way to do it would be with very accurate, full I-Q mixing to get only one sideband, but that gets very complicated. The current scheme is simpler, and works quite well. The main pieces are decade dividers (74HC390), dividers for 107 and 180 (74HC393 or 74HC4040 each), a D-FF (74HC74), a CD4046B for each PLL, and gates for the oscillators (74HC04 or 74HC86). Using an '86 allows for getting push-pull output with equal prop delays, in case I need to run it through some differential LAN LPF modules that I have on hand. It took some effort to come up with workable numbers that all fit within the constraints, but I'm sure there are many other undiscovered sets that would do it. I like your single-PLL 300/953 idea - it may be doable within 74HC speeds, and I think ceramic resonators are available at 32.0 MHz. The PLO would like the much higher reference frequency - I think any n from 8 to 150 or so will work. Scaling that by two to 600/953 , making 15.88333 MHz, with a 16 MHz resonator (I have some), fc=16. kHz, and n=76 should work too. It would be OK with 74HC for sure, and it would just fit through the LPFs, which cut off at 17 MHz. The comparison frequency fc is getting kind of low, but may be OK, depending on how much near-in phase noise I have to contend with. That was one of the reasons I opted for the two-stage approach - to avoid having a very small fc, or dealing with fractional-n ripple. I will investigate these possibilities and put together some experiments. I can directly drive the microwave section from an external synthesizer to try various reference frequencies. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
Bob, please tell me more about cascading the DFFs. I was only using one half of the '74, with the other inactive, so both are available for the task. From your description it sounds like I just run the Q from the first DFF to the D of the second, clock them together from the 1 or 10 MHz, and take the cleaned up difference signal from the Q of the second. So, I think what it means is that the same information should pass through, just delayed by one sampling clock cycle, and scrubbed of any edge uncertainty of an analog nature, that would otherwise be passed to the phase detector. Right? I would definitely do this if no additional logic packages are required. If a single-PLL type we discussed earlier is workable, I won't even need to worry about the second PLL system. It all depends on whether the phase detector frequency will be high enough. I'll be thinking through that and trying a few experiments. It's simple enough that I could even just build it and see what happens. If it's not right, then I'll just go with the previous plan, with high confidence - and a two-stage sampler. Regarding the oscillators - yes, having different signals present in common packages is what got me into this trouble in the first place. As I mentioned earlier, I had optimized the original design for compactness and minimum package count, so I had every signal in the box going every which way, all mixed up. I had used a different method for making the 10.7 MHz though - building it up by mixing various divided frequencies, then filtering it with cascaded 10.7 MHz IF filters. Most of the stuff went right around the filters anyway, since there was so much whizzing around in there. In case anyone is wondering why I'm so hung up on this 10.7 MHz thing: For this particular tracking generator project, I just need to synthesize one fixed, correct reference frequency with the simplest, most compact scheme that performs well enough. The original design evolved from using the 10.7 MHz base frequency, but it isn't actually needed per se. If anyone comes up with sets of numbers that seem to work in a single-PLL scheme, and fit the constraints evident in this discussion, please let me know. I have other tracking generator projects in the works though, that will cover most or all of the 8566B span of ~0 to 24 GHz, and need to produce various numeric and harmonic relationships for IFs and frequency control - all of these can be readily integer-derived from the fundamental 10 and 10.7 MHz references. In all cases, the ultimate reference is the 10 MHz used or produced by the 8566B, so everything is phase locked. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] How far can I push a crystal?
I've got to make a very clean 10.0594... MHz VCXO for a redo of one of my old circuits. I previously used a 10 MHz ceramic resonator, which was easy enough to push around in frequency. Of course, I have a couple dozen of those somewhere, but can't find them now that I need them again. I figured I'd just pull the ones out of the old circuit, but since I did find a whole bunch of 10 MHz quartz crystals, I'd like to revisit whether I can push one of those that far with decent results. As I recall, the results of my previous experiments in doing this were less than satisfactory, which is why I went with the ceramics. This would be a change of 60 kHz out of 10 MHz, or 0.6 percent - a helluva lot for a crystal. The frequency will be exactly phase locked to a reference. It doesn't need to have extremely high in-circuit Q or long-term stability - just tunable to that magic number - the PLL will do the rest. A conventional varicap circuit will provide the VCO-ness, while the tuning range just needs to be enough to accommodate drift and the initial setting. The power gain element will be a 74HC04 or 74HC86 section. The PLL reference will be 59.4... kHz - way above the necessary loop BW. Has anyone successfully pushed a quartz crystal that far off, with reliable (still sort of a sharp resonance) operation and no spurious modes? Any ideas? If this isn't practical, I'll just go back to the ceramic resonator (which worked just fine), but I'd like to settle it once and for all. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
Bob, are you saying they have 10.059 MHz crystals? I've never seen one anywhere, or anything even close. Ed Hi Mouser shows 16 items tighter than +/- 20 ppm accuracy. Six of them are in stock and less than $1 in single piece quantities. The cheapest is 39 cents. Bob ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
Maybe I should clarify what I meant by pushing the crystal frequency. I meant only using various topologies and electronic components in the associated circuitry, that would detune it from its natural resonance far enough to reach the new frequency, and still have it be sort of a narrow-bandwidth crystal oscillator - not doing any mechanical changes to the crystal element itself. Since the ceramic resonators seem to work well, and can be pushed (or pulled?) fairly far away by proper selection of the associated component values, I was wondering how far quartz crystals can reasonably go. I encounter this situation often - needing an oddball frequency, but preferring to use common or standard parts. The nominal choices in ceramic are quite limited, while in crystals, there are many more - but few ever seem to land at or near enough to a frequency I need. The only thing I have thought of so far is to maybe add some series R to drop the crystal Q, so broadening the resonance, and just dragging it up by extra series C, but at some point there's no point to even having the crystal there at all. I'm just trying to figure out what's possible and reasonable. I know that I can get any custom frequency by spending enough money, but that takes the challenge and fun out it sometimes. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] How far can I push a crystal?
Thanks all, for the feedback on this issue. In summary, I got these points out of the discussion on crystals: 1. The correct terminology is pulling the frequency. 2. Getting beyond about a few hundred ppm from the nominal frequency ranges from very difficult to pointless. 3. It's easier to pull down than up. It looks like it would not be worth fooling around with crystals, so I'll just use the ceramic resonators. By the way, I just tonight managed to reach the correct geological layer of stuff out in the garage, and found the missing 10 MHz resonators, and a whole tray of other parts that were in reserve for completing this project from a couple of years ago. For the curious: The 10.059... MHz is made by a PLL using the 59.... kHz reference, which is 10.7 MHz divided by 180. The 10.7 MHz is a from another VCXO (which can use a standard crystal, ceramic resonator, or ceramic IF filter - easy) that's phase locked to a 10 or 1 MHz reference, using two fixed dividers. The 10.059... MHz is used as the reference for a phase locked microwave brick oscillator, using n=120, to make 1207.1333... MHz, which is exactly one-third of 3621.4 MHz, the low-band upconversion IF of the HP8566B spectrum analyzer. The 1207.1333... MHz is harmonically mixed (m=3) with the first LO of the SA to produce the tracking signal centered in the passband of the SA. All of this is built into the modified carcass of an HP8443A tracking generator, originally built for older SA models. Using the new stuff, plus parts of the 8443A, the net result is a 50 kHz to 250 MHz tracking generator, with power up to +10 dBm, leveled within about 1 dB, and with 130 dB step attenuator range - very nice for low RF and baseband work. The 10.059... MHz is only one of many frequencies that could be multiplied by various n-values to give the same result, but it was chosen because it was very close to a standard frequency available in ceramic resonators, high enough that n didn't need to be too large, and it could be synthesized with a very simple PLL system. I had all of this built and running, but I had made the fatal engineering mistake of putting way too much stuff in too small a space. Space was tight, so I squeezed the entire LF control system and synthesizers into one small can, and necessarily optimized for minimum IC package count. Then I found that there was too much crosstalk between virtually all the signals in the box, so there was too much phase noise to work at 300 Hz and less IFBW. The problems were irreversible - sharing IC packages for multiple signal processing was an especially bad move. After many hours of rearranging signal paths, adding shielding and grounds, and changing topologies, I concluded that I had to rebuild it the right way. So here I am. The two main frequencies will be generated in separate boxes, and no ICs will contain multiple signals that aren't being processed together. This time I'll get it right, and finally wrap it up. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] hammond 77589
Since you measured the same R on opposite ends, it appears to be a center-tapped inductor. Since there's no input signal other than power in the schematic, it must be a low frequency oscillator. If you draw in the inductor on the schematic, it should make more sense. It could be that they used an AC bridge for the temperature control, so it was used as a clean, regulated-amplitude, low frequency excitation signal. AC excitation would have been easier to process back then - easy to amplify up with very high gain, and synchronously demodulate to get the feedback/control signal. DC at high gain and low offset was trickier, and often would have used a chopper-stabilized amplifier - essentially making it an AC system. This just skips the chopper part on the input end. You should find that the signal from this circuit goes to the thermistor bridge, and also to some circuitry following the amplifier to provide the demodulation. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] hammond 77589
I just noticed one more thing - if the NTC thermistor in the schematic is the actual one that senses the oven temperature, then the oscillator and bridge are one in the same. If this is the case, then the oscillator amplitude could be proportional to bridge imbalance - if the NTC is being balanced by the 1414 ohm resistor. I doubt that this is the case though, since the DC resistance of the inductor would swamp the others, so it would need to be AC-coupled into the bridge to isolate its resistance, but the schematic shows direct connection. Also, you would need the phase information to know which direction is needed to correct it, so there would have to be another signal pickoff from this circuit to get the demodulator phase info. If the thermistor is just sensing ambient temperature of the control circuit, then it's probably compensating for circuit drift and oscillator amplitude. I think this is the more likely setup, given the schematic info. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Counter OCXO behaviour
Is it manually adjusted right at the oscillator? If so, just opening it up and sticking a screwdriver in there gives it a thermal shock, and the adjusted element will have mechanical stress that has to settle out too - the value can change for a while. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Need info on HP 1 MHz ovenized XO
I want to modify an old HP XTL oven for remote voltage tuning. The HP part number is 0960-0079-1, made by Ovenaire (model OSC 49-14). These were used in HP8443A tracking generators and probably other models too. It puts out 1 MHz, and is powered from +24VDC wrt the case ground. On opening it up I found that the circuitry includes a 74LS73 dual JK FF, and a 74LS140 - very obscure - apparently a dual 4-input gate of some sort. Since there are two FFs available, it appears that the actual frequency of the crystal could be 2,3, or 4 MHz. I'm guessing that it's 3 MHz since there wouldn't be a need for extra gate decoding if it was just a divide by 2 or 4. I will measure and see, but I'd appreciate if anyone has any more info about the guts of these oscillators, so I'll have more to go on for modification and long term maintenance. The fine frequency adjustment is done with a ten-turn trimmer pot, so converting it to remote operation should be fairly straightforward. The divider circuit board is outside of the insulated XO oven, but still goes within the case. It also has the external interface connections and the fine trimmer, so easy enough to get at. I hope I won't need to modify the actual tuning range - I don't want to crack open the oven part. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Need info on HP 1 MHz ovenized XO
Sure enough, the XO runs at 3 MHz. The fine tuning pot runs from ground to about +6.8 V, which appears to be the main regulated supply. I just have to add some feedthroughs to connect up the new stuff. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Need info on HP 1 MHz ovenized XO
Thanks Nigel, The output comes right from the '140, via a small Ta cap and series R, so it makes sense that they used a line driver - they got the drive capability and the gate decoding for divide by 3 in one package. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Vectron GPSDO Oscillator stranges...
Can you completely swap the Vectron crystal oscillator modules between the Datum units? Maybe the fault lies in the other part. If they are GPSDOs, each must have a DAC somewhere driving the tuning control line to the VCXO, The 10 kHz may be the DAC serial data rate, or a PWM rate for fine tuning. A fault in that area could cause the sidebands. Maybe you really did hear data. If you can hang a fairly large (several uF or more) plastic capacitor on the tuning line without causing the whole thing to oscillate, you may be able to observe a decrease in the sideband amplitude - a sure indicator that the tuning signal carries the problem. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Vectron GPSDO Oscillator stranges...
Oops - regarding that capacitance test on the tuning line, I meant to say up to several uF or more. You can start small to see if there's any effect. It all depends on the impedance of the tuning line circuitry, and the existing amount of filtering - you may need quite a bit of C to swamp it out and show a noticeable effect. You're not looking for necessarily normal or perfectly-settled operation under the test condition, just the relative effect on the sidebands. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS Patch Antenna Electrode Tarnish
It should have virtually no effect. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] (no subject)
Yes, that's true, Simon, but remember the initial goals of simplicity and long term phase coherence, while jitter doesn't matter so much. The longer term average frequency ratio should be right on, while comparing at any particular cycles it would be awful. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 10 MHz - 16 MHz clock multiplier
I'm kind of late to the party on this one, and don't want to reopen and send it off on a tangent. I agree with the injection-lock method, but just want to suggest that since most uPs and uCs have their own CMOS oscillator built in, the simplest solution would be to just use that as intended, with a cheap 16 MHz crystal or ceramic resonator, and injection lock it to a low impedance version of the 10 MHz clock, via an RCD network. It may need a variable cap to tweak the resonator to the lock frequency. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] 10 MHz - 16 MHz clock multiplier
Actually, I was referring to an RC and Diode network in anticipation of the possible need for more signal shaping flexibility, depending on the signals and circuitry. The built-in oscillators are usually self-biased CMOS inverters intended to go with crystals, and usually a couple of small phase shift caps to ground. If you inject the right amount of reference frequency at the input, the oscillator should sync up. Since it's for a fixed frequency, the required lock range can be quite small - it needs to be enough to compensate for drift in the resonator and circuits - so the injection level probably can be small. Since the oscillator input has plenty of gain, and the reference is likely a very low impedance, I think a fairly high impedance passive coupling network to link the two should suffice, without any extra active circuitry. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] YIG oscillator drift question
Yes, something must supply the minimum magnetic field to activate the oscillator - but the current determines the field, so it's necessarily part of the drift characteristic. It also depends on whether the YIG sphere is heated - usually to 80 deg C or so, to help stabilize it. Without any other frequency control, you can expect it to stay within about +/- 10 MHz of the frequency predicted by its tuning curve. It is very linear with magnetic field, hence tuning current, until the core material begins to saturate at high flux levels, so the tuning current needs to be increased to compensate. The most common type is one-octave 2-4 GHz, used in spectrum analyzers from the 1970s on, and also the first microwave band of many generators. Another common one is the 2-6.X GHz also used for the same purposes, but covering a wider direct range. The next bands up are typically 4-8, 6-12, 8-12, and 12-18 GHz, used in generators, at least in older gear. You can expect to find about +10 to +18 dBm output power range, and fairly flat with frequency. The minimum setup is one or two power supplies, plus a variable coil current driver up to one amp or so. You don't need the heater to run it. Don't bother with the FM coil - it's not needed for basic experimenting, and it's easy to burn out if you screw up. The main coil is more robust, but it can be burned out with too much current, so don't go too far beyond the maximum tuning current spec. It can also shock you from inductive kickback, so the coil needs clamping too. It's best study the circuits of existing sweepers and such to get ideas on how it's all done. Your best bet is to get an old HP8620 or Wiltron 610 sweeper with some plug-ins. They are simple enough to dig into the guts and modify/experiment. These are regular old analog sweepers with no synthesis - look at their specs to see what kind of stability can be expected. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] GPS over fiber
That would certainly provide good distance capability and isolation, but there may be a severe penalty in signal to noise ratio (and some non-linearity) due to converting to optical and back. It may be worthwhile considering the specs of the overall E-O link versus the extra gain and SNR or NF of conventionally amplifying and driving longer coaxial cables. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Looking for info on an old WWVB receiver
I recently picked up an interesting early 1970s vintage WWVB receiver, Model 630, made by Specific Products of Monrovia, CA - that's what the adhesive sticker on the front says, and the name 1 MHz Time Base Calibrator (Utilizes WWVB accuracy of 2 parts in 10^11). There's also a pair of banana jacks labeled 1 MHz Input, a row of incandescent lamps for a signal strength indicator, and a power switch. The back says Model LF 60S, and has six RCA jacks for 100 kHz Output, 60 kHz Output, Recorder Output, Antenna Input, (divide sign) 10 Output, and Time Code Output. There's also the line cord and a +12VDC output RCA jack. I'm wondering if anyone knows anything or sources of info about this thing. With all the recent talk of WWVB changing to spread-spectrum, it may be useless anyway, except for some parts, but I'm curious about whether it's worth saving. Ed ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.