Re: [time-nuts] FE-5680A OSMT connector / RS232
The trip point is specd to be between -3 to +3, which is why 5V works, especially since the trip point is usually positive. I would set it around 1.5V. The transmitters are required to swing +/- 5V. I'm not sure the 232 allowed lower voltage as much as all we had to work with were charge pumps, first off a 5V rail and later 3V. Basically the voltage was dropped closer to the spec limit to ease the design, even if some non-compliant parts wouldn't work on the lower voltages. On 12/9/2012 11:25 PM, Hal Murray wrote: ja...@peroulas.com said: I'm not able to get it to respond to the 'S' command and when I measure the voltage on the RS232 TX pin (#2 from the left) it's always 0v. Shouldn't it be -12v when idle? Newer RS-232 allows 6V rather than 12. In practice, it's not all that uncommon for designers to save a chip and just send 5V CMOS signals. That works fine for short distances. If you can get a scope on it, sometimes embedded boxes send out a hello message at power up. I'd also check pin 3 in case you or they got things swapped. ___ 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] Sir Patrick Moore
Related to time-nuttery as astonomical observation was used for time-keeping until C20. Sir Patrick Moore, the great amateur astonomer died yesterday at the age of 89. A gentleman, astronomer, and true scholar. His death was predicatable as he's been very ill for a long time, but nonetheless a great loss. Regards, David Partridge ___ 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 DCF77 master clock data
I am logging the data from this strange master clock, and the none of the bits 42 through 44 are stuck. Good idea, though. I have just checked tonight, and the DOW is indicating as 5 (Friday), despite it only being Monday. Good idea though. On 9 December 2012 23:24, mike cook mc235...@gmail.com wrote: Le 09/12/2012 11:28, Tom Harris a écrit : Greetings Time Nuts. Time related, but unusually so: I am examining the DCF77 (see http://en.wikipedia.org/wiki/**DCF77http://en.wikipedia.org/wiki/DCF77for a refresher) data output from a master clock intended to drive a series of slave clocks operated from 24V DC, interrupted by 100ms or 200ms pulses to signal '0' '1' on the data. What I am seeing is that the day of week bits 42 through 44 inclusive do not code the day correctly, they seems to have a variable offset from the actual day, I have seen an offset from +1 to +3. The rest of the data, ncluding parity bits, are OK. Today, 9/12/2012, the day is Sunday, which should be coded as '7'. However I am seeing a value of '3', which is Wednesday. Is this normal, do wired DCF77 slave clocks transmit munged dow data to prevent them being used with someone else's master clock? If they do I am hosed since I am trying to reverse engineer the master clock. Is it just bit 44 being stuck on zero? or can each dow get varying values? -- Les chiens aboient, et la caravane passe. __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. -- Tom Harris celephi...@gmail.com ___ 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 DCF77 master clock data
That's the funny thing, the parity bit is correct for the value of DOW, just that the DOW value is wrong. On 9 December 2012 23:27, Azelio Boriani azelio.bori...@screen.it wrote: Is the parity bit P3 consistent? It protects the day of the month and that of the week. On Sun, Dec 9, 2012 at 1:24 PM, mike cook mc235...@gmail.com wrote: Le 09/12/2012 11:28, Tom Harris a écrit : Greetings Time Nuts. Time related, but unusually so: I am examining the DCF77 (see http://en.wikipedia.org/wiki/DCF77 for a refresher) data output from a master clock intended to drive a series of slave clocks operated from 24V DC, interrupted by 100ms or 200ms pulses to signal '0' '1' on the data. What I am seeing is that the day of week bits 42 through 44 inclusive do not code the day correctly, they seems to have a variable offset from the actual day, I have seen an offset from +1 to +3. The rest of the data, ncluding parity bits, are OK. Today, 9/12/2012, the day is Sunday, which should be coded as '7'. However I am seeing a value of '3', which is Wednesday. Is this normal, do wired DCF77 slave clocks transmit munged dow data to prevent them being used with someone else's master clock? If they do I am hosed since I am trying to reverse engineer the master clock. Is it just bit 44 being stuck on zero? or can each dow get varying values? -- Les chiens aboient, et la caravane passe. ___ 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 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. -- Tom Harris celephi...@gmail.com ___ 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] Switching oscillators
Hi On Dec 9, 2012, at 9:07 PM, WB6BNQ wb6...@cox.net wrote: Bob, I don't know where you got your multiple versions. Quite possibly from having spent the last 40 years in the business of designing and building OCXO's. RETRACE deals with the object in a fully realized steady state operating condition, then removal of power for a period of time (usually stated in real specs as 24 hours) and then re application of power with a stated result. Really quite simple in concept. Its cousin, WARMUP, would describe a similar process assuming both are at a non operating nominal environment ambient temperature. Usually a RETRACE spec will state some amount of off period, like 24 hours. Unless we are talking super hot molten metal or a nuclear reaction, usually most things get to nominal room temperature inside of 24 hours. So, in that case RETRACE and WARMUP would mean virtually the same Except that you see them as independent specifications on OCXO's talking about different parts of the process. Bob thing. BillWB6BNQ Bob Camp wrote: Hi That's one interpretation and one method of measurement. The other method is to measure frequency shift from say one hour after power on to 24 hours after power on. A lot depends on the requirements of the system the OCXO was intended to be used in. If you use the measure / power off / warmup / measure approach, you need to: 1) Define the time on before the first measure 2) Define the power off time 3) Define the warmup time Change any of those numbers and the retrace number will be different. Generally the temperature(s) involved are also defined. The most common case is that all of this is done at 20 or 25C. Bob On Dec 9, 2012, at 8:47 PM, Tom Miller tmil...@skylinenet.net wrote: So if I understand it, you allow the OCXO several days of warm-up to set the frequency. Then when turned off for a while, then restarted. After some warm-up period, the retrace spec would give an indication of how close the frequency will be? Thanks, Tom - Original Message - From: Bob Camp li...@rtty.us To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Sunday, December 09, 2012 7:58 PM Subject: Re: [time-nuts] Switching oscillators Hi Retrace assumes that the oscillator has some normal frequency that only moves according to the aging rate. Retrace occurs after the oscillator has been off power for some period. The rate of change is greater than the aging rate. Warmup and retrace are obviously inter-related. Warmup is generally described as a short term (sub 1 hour) process. Retrace is often looked at as a day to multiple day sort of thing. Since none of this is tightly defined, you will see various specs looking at the same issues a bit differently. Often those differences roll up to some sort of system requirement. Bob On Dec 9, 2012, at 7:45 PM, Tom Miller tmil...@skylinenet.net wrote: Hi Bob, Can you give a good definition of retrace as it applies here? Thanks, Tom - Original Message - From: Bob Camp li...@rtty.us To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Sunday, December 09, 2012 7:24 PM Subject: Re: [time-nuts] Switching oscillators Hi Your TCXO will have the same sort of retrace issues as your OCXO. Past some number of minutes (5,10,15…) you will always be better with a modern OCXO than with a TCXO. Bob On Dec 9, 2012, at 7:05 PM, WB6BNQ wb6...@cox.net wrote: Hi Joe, I think you all are not looking at this correctly. 1.First, as has been pointed out, a TCXO will vary around till the environment it is in has returned to its nominal operating temperature. 2.A typical TCXO is nominally spec'ed around +/- 0.5 x 10-7 neighborhood. Not a stellar number. 3.The real spec to look at is the RETRACE factor of a good OCXO. Many of the modern PCB CAN manufactures do not or are quite hazy on this point. Vectron, for example, on their double oven high performance WIDGET (model DX-170) claims a warmup time of 5 minutes to +-10ppb of final frequency, however, they also include this cryptic statement (1 hour reading) @ +25DEGC on the same spec. I am not sure, but it suggests that they are reading the final frequency at the one hour point after turn-on. Taking it at face value, it suggest that the oscillator is within +/- 1 x 10-8 at 5 minutes. That is a whole decade better than the TCXO under any condition. Looking at something real like the HP 10811A/B Quartz Crystal Oscillator, you will see they spec the retrace as Warmup 10 min. after turn-on within 5 X 10-9 of final value, at 25DEGC and 20 Vdc. See Notes 1 2. Notes: 1. For oscillator off-time less than 24 hours. 2. Final value is defined as frequency 24 hours after turn-on. Here, we are talking about two whole decades better than
Re: [time-nuts] Cinox source
Hi Perkin Elmer does everything that EGG used to do - same company, new name. Bob On Dec 9, 2012, at 11:07 PM, Jim Lux jim...@earthlink.net wrote: On 12/9/12 5:38 PM, Bob Camp wrote: HI Princeton was part of EGG for quite a while. EGG bought out Perkin Elmer in the 90's. The EGG name was not as warm and fuzzy as Perkin Elmer. The switched over the name of the combined company to Perkin Elmer a bit after they got together. P-E is also who does all of EGG's triggered spark gaps and such (I guess that's peripherally time nutty.. generating precision HV pulses with sub-ns accuracy) .. I think they also absorbed Physics International. ___ 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 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] FE-5680A OSMT connector / RS232
Hi You may have a TTL output level version of the FE. Bob On Dec 10, 2012, at 1:24 AM, James Peroulas ja...@peroulas.com wrote: For the record, my device did have an OSMT (not Hirose) connector on the DDS board. I'm having trouble getting the internal RS232 port on the DDS board to work and I was wondering if there were any tricks to it? I'm not able to get it to respond to the 'S' command and when I measure the voltage on the RS232 TX pin (#2 from the left) it's always 0v. Shouldn't it be -12v when idle? Thanks, James On Mon, Nov 26, 2012 at 11:22 PM, James Peroulas ja...@peroulas.com wrote: Was wondering if anyone ever confirmed the type of RF connector found internally in the FE-5860a? I've seen several posts calling it a Hirose U.FL, but the linked thread suggests that it might actually be an OSMT connector: http://www.vklogger.com/forum/viewtopic.php?t=10289 Or is everyone simply soldering cable directly to the center pin? Thanks, James -- *Integrity is a binary state - either you have it or you don’t.* - John Doerr -- *Integrity is a binary state - either you have it or you don’t.* - John Doerr ___ 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 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] Sir Patrick Moore
On 10 December 2012 09:24, David C. Partridge david.partri...@perdrix.co.uk wrote: Related to time-nuttery as astonomical observation was used for time-keeping until C20. Sir Patrick Moore, the great amateur astonomer died yesterday at the age of 89. Although considered an amateur, I assume he got paid for presenting The Sky At Night. According to Wikipediat at least, he was an author of over 70 books on astronomy. So perhaps amateur is not quite the right word. A gentleman, astronomer, and true scholar. His death was predicatable as he's been very ill for a long time, but nonetheless a great loss. I think the death of us is all predictable! But I sorry to hear of his passing away. I suspect there are a good few people who have astronomy PhDs, which would not have studied astronomy if it was not for Patrick Moore. He certainly insprired me to buy a telescope (4.5 reflector), but I never studied the subject professionally. Regards, David Partridge Dave ___ 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] Nifty MINI TIC for DMTD work please hold off, 4 channel pic...
Bert, I have been waiting for more details to become available. Thanks for the update. Not sure exactly what this means: but I have to depend on volunteers to do the drawings and they have to make their time available when convenient. Depending on if the input is in some kind of standard, or a transmittable sketch format, and what the desired output is, maybe I could help. Contact me if you think another cook could possibly help the broth. -Rex in San Jose, CA On 12/9/2012 3:20 AM, ewkeh...@aol.com wrote: We have a choice of two dual mixers, my copy of the original NBS with minor changes and Bill Riley's unit. Comparison tests are ongoing to pick the best. If Bill's shows better results I plan on laying one out using leaded components. 2 channel and 4 channel counters are completed and are being used for testing. Corby has done some more tests plotting single channel using the period mode. Great results. Can use phase or period. A documentation package is being prepared to be placed on Didier's site, but I have to depend on volunteers to do the drawings and they have to make their time available when convenient. Juerg is continuing his work on programming a PIC for LCD display but that will only be an added feature one can use the counters direct with a PC and as you may have noticed Corby sold his SR 620. Bert Kehren In a message dated 12/8/2012 5:50:35 P.M. Eastern Standard Time, w1...@earthlink.net writes: What is the status of this project ? I may have missed a few e-mails. Thanks, Dick, W1KSZ ___ 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 DCF77 master clock data
If it were me, I'd suspect that my window of bits is off to the left or right by one bit. The old first bit is called zero problem. And the relationship between 7- 3 and 5- 1 makes me wonder. Chris On 12/10/2012 5:00 AM, Tom Harris wrote: I am logging the data from this strange master clock, and the none of the bits 42 through 44 are stuck. Good idea, though. I have just checked tonight, and the DOW is indicating as 5 (Friday), despite it only being Monday. Good idea though. On 9 December 2012 23:24, mike cook mc235...@gmail.com wrote: Le 09/12/2012 11:28, Tom Harris a écrit : Greetings Time Nuts. Time related, but unusually so: I am examining the DCF77 (see http://en.wikipedia.org/wiki/**DCF77http://en.wikipedia.org/wiki/DCF77for a refresher) data output from a master clock intended to drive a series of slave clocks operated from 24V DC, interrupted by 100ms or 200ms pulses to signal '0' '1' on the data. What I am seeing is that the day of week bits 42 through 44 inclusive do not code the day correctly, they seems to have a variable offset from the actual day, I have seen an offset from +1 to +3. The rest of the data, ncluding parity bits, are OK. Today, 9/12/2012, the day is Sunday, which should be coded as '7'. However I am seeing a value of '3', which is Wednesday. Is this normal, do wired DCF77 slave clocks transmit munged dow data to prevent them being used with someone else's master clock? If they do I am hosed since I am trying to reverse engineer the master clock. Is it just bit 44 being stuck on zero? or can each dow get varying values? -- Les chiens aboient, et la caravane passe. __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ 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] EIP545A 18GHz counter query
I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex On 12/9/2012 8:41 AM, paul swed wrote: Chris TP4 is not all that helpful its a control signal. So let me back up for a minute. What are we troubleshooting? I was thinking band 2 was semi working and you still had a totally dead band 3. My comments have been around trying to see what was going on with 3. Other comment. I think we should take this offline. Most likely driving time-nuts nuts. :-) Do you use skype? Regards Paul ___ 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] Switching oscillators
Not sure why you would want to do this but, if you have a Rb, such as an LPRO-101 or others, there is a signal that lets you know the Rb is 'locked'. You could use that to drive a relay that would switch the output of the Rb to the external input of your device. This presumes you do not have to mechanically flip a switch on the device to select internal or external references. On an OCXO, I guess you could sense the oven current and, when it goes low, indicating that the oven is 'warm', that could be used to drive the relay to send the output of the OCXO to the device. Otherwise, you would need a 'house standard', that is always on, and measure how far apart the OCXO is from your 'house standard' and switch when your OCXO meets certain performance specs. In that case, why not use the 'house standard' all the time? Might be easier to build a battery back-up for your 'house standard'. Hope this helps. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Joseph Gray Sent: Sunday, December 09, 2012 3:13 PM To: Discussion of precise time and frequency measurement Subject: [time-nuts] Switching oscillators I have a device that has an internal TCXO. I want to feed it with an external OCXO, but I don't want to completely replace the TCXO. Here is the scenario. On initial power on, or after a power loss, I want the internal TCXO to be used. Once the OCXO is up, I want to switch to it. How could this be done easily and cheaply? Joe Gray W5JG ___ 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 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] Sir Patrick Moore
On 12/10/2012 7:10 AM, David Kirkby wrote: On 10 December 2012 09:24, David C. Partridge david.partri...@perdrix.co.uk wrote: Related to time-nuttery as astonomical observation was used for time-keeping until C20. Sir Patrick Moore, the great amateur astonomer died yesterday at the age of 89. Although considered an amateur, I assume he got paid for presenting The Sky At Night. According to Wikipediat at least, he was an author of over 70 books on astronomy. So perhaps amateur is not quite the right word. If I'm remembering correctly, he was a guest several times on the BBC radio show Just A Minute and was absolutely wonderful -- ideal for that show because he spoke about 600 words per minute and made perfect sense while doing so. John ___ 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] EIP545A 18GHz counter query
I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex 10/12/2012 13:40 Will do Rex, but I was worried this thread may not have been strictly on topic for this reflector. Comments? -- Best Regards, Chris Wilson. ___ 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] Of possible interest
On 12/9/12 10:14 PM, DaveH wrote: http://www.scmagazine.com.au/News/325731,researchers-find-crippling-flaws-in -global-gps.aspx ___ 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. It is interesting.. it's more about finding software vulnerabilities in the receiver than attacking the GPS signal (although exercising the vulnerability requires spoofing in some cases). It doesn't come as a huge surprise that there are receivers with software bugs in the nav computation. ___ 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] Of possible interest
On 12/9/12 10:14 PM, DaveH wrote: http://www.scmagazine.com.au/News/325731,researchers-find-crippling-flaws-in -global-gps.aspx ___ 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. the actual paper is at: http://users.ece.cmu.edu/~dbrumley/courses/18487-f12/readings/Nov28_GPS.pdf ___ 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] EIP545A 18GHz counter query
Me too. I have two 545A's, one of which works and the other of which has issues with Band 3, failing to read in a very interesting pattern suggesting a YIG filter issue, probably the control circuit. Just no time to chase it right now. Please post the resolution. Thanks. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rex Sent: Monday, December 10, 2012 6:50 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] EIP545A 18GHz counter query I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex On 12/9/2012 8:41 AM, paul swed wrote: Chris TP4 is not all that helpful its a control signal. So let me back up for a minute. What are we troubleshooting? I was thinking band 2 was semi working and you still had a totally dead band 3. My comments have been around trying to see what was going on with 3. Other comment. I think we should take this offline. Most likely driving time-nuts nuts. :-) Do you use skype? Regards Paul ___ 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 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] EIP545A 18GHz counter query
Chris and I took this offline so as not to bother the time-nuts thread. Hopefully we can work it through. Chris is lucky as he only has a band 2 problem. Band 3 in nastier o deal with. Chris since several have expressed interest perhaps we just add them to your new thread. We can still skype for the real time work. Joe to your 545. Lots of possibilities. There is the yig tuning circuit driven by a dac driven by a 6.2V zener reference. I have run into bad zeners more then I care to count. Plus the other part of the equation is the locked comb source. So as I say band 3 is nasty lots can go wrong besides someone simply frying the front end. Regards Paul On Mon, Dec 10, 2012 at 9:00 AM, J. L. Trantham jlt...@att.net wrote: Me too. I have two 545A's, one of which works and the other of which has issues with Band 3, failing to read in a very interesting pattern suggesting a YIG filter issue, probably the control circuit. Just no time to chase it right now. Please post the resolution. Thanks. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rex Sent: Monday, December 10, 2012 6:50 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] EIP545A 18GHz counter query I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex On 12/9/2012 8:41 AM, paul swed wrote: Chris TP4 is not all that helpful its a control signal. So let me back up for a minute. What are we troubleshooting? I was thinking band 2 was semi working and you still had a totally dead band 3. My comments have been around trying to see what was going on with 3. Other comment. I think we should take this offline. Most likely driving time-nuts nuts. :-) Do you use skype? Regards Paul ___ 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 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 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] 31bit ADC, 1000 samples per second
31-bit is 186dB... with what do you compare that device? What kind of reference is needed? On Sun, Dec 9, 2012 at 11:18 PM, Bob Camp li...@rtty.us wrote: Hi It's the big brother of the ADS127x parts that I mentioned a little while back. I'm sure it works quite well. The 127x's are fine parts. The internal filtering combined with the chopper is a very nice combination. Bob On Dec 9, 2012, at 5:13 PM, Poul-Henning Kamp p...@phk.freebsd.dk wrote: I'm doing some work with a couple of geophones (long story...) and stumbled across TI's ADS1282 geophone ADC. It has some pretty amazing specs for low frequencies, and having played with the EVAL board for an evening here in my lab, I'm pretty impressed. People working with close-in phase-noise measurements may want to consider this chip... -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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 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 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] EIP545A 18GHz counter query
Or join and post to the EIP_Microwave Yahoo group - lots of good info showing up there! Steve WB0DBS On Dec 10, 2012, at 8:00 AM, J. L. Trantham jlt...@att.net wrote: Me too. I have two 545A's, one of which works and the other of which has issues with Band 3, failing to read in a very interesting pattern suggesting a YIG filter issue, probably the control circuit. Just no time to chase it right now. Please post the resolution. Thanks. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rex Sent: Monday, December 10, 2012 6:50 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] EIP545A 18GHz counter query I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex On 12/9/2012 8:41 AM, paul swed wrote: Chris TP4 is not all that helpful its a control signal. So let me back up for a minute. What are we troubleshooting? I was thinking band 2 was semi working and you still had a totally dead band 3. My comments have been around trying to see what was going on with 3. Other comment. I think we should take this offline. Most likely driving time-nuts nuts. :-) Do you use skype? Regards Paul ___ 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 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 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] 31bit ADC, 1000 samples per second
In message CAL8XPmN8PbcSEfmRCiude+sYCX3kawBF6h4T0=n0n0emxlx...@mail.gmail.com, Azelio Boriani writes: 31-bit is 186dB... with what do you compare that device? What kind of reference is needed? It's not 31 noise-free bits, but the SNR is in the 120-130dB territory. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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] EIP545A 18GHz counter query
I'm already over there. The thread just migrated back here. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Steve Sent: Monday, December 10, 2012 8:43 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] EIP545A 18GHz counter query Or join and post to the EIP_Microwave Yahoo group - lots of good info showing up there! Steve WB0DBS On Dec 10, 2012, at 8:00 AM, J. L. Trantham jlt...@att.net wrote: Me too. I have two 545A's, one of which works and the other of which has issues with Band 3, failing to read in a very interesting pattern suggesting a YIG filter issue, probably the control circuit. Just no time to chase it right now. Please post the resolution. Thanks. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Rex Sent: Monday, December 10, 2012 6:50 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] EIP545A 18GHz counter query I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex On 12/9/2012 8:41 AM, paul swed wrote: Chris TP4 is not all that helpful its a control signal. So let me back up for a minute. What are we troubleshooting? I was thinking band 2 was semi working and you still had a totally dead band 3. My comments have been around trying to see what was going on with 3. Other comment. I think we should take this offline. Most likely driving time-nuts nuts. :-) Do you use skype? Regards Paul ___ 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 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 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 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] 31bit ADC, 1000 samples per second
31 bits is just a convenient word size that comes out of the on-chip filter, and does not really relate to performance. Better to look at SNR which, at worst case, is 120dB in high-res mode. That indicates performance just under 20 bits, typical for better monolithic sigma-delta converters. Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. Bob LaJeunesse From: Azelio Boriani azelio.bori...@screen.it To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Mon, December 10, 2012 9:39:30 AM Subject: Re: [time-nuts] 31bit ADC, 1000 samples per second 31-bit is 186dB... with what do you compare that device? What kind of reference is needed? On Sun, Dec 9, 2012 at 11:18 PM, Bob Camp li...@rtty.us wrote: Hi It's the big brother of the ADS127x parts that I mentioned a little while back. I'm sure it works quite well. The 127x's are fine parts. The internal filtering combined with the chopper is a very nice combination. Bob On Dec 9, 2012, at 5:13 PM, Poul-Henning Kamp p...@phk.freebsd.dk wrote: I'm doing some work with a couple of geophones (long story...) and stumbled across TI's ADS1282 geophone ADC. It has some pretty amazing specs for low frequencies, and having played with the EVAL board for an evening here in my lab, I'm pretty impressed. People working with close-in phase-noise measurements may want to consider this chip... -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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 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 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 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 DCF77 master clock data
Le 10/12/2012 13:44, Chris Howard a écrit : If it were me, I'd suspect that my window of bits is off to the left or right by one bit. The old first bit is called zero problem. And the relationship between 7- 3 and 5- 1 makes me wonder. Chris On 12/10/2012 5:00 AM, Tom Harris wrote: I am logging the data from this strange master clock, and the none of the bits 42 through 44 are stuck. Good idea, though. I have just checked tonight, and the DOW is indicating as 5 (Friday), despite it only being Monday. Good idea though. I guess you will need to log a weeks worth to see if there is a logical pattern. How about flipped bit 44 ? That would give the seen Sunday and Monday values. Tomorrow would be 6. Mon = 101 5 Tue = 110 6 Wed = 111 7 Thu = 000 0 Fri = 001 1 Sat = 010 2 Sun = 011 3 The parity being good may mean that it is not an error, but a feature. On 9 December 2012 23:24, mike cook mc235...@gmail.com wrote: Le 09/12/2012 11:28, Tom Harris a écrit : Greetings Time Nuts. Time related, but unusually so: I am examining the DCF77 (see http://en.wikipedia.org/wiki/**DCF77http://en.wikipedia.org/wiki/DCF77for a refresher) data output from a master clock intended to drive a series of slave clocks operated from 24V DC, interrupted by 100ms or 200ms pulses to signal '0' '1' on the data. What I am seeing is that the day of week bits 42 through 44 inclusive do not code the day correctly, they seems to have a variable offset from the actual day, I have seen an offset from +1 to +3. The rest of the data, ncluding parity bits, are OK. Today, 9/12/2012, the day is Sunday, which should be coded as '7'. However I am seeing a value of '3', which is Wednesday. Is this normal, do wired DCF77 slave clocks transmit munged dow data to prevent them being used with someone else's master clock? If they do I am hosed since I am trying to reverse engineer the master clock. Is it just bit 44 being stuck on zero? or can each dow get varying values? -- Les chiens aboient, et la caravane passe. __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ 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. -- Les chiens aboient, et la caravane passe. ___ 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] 31bit ADC, 1000 samples per second
I see that there is a high resolution 130dB mode too that can give 21 bits @ 250 samples per second, good for very slow, high resolution DMTD (but also very stable voltage reference and ADC temperature, seems very challenging). On Mon, Dec 10, 2012 at 4:26 PM, Robert LaJeunesse rlajeune...@sbcglobal.net wrote: 31 bits is just a convenient word size that comes out of the on-chip filter, and does not really relate to performance. Better to look at SNR which, at worst case, is 120dB in high-res mode. That indicates performance just under 20 bits, typical for better monolithic sigma-delta converters. Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. Bob LaJeunesse From: Azelio Boriani azelio.bori...@screen.it To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Mon, December 10, 2012 9:39:30 AM Subject: Re: [time-nuts] 31bit ADC, 1000 samples per second 31-bit is 186dB... with what do you compare that device? What kind of reference is needed? On Sun, Dec 9, 2012 at 11:18 PM, Bob Camp li...@rtty.us wrote: Hi It's the big brother of the ADS127x parts that I mentioned a little while back. I'm sure it works quite well. The 127x's are fine parts. The internal filtering combined with the chopper is a very nice combination. Bob On Dec 9, 2012, at 5:13 PM, Poul-Henning Kamp p...@phk.freebsd.dk wrote: I'm doing some work with a couple of geophones (long story...) and stumbled across TI's ADS1282 geophone ADC. It has some pretty amazing specs for low frequencies, and having played with the EVAL board for an evening here in my lab, I'm pretty impressed. People working with close-in phase-noise measurements may want to consider this chip... -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org| TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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 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 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 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 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] EIP545A 18GHz counter query
Yes. There is a Yahoo Group spwecifically about EIP products. -John I have an EIP counter with a problem, stashed, that I need to look at eventually. So have been following the thread a bit. If you guys get somewhere conclusive (or not), might be nice to post a summary of all important details, techniques, or circuit notes beyond the manual, found, to this thread for closing to posterity searchers. (And me.) And, in the mean time, good luck, bon debugging. -Rex 10/12/2012 13:40 Will do Rex, but I was worried this thread may not have been strictly on topic for this reflector. Comments? -- Best Regards, Chris Wilson. ___ 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 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] 31bit ADC, 1000 samples per second
Hi What is atypical about these parts is: 1) The excellent noise performance inside 10 Hz 2) The built in Nyquist filters The first is very impressive compared to just about anything else out there. The second is available in some, but not all similar parts. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Robert LaJeunesse Sent: Monday, December 10, 2012 10:27 AM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] 31bit ADC, 1000 samples per second 31 bits is just a convenient word size that comes out of the on-chip filter, and does not really relate to performance. Better to look at SNR which, at worst case, is 120dB in high-res mode. That indicates performance just under 20 bits, typical for better monolithic sigma-delta converters. Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. Bob LaJeunesse From: Azelio Boriani azelio.bori...@screen.it To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Mon, December 10, 2012 9:39:30 AM Subject: Re: [time-nuts] 31bit ADC, 1000 samples per second 31-bit is 186dB... with what do you compare that device? What kind of reference is needed? On Sun, Dec 9, 2012 at 11:18 PM, Bob Camp li...@rtty.us wrote: Hi It's the big brother of the ADS127x parts that I mentioned a little while back. I'm sure it works quite well. The 127x's are fine parts. The internal filtering combined with the chopper is a very nice combination. Bob On Dec 9, 2012, at 5:13 PM, Poul-Henning Kamp p...@phk.freebsd.dk wrote: I'm doing some work with a couple of geophones (long story...) and stumbled across TI's ADS1282 geophone ADC. It has some pretty amazing specs for low frequencies, and having played with the EVAL board for an evening here in my lab, I'm pretty impressed. People working with close-in phase-noise measurements may want to consider this chip... -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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 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 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 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 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] Switching oscillators
Hi The oven monitor pin on most OCXO's is not a real good thing to switch directly off of. They generally go low well before the output is very stable. A timer is run off of the oven monitor pin would give you a pretty good switch point (switch X minutes after the oven monitor shows good). Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of J. L. Trantham Sent: Monday, December 10, 2012 8:18 AM To: 'Discussion of precise time and frequency measurement' Subject: Re: [time-nuts] Switching oscillators Not sure why you would want to do this but, if you have a Rb, such as an LPRO-101 or others, there is a signal that lets you know the Rb is 'locked'. You could use that to drive a relay that would switch the output of the Rb to the external input of your device. This presumes you do not have to mechanically flip a switch on the device to select internal or external references. On an OCXO, I guess you could sense the oven current and, when it goes low, indicating that the oven is 'warm', that could be used to drive the relay to send the output of the OCXO to the device. Otherwise, you would need a 'house standard', that is always on, and measure how far apart the OCXO is from your 'house standard' and switch when your OCXO meets certain performance specs. In that case, why not use the 'house standard' all the time? Might be easier to build a battery back-up for your 'house standard'. Hope this helps. Joe -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Joseph Gray Sent: Sunday, December 09, 2012 3:13 PM To: Discussion of precise time and frequency measurement Subject: [time-nuts] Switching oscillators I have a device that has an internal TCXO. I want to feed it with an external OCXO, but I don't want to completely replace the TCXO. Here is the scenario. On initial power on, or after a power loss, I want the internal TCXO to be used. Once the OCXO is up, I want to switch to it. How could this be done easily and cheaply? Joe Gray W5JG ___ 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 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 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] Counter
Is shipped, tracking number 1Z30VR960337713831 due Wednesday Didier Sent from my Droid Razr 4G LTE wireless tracker. ___ 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
Thanks, Didier, if you ever find a mainframe for it, please keep me in mind. Thanks, -Don -- From: shali...@gmail.com Sent: Monday, December 10, 2012 11:23 AM To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: [time-nuts] Counter Is shipped, tracking number 1Z30VR960337713831 due Wednesday Didier Sent from my Droid Razr 4G LTE wireless tracker. ___ 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 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
Well, that obviously went to the wrong place. That was a private message, sorry for the bandwidth. Didier On Mon, Dec 10, 2012 at 12:06 PM, dlewis6767 dlewis6...@austin.rr.comwrote: Thanks, Didier, if you ever find a mainframe for it, please keep me in mind. Thanks, -Don --** From: shali...@gmail.com Sent: Monday, December 10, 2012 11:23 AM To: Discussion of precise time and frequency measurement time-nuts@febo.com Subject: [time-nuts] Counter Is shipped, tracking number 1Z30VR960337713831 due Wednesday Didier Sent from my Droid Razr 4G LTE wireless tracker. __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ 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] 31bit ADC, 1000 samples per second
rlajeune...@sbcglobal.net said: Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. The initial post called it TI's ADS1282 geophone ADC. I'm pretty sure that the geologists using geophones don't care about super low frequencies so they probably won't care if the reference changes due to temperature shifts. Some geologists do care about super low frequencies. That would be continental drift. -- These are my opinions. I hate spam. ___ 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] 31bit ADC, 1000 samples per second
Hi When your sound source is a thumper truck, you do indeed care about some pretty low frequencies. Bob -Original Message- From: time-nuts-boun...@febo.com [mailto:time-nuts-boun...@febo.com] On Behalf Of Hal Murray Sent: Monday, December 10, 2012 1:33 PM To: Discussion of precise time and frequency measurement Subject: Re: [time-nuts] 31bit ADC, 1000 samples per second rlajeune...@sbcglobal.net said: Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. The initial post called it TI's ADS1282 geophone ADC. I'm pretty sure that the geologists using geophones don't care about super low frequencies so they probably won't care if the reference changes due to temperature shifts. Some geologists do care about super low frequencies. That would be continental drift. -- These are my opinions. I hate spam. ___ 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 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] 31bit ADC, 1000 samples per second
It is 31 bits with no missing codes. Usually missing codes is of concern in feedback systems, but I don't see the use in a geophone. Perhaps they will average the digital signal further to reduce the noise, hence the noisy bits. ___ 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] 31bit ADC, 1000 samples per second
I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) On Mon, Dec 10, 2012 at 2:19 PM, li...@lazygranch.com wrote: It is 31 bits with no missing codes. Usually missing codes is of concern in feedback systems, but I don't see the use in a geophone. Perhaps they will average the digital signal further to reduce the noise, hence the noisy bits. ___ 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 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] 31bit ADC, 1000 samples per second
Am 10.12.2012 21:53, schrieb Didier Juges: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) With that much noise it is really guaranteed that no code will be missing. :-) Gerhard ___ 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] 31bit ADC, 1000 samples per second
How long does it take to prove it? And what's the point? On Mon, Dec 10, 2012 at 3:04 PM, Gerhard Hoffmann dk...@arcor.de wrote: Am 10.12.2012 21:53, schrieb Didier Juges: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) With that much noise it is really guaranteed that no code will be missing. :-) Gerhard __**_ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/** mailman/listinfo/time-nutshttps://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there. ___ 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] 31bit ADC, 1000 samples per second
Regarding feedback loops, I had a brain fart. I mean monotonic and no missing codes, not no missing codes in general. I think it was Fluke or Analog that had a small booklet on understanding data converter specifications. ___ 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] 31bit ADC, 1000 samples per second
Hi I suspect you can prove it mathematically. You could also just sit there and watch what it puts out for a year or so. With a reasonable ramp it likely would put out all codes. That's not to say you could prove they are in order, only that you saw all 4 billion codes. More or less: 1,000 samples a second, 4 billion codes - you need 4 million seconds if everything works perfectly. 10X that number is probably adequate to catch them all. Bob On Dec 10, 2012, at 3:53 PM, Didier Juges shali...@gmail.com wrote: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) On Mon, Dec 10, 2012 at 2:19 PM, li...@lazygranch.com wrote: It is 31 bits with no missing codes. Usually missing codes is of concern in feedback systems, but I don't see the use in a geophone. Perhaps they will average the digital signal further to reduce the noise, hence the noisy bits. ___ 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 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 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] 31bit ADC, 1000 samples per second
Integrating converters including delta-sigma converters can be no missing codes by design without being able to take advantage of the full resolution that implies. Integral nonlinearity, drift, and noise will limit the usable resolution. On Mon, 10 Dec 2012 15:08:38 -0600, Didier Juges shali...@gmail.com wrote: How long does it take to prove it? And what's the point? On Mon, Dec 10, 2012 at 3:04 PM, Gerhard Hoffmann dk...@arcor.de wrote: Am 10.12.2012 21:53, schrieb Didier Juges: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) With that much noise it is really guaranteed that no code will be missing. :-) Gerhard ___ 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] 31bit ADC, 1000 samples per second
The gain drift is specified at 2ppm per degree C. There are provisions for a calibration cycle but that requires external multiplexing. On Mon, 10 Dec 2012 07:26:55 -0800 (PST), Robert LaJeunesse rlajeune...@sbcglobal.net wrote: . . . typical for better monolithic sigma-delta converters. Nonetheless 20 bits indicates a matching reference would have to hold 1PPM over temperature, no simple task. Bob LaJeunesse ___ 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] Thunderbolt oven / non-stable operating temperature
Ended up with a gently used trimble thunderbolt a few months ago, and been trying to figure out the best settings to use (time constant, damping, etc.) for best performance. I got distracted though. There is something which I find rather annoying, and I'm spending more time messing with this factor than actually observing the performance at different damping / time constant / etc... The temperature seems to be aprox 20C above ambient temperature rather than a steady 40C. I thought the purpose of the oven was to keep the temperature stable. When I adjust the thermostat because I'm chilly, the thunderbolt gets warmer too (but not by as much) I've tried (seemingly, to me) everything: light insulation putting the unit in a spot away from drafts putting the unit in a place away from drafts in a shoebox with light insulation with a fan blowing in it (currently in a cupboard in this state, and the temp seems to fluctuate the least, but still by more than 1-2 degrees centigrade) I feel like shouldn't need to fuss with ambient conditions this heavily for an OCXO, and find myself researching construction / design for a DIY outer-oven to wrap the thunderbolt in. Anyone have experience with non-stable temperature on a trimble thunderbolt? I've also noticed (monitoring with Lady Heather's Disciplined Oscillator Control Program v3.10 by John Miles aka KE5FX) the thunderbolt unit sometimes flags an alarm briefly (and I rarely ever manage to write it down before the screen updates or the alarm goes away) and the last one I was able to catch before it went away was related to temperature. Should I be getting a steady / non-flucturating 40C or something if the oven is working normally? Thanks in advance, Sarah ___ 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] Thunderbolt oven / non-stable operating temperature
Hi The oven in the OCXO keeps the crystal at a constant temperature. It is normal for the TBolt it's self to idle 10 to 20C above ambient. The temperature sensor is located near the 9 pin D connector. It shows a temperature somewhere in between the ambient and the OCXO case temperature. Some have mounted speed controlled fans to keep the TBolt case at a constant temperature. Bob On Dec 10, 2012, at 5:15 PM, Sarah White kuze...@gmail.com wrote: Ended up with a gently used trimble thunderbolt a few months ago, and been trying to figure out the best settings to use (time constant, damping, etc.) for best performance. I got distracted though. There is something which I find rather annoying, and I'm spending more time messing with this factor than actually observing the performance at different damping / time constant / etc... The temperature seems to be aprox 20C above ambient temperature rather than a steady 40C. I thought the purpose of the oven was to keep the temperature stable. When I adjust the thermostat because I'm chilly, the thunderbolt gets warmer too (but not by as much) I've tried (seemingly, to me) everything: light insulation putting the unit in a spot away from drafts putting the unit in a place away from drafts in a shoebox with light insulation with a fan blowing in it (currently in a cupboard in this state, and the temp seems to fluctuate the least, but still by more than 1-2 degrees centigrade) I feel like shouldn't need to fuss with ambient conditions this heavily for an OCXO, and find myself researching construction / design for a DIY outer-oven to wrap the thunderbolt in. Anyone have experience with non-stable temperature on a trimble thunderbolt? I've also noticed (monitoring with Lady Heather's Disciplined Oscillator Control Program v3.10 by John Miles aka KE5FX) the thunderbolt unit sometimes flags an alarm briefly (and I rarely ever manage to write it down before the screen updates or the alarm goes away) and the last one I was able to catch before it went away was related to temperature. Should I be getting a steady / non-flucturating 40C or something if the oven is working normally? Thanks in advance, Sarah ___ 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 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] Thunderbolt oven / non-stable operating temperature
Hi Sarah, The temperature you are seeing is that of a Dallas IC mounted on the main PC board. You are seeing the ambient temp of the board, not inside the OCXO oven. Regards, Tom - Original Message - From: Sarah White kuze...@gmail.com To: Discussion of precise time and frequency measurement time-nuts@febo.com Sent: Monday, December 10, 2012 5:15 PM Subject: [time-nuts] Thunderbolt oven / non-stable operating temperature Ended up with a gently used trimble thunderbolt a few months ago, and been trying to figure out the best settings to use (time constant, damping, etc.) for best performance. I got distracted though. There is something which I find rather annoying, and I'm spending more time messing with this factor than actually observing the performance at different damping / time constant / etc... The temperature seems to be aprox 20C above ambient temperature rather than a steady 40C. I thought the purpose of the oven was to keep the temperature stable. When I adjust the thermostat because I'm chilly, the thunderbolt gets warmer too (but not by as much) I've tried (seemingly, to me) everything: light insulation putting the unit in a spot away from drafts putting the unit in a place away from drafts in a shoebox with light insulation with a fan blowing in it (currently in a cupboard in this state, and the temp seems to fluctuate the least, but still by more than 1-2 degrees centigrade) I feel like shouldn't need to fuss with ambient conditions this heavily for an OCXO, and find myself researching construction / design for a DIY outer-oven to wrap the thunderbolt in. Anyone have experience with non-stable temperature on a trimble thunderbolt? I've also noticed (monitoring with Lady Heather's Disciplined Oscillator Control Program v3.10 by John Miles aka KE5FX) the thunderbolt unit sometimes flags an alarm briefly (and I rarely ever manage to write it down before the screen updates or the alarm goes away) and the last one I was able to catch before it went away was related to temperature. Should I be getting a steady / non-flucturating 40C or something if the oven is working normally? Thanks in advance, Sarah ___ 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 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] Thunderbolt oven / non-stable operating temperature
Sarah, The reported temperature is not the oven temp, but rather the temp of the circuit board just behind the DB-9 serial connector. As far as I know, the actual oven temp is not available outside the OCXO. Regards, geo On Mon, Dec 10, 2012 at 5:15 PM, Sarah White kuze...@gmail.com wrote: Ended up with a gently used trimble thunderbolt a few months ago, and been trying to figure out the best settings to use (time constant, damping, etc.) for best performance. I got distracted though. There is something which I find rather annoying, and I'm spending more time messing with this factor than actually observing the performance at different damping / time constant / etc... The temperature seems to be aprox 20C above ambient temperature rather than a steady 40C. I thought the purpose of the oven was to keep the temperature stable. When I adjust the thermostat because I'm chilly, the thunderbolt gets warmer too (but not by as much) I've tried (seemingly, to me) everything: light insulation putting the unit in a spot away from drafts putting the unit in a place away from drafts in a shoebox with light insulation with a fan blowing in it (currently in a cupboard in this state, and the temp seems to fluctuate the least, but still by more than 1-2 degrees centigrade) I feel like shouldn't need to fuss with ambient conditions this heavily for an OCXO, and find myself researching construction / design for a DIY outer-oven to wrap the thunderbolt in. Anyone have experience with non-stable temperature on a trimble thunderbolt? I've also noticed (monitoring with Lady Heather's Disciplined Oscillator Control Program v3.10 by John Miles aka KE5FX) the thunderbolt unit sometimes flags an alarm briefly (and I rarely ever manage to write it down before the screen updates or the alarm goes away) and the last one I was able to catch before it went away was related to temperature. Should I be getting a steady / non-flucturating 40C or something if the oven is working normally? Thanks in advance, Sarah ___ 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 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] 31bit ADC, 1000 samples per second
I agree, but there is a difference between something that you should be able to achieve by design and being able to prove it, or even if it's useful in practice. In a case like this, the wide span between resolution and effective bits makes me wonder what's the point of advertising the former other than bragging rights? On Mon, Dec 10, 2012 at 3:38 PM, David davidwh...@gmail.com wrote: Integrating converters including delta-sigma converters can be no missing codes by design without being able to take advantage of the full resolution that implies. Integral nonlinearity, drift, and noise will limit the usable resolution. On Mon, 10 Dec 2012 15:08:38 -0600, Didier Juges shali...@gmail.com wrote: How long does it take to prove it? And what's the point? On Mon, Dec 10, 2012 at 3:04 PM, Gerhard Hoffmann dk...@arcor.de wrote: Am 10.12.2012 21:53, schrieb Didier Juges: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) With that much noise it is really guaranteed that no code will be missing. :-) Gerhard ___ 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 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] Thunderbolt oven / non-stable operating temperature
I feel like shouldn't need to fuss with ambient conditions this heavily for an OCXO, and find myself researching construction / design for a DIY outer-oven to wrap the thunderbolt in. Anyone have experience with non-stable temperature on a trimble thunderbolt? The temperature you see is not from the OCXO but from the thermometer chip near the RS-232 connector. The DS1620 is probably at fault and the rest of the Thunderbolt is operating as it should. I've see a lot of Thunderbolts and the most common failure mode of the non E revision DS1620 thermometer is for them to display -55 degrees C although it could just have erratic output like yours. The erratic temperature that it is reporting could cause the processor to try to compensate for the erratic jumps and cause the Thunderbolt output to be less stable. The DS1620 should be replaced. -Arthur ___ 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] 31bit ADC, 1000 samples per second
In message tilcc892a6ntfk64t2nljm92idr92df...@4ax.com, David writes: The gain drift is specified at 2ppm per degree C. There are provisions for a calibration cycle but that requires external multiplexing. That's the really smart thing about this particular chip: You can drive the geophone calibration signal in through the second input and out onto the first, which eliminates the external calibration mux and the errors that usually cause. And yes, in seismology you do care about frequencies from 0.1Hz and up (0.01Hz if you're involved with the CTBT) and most voltage references will be stable enough for that. For longer term you calibrate your entire system, starting with the geophone, so the voltage reference is caught that way. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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] Thunderbolt oven / non-stable operating temperature
On 12/10/2012 5:22 PM, Bob Camp wrote: Hi The oven in the OCXO keeps the crystal at a constant temperature. It is normal for the TBolt it's self to idle 10 to 20C above ambient. The temperature sensor is located near the 9 pin D connector. It shows a temperature somewhere in between the ambient and the OCXO case temperature. Some have mounted speed controlled fans to keep the TBolt case at a constant temperature. Bob Bob, others: Ah, so then it's probably fine. Thanks, the other posts which followed seem to be in agreement with yours. ... Now then. Suppose I'll likely return my efforts to finding a time constant that makes sense for my particular tbolt / usage needs. ___ 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] 31bit ADC, 1000 samples per second
In message camqqfunudahc9rdrk0g8hxo_kohs+bndeilnqspq1_b2nqk...@mail.gmail.com , Didier Juges writes: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) No missing codes is guaranteed by the design, you'd have to screw up the digital filters to come out with missing codes. That's not impossible to do, but the fix is simple: sufficient precision in the filters, hence the 31 bits. The 32nd bit is an overrange bit btw, if it tracks the sign bit you're fine, if it is inversed you are out of range. Smart detail. -- Poul-Henning Kamp | UNIX since Zilog Zeus 3.20 p...@freebsd.org | TCP/IP since RFC 956 FreeBSD committer | BSD since 4.3-tahoe Never attribute to malice what can adequately be explained by incompetence. ___ 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] Thunderbolt oven / non-stable operating temperature
I believe that the high temperature alarm you see is triggered at 50 degrees C. If that is what you're seeing without artificially raising the temperature of the Thunderbolt by insulating it so it can't radiate the heat, what I said about replacing the chip is correct but if it is staying within a few degrees over the course of the day and is in the 40 degree C range without being insulated, the DS1620 thermometer chip is o.k.. -Arthur ___ 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] Thunderbolt oven / non-stable operating temperature
On 12/10/2012 6:10 PM, Arthur Dent wrote: I believe that the high temperature alarm you see is triggered at 50 degrees C. If that is what you're seeing without artificially raising the temperature of the Thunderbolt by insulating it so it can't radiate the heat, what I said about replacing the chip is correct but if it is staying within a few degrees over the course of the day and is in the 40 degree C range without being insulated, the DS1620 thermometer chip is o.k.. My particular thunderbolt seems to be from 2004, so I guess that manual is at or around checked: ThunderBolt[tm] GPS Disciplined Clock User Guide Version 5.0 5.0 is a version of the manual published in 2003 Table C.2.2 Environmental Specifications: Operating Temp: -0°C to +60°C Storage Temp: -40°C to +85°C ... so I guess I shouldn't worry about a few degrees of difference in the reported temp, especially considering that the sensor in question isn't on the OCXO itself. ___ 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] Thunderbolt oven / non-stable operating temperature
Sarah, That temperature sensor does have an effect on the final outcome as it is part of the internal equation. So buffering the ambient temperature is important. You do not need to go crazy, but having it contained in a box with some small amount of heat applied and maintained by some controlling mechanism would be a good way to go. The amount of heat depends upon what extremes your location experiences over the day/week/year and the effective insulation of the container. Ideally you would want no temperature change, but, obviously, that is not practical, so a one degree variance would be a reasonable goal. BillWB6BNQ Sarah White wrote: On 12/10/2012 6:10 PM, Arthur Dent wrote: I believe that the high temperature alarm you see is triggered at 50 degrees C. If that is what you're seeing without artificially raising the temperature of the Thunderbolt by insulating it so it can't radiate the heat, what I said about replacing the chip is correct but if it is staying within a few degrees over the course of the day and is in the 40 degree C range without being insulated, the DS1620 thermometer chip is o.k.. My particular thunderbolt seems to be from 2004, so I guess that manual is at or around checked: ThunderBolt[tm] GPS Disciplined Clock User Guide Version 5.0 5.0 is a version of the manual published in 2003 Table C.2.2 Environmental Specifications: Operating Temp: -0°C to +60°C Storage Temp: -40°C to +85°C ... so I guess I shouldn't worry about a few degrees of difference in the reported temp, especially considering that the sensor in question isn't on the OCXO itself. ___ 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 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] 31bit ADC, 1000 samples per second
On Mon, 10 Dec 2012 23:01:49 +, Poul-Henning Kamp p...@phk.freebsd.dk wrote: In message tilcc892a6ntfk64t2nljm92idr92df...@4ax.com, David writes: The gain drift is specified at 2ppm per degree C. There are provisions for a calibration cycle but that requires external multiplexing. That's the really smart thing about this particular chip: You can drive the geophone calibration signal in through the second input and out onto the first, which eliminates the external calibration mux and the errors that usually cause. I see now from the data sheet that the 3000 ppm typical gain mismatch applies between the PGA gain settings and not between the multiplexor input channels so you can use the internal multiplexor for external calibration. The other instrumentation delta-sigma ADCs I am familiar with do a gain and offset calibration for every conversion which knocks the drift down by an order of magnitude unless you disable that feature. And yes, in seismology you do care about frequencies from 0.1Hz and up (0.01Hz if you're involved with the CTBT) and most voltage references will be stable enough for that. Voltage reference 1/f noise will be a problem but only because the ADS1282 input stage is already chopped allowing that kind of sensitivity at low frequencies. Competing instrumentation delta-signal converters have the same problem if you consider that a problem. For longer term you calibrate your entire system, starting with the geophone, so the voltage reference is caught that way. What they need is a ratiometric geophone . . . :) ___ 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] 31bit ADC, 1000 samples per second
Hi It's not at all uncommon for the seismic guys to go a bit uber-nuts on clean supplies and references. Bob On Dec 10, 2012, at 7:54 PM, David davidwh...@gmail.com wrote: On Mon, 10 Dec 2012 23:01:49 +, Poul-Henning Kamp p...@phk.freebsd.dk wrote: In message tilcc892a6ntfk64t2nljm92idr92df...@4ax.com, David writes: The gain drift is specified at 2ppm per degree C. There are provisions for a calibration cycle but that requires external multiplexing. That's the really smart thing about this particular chip: You can drive the geophone calibration signal in through the second input and out onto the first, which eliminates the external calibration mux and the errors that usually cause. I see now from the data sheet that the 3000 ppm typical gain mismatch applies between the PGA gain settings and not between the multiplexor input channels so you can use the internal multiplexor for external calibration. The other instrumentation delta-sigma ADCs I am familiar with do a gain and offset calibration for every conversion which knocks the drift down by an order of magnitude unless you disable that feature. And yes, in seismology you do care about frequencies from 0.1Hz and up (0.01Hz if you're involved with the CTBT) and most voltage references will be stable enough for that. Voltage reference 1/f noise will be a problem but only because the ADS1282 input stage is already chopped allowing that kind of sensitivity at low frequencies. Competing instrumentation delta-signal converters have the same problem if you consider that a problem. For longer term you calibrate your entire system, starting with the geophone, so the voltage reference is caught that way. What they need is a ratiometric geophone . . . :) ___ 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 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] 31bit ADC, 1000 samples per second
If it *was* missing codes in reality, it would not be the first time TI made that kind of mistake. One of their not so early multi-slope integrating converters was advertised to *not* return both plus zero and minus zero which was a big feature at the time . . . but it did. Almost nobody noticed that error for years. On Mon, 10 Dec 2012 16:28:52 -0600, Didier Juges shali...@gmail.com wrote: I agree, but there is a difference between something that you should be able to achieve by design and being able to prove it, or even if it's useful in practice. In a case like this, the wide span between resolution and effective bits makes me wonder what's the point of advertising the former other than bragging rights? On Mon, Dec 10, 2012 at 3:38 PM, David davidwh...@gmail.com wrote: Integrating converters including delta-sigma converters can be no missing codes by design without being able to take advantage of the full resolution that implies. Integral nonlinearity, drift, and noise will limit the usable resolution. On Mon, 10 Dec 2012 15:08:38 -0600, Didier Juges shali...@gmail.com wrote: How long does it take to prove it? And what's the point? On Mon, Dec 10, 2012 at 3:04 PM, Gerhard Hoffmann dk...@arcor.de wrote: Am 10.12.2012 21:53, schrieb Didier Juges: I do not understand how anyone can guaranty no missing codes when the lower 11 bits are essentially noise? (31 bits resolution versus 20 effective bits) With that much noise it is really guaranteed that no code will be missing. :-) Gerhard ___ 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 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 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] Thunderbolt oven / non-stable operating temperature
Hi Charles, Well, perhaps you are not looking close enough. That is you need to be observing at a finer level of comparison. The changes, observed here and at another location, are in parts in 10-10 to 10-11 range, sometimes larger. At one of the locations there was a direct correlation to the air conditioning cycle. BillWB6BNQ Charles P. Steinmetz wrote: That temperature sensor does have an effect on the final outcome as it is part of the internal equation. So buffering the ambient temperature is important. I've heard this before, but the evidence I have seen does not seem to support the proposition. While switching the Dallas chip in one, I used the opportunity to bring the chip temporarily outside of the Tbolt housing on a cable to investigate whether the Tbolt makes any internal use of the temperature data. Neither freeze spray nor bringing a soldering iron near the chip, when it was outside of the Tbolt housing and the Tbolt housing was well insulated from the changes in chip temperature, seemed to have any effect on the operation of the Tbolt, either normal or in holdover. I have also run Tbolts with the newer (wrong) temperature chips for long periods, and have not observed any systematic differences in performance between them and units with the older chips, either in normal operation or in holdover. In Tbolts with the newer chips, the reported temperature often has little connection with the actual temperature and, at times, jumps abruptly, yet the Thunderbolts operate normally with no corresponding jumps in operating parameters. My supposition/conclusion is that the temperature sensor was provided so telcom operators could get a rough idea of the temperature in remote cell-site transmitter shacks, not for internal use by the Tbolt. As long as the Tbolt is housed so that its reported temperature does not change too rapidly, the oven control loop will keep the crystal very close to its set temperature over a wide range of ambient temperatures. I have used this approach and have also actively controlled the housing temperature, and have not observed any material difference in frequency or timing stability between the two approaches. Best regards, Charles ___ 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 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] 5061A Beam Current question
I've got a couple questions on a early S/N 5061A I got a few days ago. It took a couple days but the Ion Pump current is now down to zero or at least less then one division on the meter. I'm able to get the unit to go into continuous operation mode by following the TURN-ON PROCEDURES but have a few qestions 1. The unit 'sings' I'm tone deaf but its a high pitch. It could take some doing to isolate where it is coming from. Any ideas ? 2. How much should the beam current change when turning the OSC FREQUENCY COARSE adjustment ? On time about 45 to 60 minutes The current scribbled on the door is 19, I'm running pretty much 30. When making the adjustment the difference between the peak and any other position is only plus/minus a max of 2 or 3. 3 And i guess for last, are there any false modes that will turn off the ALARM and turn on the Con't Operation light ? Thanks -pete PS Anyone have a side panel without the front handle broken ? ___ 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] Of possible interest
On 12/10/2012 02:53 PM, Jim Lux wrote: On 12/9/12 10:14 PM, DaveH wrote: http://www.scmagazine.com.au/News/325731,researchers-find-crippling-flaws-in -global-gps.aspx ___ 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. the actual paper is at: http://users.ece.cmu.edu/~dbrumley/courses/18487-f12/readings/Nov28_GPS.pdf We had one incident here in Sweden when a complete network went down due to a GPS bug. As PRN32 became enabled, it got selected and tracking data was collected in a sat[32] vector, which is obviously unsafe so it overwrote critical data and crashed. In fact, they all did this within a minute or so as it came up above the horizon. It took a bit of advanced guesswork from my side to find that flaw. It was one of those blame the solar flare incidents, but we had data from the solar flare measurements that showed a small deviation, the day after. So yes, you can attack this way. Cheers, Magnus ___ 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.