Re: [time-nuts] HP 10811 Response I Replies
Does your Mode B 10811 use a different xtral or same xtral, different frequency? What does Mode B do better, ... and worse? Why was it not used in the 10811? When trouble shooting a freq stability problem, usually the more data the better. What can really help to narrow down the cause or see the cause/effect, is to study a High Speed (100+ sps), High Resolution (1e-12) plot of the frequency change with time. That way one can see if the freq change is instantaneous or happens in steps, is the change monotonic, if there is a time constant to the freq change what is the TC's value and shape, is the freq change in fixed values or random noisy steps and now and if the freq returns to the original value. With a good freq plot, the source of the problem can often be narrow down to a single possible source. And the best tool that I know of for making that kind of plot is a TPLL-2.0 tester. Here is a freq plot that has nothing to do with the existing problem, It is just an example where a TPLL-1.0 freq plot is useful in identifying the source of some noise. http://www.febo.com/pipermail/time-nuts/attachments/20100615/e4c25279/attachment-0001.gif and another example showing a plot of purposely induced Noise that only a few instruments could even measure: http://www.thegleam.com/ke5fx/tpll/disturb_zoom.gif ws ** WarrenS wrote: I have to wonder if the unit being tested had its high impedance oven control points lifted off the PCB board and on Teflon standoffs like the production units? ws It was a production unit, no modifications whatsoever. The oven change is an interesting theory; I never thought of that. I have a mode B 10811 on hand that I could use to test that theory. Rick Karlquist N6RK ___ 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 Response I Replies
WarrenS wrote: Does your Mode B 10811 use a different xtral or same xtral, different frequency? What does Mode B do better, ... and worse? Why was it not used in the 10811? Mode B is the same crystal on a different frequency (about 10.95 MHz). It does not have a turnover and the tempco is much higher than mode C. It makes a very sensitive thermometer and is therefore useful for evaluating the oven. The reason is not used in a production 10811 is obviously that for timebase use, you definitely don't want a high tempco. Rick ___ 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 Response I Replies
On 23/09/11 05:40, Richard (Rick) Karlquist wrote: On 9/22/2011 5:17 PM, ws at Yahoo wrote: Within minutes the frequency changed more than the spec For humidity to get thru something like that it takes weeks or more it does it at all. That fast of reaction, Sure sounds like some other effect like blowing a little air on the case or loading the osc output with water in the output cable etc, etc. I think it is safe to say the effect was not due to water inside, unless there was a hole. ws The effect we saw was like parts in 10^8 of something. Way too big to be related to output loading or air on the case. In any event, the air blowing on the case was constant during the test. We saw that when we didn't even try to seal the 10811 and also when we tried to seal it. Blowing on the case? Did you consider the increased cooling that higher humidity provides? 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.
Re: [time-nuts] HP 10811 Response I Replies
I have made plots of the effects of everything I can find that effects the freq of a HP10811. Most things are much slower than minutes, more like an hour time constant, such as anything effecting the outside case's temperature OR the effect is much faster than minutes, things such as voltage and load changes. The only thing that I've seen that responds to instantaneous changes within minutes is the oven loop. If one assumed Rick's data is correct, then one likely cause of what was happing (not why it was happening) is that the oven set point temperature was changed. That is the ONLY thing I've seen that can cause that sort of freq change over that kind of time period. It does not take much, a degree or so change in the oven's set point could do it. If I saw a real plot of freq change with time, I could be more sure of the cause. But even if the exact reason is not known, This string does show, If you want the best from your 10811 then it is NOT a good idea to heat it up in an oven to near its limits and then dump in water. I have to wonder if the unit being tested had its high impedance oven control points lifted off the PCB board and on Teflon standoffs like the production units? If not just blowing with your mouth on theses sensitive points could cause this type of freq change. ws ** Within minutes the frequency changed more than the spec Rick For humidity to get thru something like that it takes weeks or more IF it does it at all. That fast of reaction, Sure sounds like some other effect like blowing a little air on the case or loading the osc output with water in the output cable etc, etc. I think it is safe to say the effect was not due to water inside, unless there was a hole. ws *** The effect we saw was like parts in 10^8 of something. Way too big to be related to output loading or air on the case. In any event, the air blowing on the case was constant during the test. We saw that when we didn't even try to seal the 10811 and also when we tried to seal it. Richard (Rick) Karlquist ** Blowing on the case? Did you consider the increased cooling that higher humidity provides? 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.
Re: [time-nuts] HP 10811 Response I Replies
WarrenS wrote: I have to wonder if the unit being tested had its high impedance oven control points lifted off the PCB board and on Teflon standoffs like the production units? ws It was a production unit, no modifications whatsoever. The oven change is an interesting theory; I never thought of that. I have a mode B 10811 on hand that I could use to test that theory. Rick Karlquist N6RK ___ 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 Response I Replies
Magnus Danielson wrote: On 23/09/11 05:40, Richard (Rick) Karlquist wrote: Blowing on the case? Did you consider the increased cooling that higher humidity provides? Cheers, Magnus Against, the frequency change we saw was considerably more than what we got from changing the temperature 50 degrees. rick ___ 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 Response I Replies
Maybe you have a bad joint in the tuning circuit and the humidity makes the actual tuning voltage vary. -John === Magnus Danielson wrote: On 23/09/11 05:40, Richard (Rick) Karlquist wrote: Blowing on the case? Did you consider the increased cooling that higher humidity provides? Cheers, Magnus Against, the frequency change we saw was considerably more than what we got from changing the temperature 50 degrees. rick ___ 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] HP 10811 Response I Replies
On 23/09/11 23:25, Rick Karlquist wrote: Magnus Danielson wrote: On 23/09/11 05:40, Richard (Rick) Karlquist wrote: Blowing on the case? Did you consider the increased cooling that higher humidity provides? Cheers, Magnus Against, the frequency change we saw was considerably more than what we got from changing the temperature 50 degrees. Which would not be well explained by that effect, true. 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.
[time-nuts] HP 10811 Response I Replies
Esteemed Colleagues, I will attempt to make a coherent technical reply to the remarks made about my 10811 posting. But first a bit of qualification. My remarks refer to the 10811 and that vintage of OCXO’s. Wrote: Doing what you describe will result in a very sensitive humidity sensor, having eliminated the thermometer effect. I do not understand. I believed that since the OCXO temperature will be substationally higher than the surrounding temps, any residual moisture would migrate to a lower temperature. The fiberglass insulation inside the vacuum flask prevents air currents, reduces air volume, and provides mechanical isolation. If I err, what would be a better solution? There are similar, well lauded, designs using large blocks of styrofoam instead (Shera). Why would foam be better that a vacuum bottle? Wrote: In any event, none of this affects crystal aging or frequency jumps. So true, but I never said it would. All I proposed was to as much as possible eliminate any exterior variables so all one would see is only the crystal aging characteristics. Wrote: This is what limits the E1938A which is hermetic and has a thermal gain 1000 times better than the 10811. I know it is a splendid OCXO, a much later design and costs your new-born child IF a used one can be found. Rick, you have probably forgotten more about OCXO’s than I’ll ever know. My questions are not argumentative but are for seeking knowledge. Wrote: Just how good is the thermos bottle in this case? (as in degrees / watt). You can get some very good vacuum ones and some pretty poor Styrofoam ones. I don’t know anything about its degrees/watt. It is a true wide mouth glass vacuum thermos with a protective plastic inner lining that I bought a long, long time ago. It was sold for taking hot soup to work. Speaking of Styrofoam. If you want to use one try to get one from Omaha Steaks. Wrote: Imagine that the set point is variable, and can be set below the desired temperature. Then imagine that the set point can approach the desired temperature more closely as it gets closer to the desired temperature. ... And you will have discovered (100 years late) the PID controller. It still works on the concept of successive approximation. It may have less over-shoot on start-up and the amount of over-shoot may not have a negligible effect. But it still will be there. Perhaps if one was designing a new product that might be the best solution. However my remarks were about the HP 10811. It isn’t applicable in this situation. Wrote: 30 years ago I was designing PID controllers, with a little microprocessor magic, that could quickly arrive at the set point temperature and never, I repeat, never, exceed that temperature. Someone's internal organs would have become toast if it did. From 1976 to 2001 I repaired life support equipment including hemodialysis machines (factory trained) and blood warmers. They only used thermistor bridge or mechanical thermostats. And yes we had to make sure the blood temperature never, ever went over 102F. Probably at that time reliable µprocessor control was not generally cost effective. But I don’t think that PIC technology you did would apply to OCXO’s. Here’s why: heat is added to either large amount of cold fluids entering the body or to substitute for the body’s inability to maintain proper body heat (radiant infant warmers). You have a large ongoing thermal sink. In an extremely well insulated OCXO as I understand it, there is almost no thermal sink once the set point is reached. And then there is the cost issue. Again, not to argue with your type of controller. I’m just considering the current surplus OCXO’s that are available. Regards, Perrier ___ 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 Response I Replies
Perry Sandeen wrote: Wrote: Doing what you describe will result in a very sensitive humidity sensor, having eliminated the thermometer effect. I do not understand. I believed that since the OCXO temperature will be substationally higher than the surrounding temps, any residual moisture would migrate to a lower temperature. The fiberglass insulation inside Heating up a space does not change the absolute humidity AFAIK. It only changes the relative humidity. We did tests where we sealed a 10811 inside a box that was held together with so-called hermetic epoxy. We put it in an environmental chamber at a constant temperature and constant low humidity and let it stabilize. We then increased the humidity to something like 80%, while holding the temperature constant. Within minutes the frequency changed more than the spec for the entire temperature range. Therefore, you should do your experiment with the hermetic version of the 10811. The hermetic version is soldered shut, rather than using epoxy, which turns out not to be hermetic, no matter what they claim. Rick Karlquist ___ 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 Response I Replies
You had a leak. If epoxy was really as bad as you indicate, it would not be usable for holding pressure, or mild vacuum, and yet it is. Somehow, someway you left a big hole in the bucket. -Chuck Harris Rick Karlquist wrote: Perry Sandeen wrote: Wrote: Doing what you describe will result in a very sensitive humidity sensor, having eliminated the thermometer effect. I do not understand. I believed that since the OCXO temperature will be substationally higher than the surrounding temps, any residual moisture would migrate to a lower temperature. The fiberglass insulation inside Heating up a space does not change the absolute humidity AFAIK. It only changes the relative humidity. We did tests where we sealed a 10811 inside a box that was held together with so-called hermetic epoxy. We put it in an environmental chamber at a constant temperature and constant low humidity and let it stabilize. We then increased the humidity to something like 80%, while holding the temperature constant. Within minutes the frequency changed more than the spec for the entire temperature range. Therefore, you should do your experiment with the hermetic version of the 10811. The hermetic version is soldered shut, rather than using epoxy, which turns out not to be hermetic, no matter what they claim. Rick Karlquist ___ 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] HP 10811 Response I Replies
A problem I've seen when using the hermetic sealed soldered version of the 10811 from a dual oven unit, is when the case is sealed the osc makes a good barometer because of changes in its case due to barometric changes. A 1 inch difference (such as 30 to 29) caused something on the order of 1e-11 freg change. To see the effect just need to put a little pressure on the case So now I'm trying to determine which is worse, leaving the freq adj screw tight and sealed or leaking a little, inside a box containing desiccant. ws ** Heating up a space does not change the absolute humidity AFAIK. It only changes the relative humidity. We did tests where we sealed a 10811 inside a box that was held together with so-called hermetic epoxy. We put it in an environmental chamber at a constant temperature and constant low humidity and let it stabilize. We then increased the humidity to something like 80%, while holding the temperature constant. Within minutes the frequency changed more than the spec for the entire temperature range. Therefore, you should do your experiment with the hermetic version of the 10811. The hermetic version is soldered shut, rather than using epoxy, which turns out not to be hermetic, no matter what they claim. Rick Karlquist ___ 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 Response I Replies
Within minutes the frequency changed more than the spec For humidity to get thru something like that it takes weeks or more it does it at all. That fast of reaction, Sure sounds like some other effect like blowing a little air on the case or loading the osc output with water in the output cable etc, etc. I think it is safe to say the effect was not due to water inside, unless there was a hole. ws * Chuck Harris cfharris at erols.com You had a leak. If epoxy was really as bad as you indicate, it would not be usable for holding pressure, or mild vacuum, and yet it is. Somehow, someway you left a big hole in the bucket. -Chuck Harris Rick Karlquist wrote: Perry Sandeen wrote: Wrote: Doing what you describe will result in a very sensitive humidity sensor, having eliminated the thermometer effect. I do not understand. I believed that since the OCXO temperature will be substationally higher than the surrounding temps, any residual moisture would migrate to a lower temperature. The fiberglass insulation inside Heating up a space does not change the absolute humidity AFAIK. It only changes the relative humidity. We did tests where we sealed a 10811 inside a box that was held together with so-called hermetic epoxy. We put it in an environmental chamber at a constant temperature and constant low humidity and let it stabilize. We then increased the humidity to something like 80%, while holding the temperature constant. Within minutes the frequency changed more than the spec for the entire temperature range. Therefore, you should do your experiment with the hermetic version of the 10811. The hermetic version is soldered shut, rather than using epoxy, which turns out not to be hermetic, no matter what they claim. Rick Karlquist ___ 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 Response I Replies
Humidity is a confusing subject to many engineers and scientists. Unlike most parameters it is a quantity with two input variables, concentration and temperature. There are many ways to combine these to give different units. As a research scientist I spent most of my career working with composite materials which exhibit great sensitivity to humidity. Most composite materials respond to the Relative Humidity with only a small temperature dependance. So 80% RH has the same effect at any temperature. Note that the Absolute Humidity varies exponentially with temperature. For a fixed Absolute Humidity (say 10gms /m3) and at 70% RH the relative humidity changes about 10% per degree celsius. So if you have a sealed container with some water vapour in it the RH will vary about 10% per C*. If you have a fixed ambient humidity, a heated enclosure will have a humidity that falls 10%/K as the temperature rises. Now there are many grandiose environmental chambers sold to scientists and engineers that perform poorly. They have internal temperature gradients, so even if the concentration of water vapour is uniform the distribution of relative humidity is not. If a chamber set to 80% RH has a 2 degree gradient it could have internal condensation. The problems are made worse by the plethora of nearly useless humidity probes made by manufacturers who are having a bidding war based on claimed specs. Since there are very few facilities to calibrate humidity sensors, and no company can make a dollar by having their humidity measured more accurately, there is no pressure to improve instrument quality and the situation remains that there is a lot of misunderstanding about humidity. The instruments I have built have an inaccuracy of less than 0.1% RH, and I have built isothermal chambers that can be programmed to 0.1% RH. They are based on calculable processes for calibration, and so have absolute calibration. In the case of quartz crystals in ovens, when the oven is 30K above ambient the relative humidity is very low, so you would expect there to be very little absorbed or adsorbed water to interfere with stability. The main effects are surface leakage on hydrophilic surfaces and dielectric absorption in composite material insulators. There is a second order effect that the dielectric constant of air changes with absolute humidity. Humidity sensitivity would seem to me to be a problem of the measurement system rather than the item being tested. Cheers, Neville Michie On 23/09/2011, at 9:07 AM, Rick Karlquist wrote: Perry Sandeen wrote: Wrote: Doing what you describe will result in a very sensitive humidity sensor, having eliminated the thermometer effect. I do not understand. I believed that since the OCXO temperature will be substationally higher than the surrounding temps, any residual moisture would migrate to a lower temperature. The fiberglass insulation inside Heating up a space does not change the absolute humidity AFAIK. It only changes the relative humidity. We did tests where we sealed a 10811 inside a box that was held together with so-called hermetic epoxy. We put it in an environmental chamber at a constant temperature and constant low humidity and let it stabilize. We then increased the humidity to something like 80%, while holding the temperature constant. Within minutes the frequency changed more than the spec for the entire temperature range. Therefore, you should do your experiment with the hermetic version of the 10811. The hermetic version is soldered shut, rather than using epoxy, which turns out not to be hermetic, no matter what they claim. Rick Karlquist ___ 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] HP 10811 Response I Replies
On 9/22/2011 5:17 PM, ws at Yahoo wrote: Within minutes the frequency changed more than the spec For humidity to get thru something like that it takes weeks or more it does it at all. That fast of reaction, Sure sounds like some other effect like blowing a little air on the case or loading the osc output with water in the output cable etc, etc. I think it is safe to say the effect was not due to water inside, unless there was a hole. ws The effect we saw was like parts in 10^8 of something. Way too big to be related to output loading or air on the case. In any event, the air blowing on the case was constant during the test. We saw that when we didn't even try to seal the 10811 and also when we tried to seal it. Rick Karlquist ___ 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.