[time-nuts] Thunderbolt stability and ambient temperature

[Arnold Tibus]

Ed, 
and the big group

I did run such a 10811 without the outer heater connected and without 
any insulation around as replacement for the original insulation and box 
but I did not see any fluctuations beside the normal behaviour of a 10811.

(I have to admit that I did not yet run precision records, neither for the 
heater current nor for the frequency output. I see what I have to do soon!)

This model is obviously adjusted inside to work best in an encreased 
environment of around 40 to 45 deg. C . 

I think as Ulrich does explain, blocking the heat flow significantly must 
deteriorate the regulating loop parameters. The life time of the oscillator 
electronics outside the stabilized crystal element will as well decrease...

A good idea in fact to put the standard in a stable area and very slow 
varying temp range of around 15 deg. C underneath the house 
fundaments...I will think about!

Did anybody ever think to apply wood as stabilizing element around an
oscillator? Wood behaves a bit strange, it is on one side a not bad 
insulator and it does have as well a quite high heat storage capacity! 

(Oak with around 2.4 J/g, K., Lambda 0.17 W/m.K,,
Pine around 2.7 J/g.K., Lambda 0.14 W/m.K.
Water around 4.18 J/g.K., Lambda 0,604 W/m.K.
Copper has 0.385 2.4 J/g, K., Lambda 401 W/m.K. at 25 deg. C. )
I consider wood a very intersting material, cheap and easy to work...

Arnold



On Thu, 11 Jun 2009 09:22:00 -0600, Ed Palmer wrote:

>Ulrich, when you changed the thermal characteristics around your 10811, 
>how 'crazy' did it become?

>According to your theory, there should be differences in the schematics 
>of the oven controllers between the single and double oven variants of 
>the 10811.  Last year I picked up a couple of those double oven 10811s 
>from China that had the outer insulation removed.  I wonder if I could 
>see any anomalies in its operation with no outer insulation and only the 
>inner oven operating.

>Ed

>Ulrich Bangert wrote:
>>> I'm not talking about crystal ovens, I'm talking about 
>>> timekeeping kit in general: how to filter high frequency 
>>> temperature fluctuations out, so the PLL does not have to 
>>> deal with them.
>>>     
>>
>> Ok, but in this general case you have to accept the fact, that filtering
>> high temperature fluctuations is only ONE part of the engineering job. The
>> second task (which I tried to point at) is: When you do already have a
>> closed loop temperature regulation system inside what you plan to built
>> around it, then you need to manage the filtering without mal interaction of
>> the local temperature controller with your additional filtering stuff. The
>> art is to make the closed temperature loop inside see only smaller
>> temperature fluctuations but No changes otherwise. 
>>
>> If you apply an overall change in temperature insulation to the ambient you
>> will make the work of the inside temperature controller a lot harder because
>> it is not prepared for that. In the aluminium box that i wrote about I have
>> for an experiment replaced the air between the 10811 and the outer box with
>> a material that has a significant lower temperature conductivity than
>> standing air. You can easily watch the temperature controller go crazy with
>> that. For a second experiment I have replaced the air with a material having
>> lots more of thermal conductivity than air and you can watch the temperature
>> controller go crazy with that too. You don't need sophisticated temperature
>> measurement equipment to see the controller go crazy, just watch the
>> frequency of the oscillator. This will tell you all about it. 
>>
>> Things are different, if you have an influence on the controller's
>> regulation parameters (which you do not have with a ready bought device). If
>> you do have, then your additional provisions for filtering and insulation
>> can easily be included into the thermal model that the temperature
>> controller has to handle and will lead to different operation parameters of
>> the controller. That will of course work and that is why I compared your
>> suggestions to designing a new oven (for xtal or rubidium or whatever). 
>>
>> Best regards
>> Ulrich Bangert
>>
>>   
>>> -----Ursprungliche Nachricht-----
>>> Von: time-nuts-bounces at febo.com 
>>> [mailto:time-nuts-bounces at febo.com] Im Auftrag von Poul-Henning Kamp
>>> Gesendet: Donnerstag, 11. Juni 2009 12:50
>>> An: Discussion of precise time and frequency measurement
>>> Betreff: Re: [time-nuts] Thunderbolt stability and ambient temperature
>>>
>>>
>>> In message <E182724B9FCE4B9F9E914BCC4F374401 at athlon>, "Ulrich 
>>> Bangert" writes:
>>>     
>>>>> For PLL steered devices, you want your device enclosed by a
>>>>> thermal mass which is again enclosed by a layer of thermal 
>>>>> isolation.  The goal is to filter/average all rapid (daily ?) 
>>>>> external temperature influences, only letting through such 
>>>>> slow variations (seasonal ?) which the PLL can comfortably 
>>>>>         
>>> cope with.
>>>     
>>>> My impression with this thread is that two sligthly different things 
>>>> are being discussed here:
>>>>
>>>> Some posters explain how a crystal oven shall be designed as 
>>>> Poul-Henning does with his words above. Absolute correct, no doubt 
>>>> about it!
>>>>       
>>> I'm not talking about crystal ovens, I'm talking about 
>>> timekeeping kit in general: how to filter high frequency 
>>> temperature fluctuations out, so the PLL does not have to 
>>> deal with them.
>>>
>>> The above advice applies to GPSDO's and Rb's and Cs's as well 
>>> as TCXOs, OCXOs and voltage references.
>>>
>>> Poul-Henning
>>>
>>> -- 
>>> Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
>>> phk at 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.
>>>