[time-nuts] Z3805A cooling requirements?

Bob Camp lists at rtty.us
Sun Dec 16 23:38:49 UTC 2012


Hi

Ok where to start.

Some typical numbers:

A DOCXO likely will be specified at around 0.3 ppb peak to peak over -30 to +70C. That comes out to 0.003 ppb per degree. A single OCOX likely will be specified at around 10 ppb peak to peak over -30 to +70C. That's 0.1 ppb per degree. Yes, there are a number of variations, but those numbers are not at all crazy. eBay will happy cough up a lot of parts with those sort of numbers on them. 

---------------------------

Both devices will heat the crystal to a constant temperature when the ambient (however it's measured) is in the range of -30 to +70C. If it's below -30, there may not be enough oven power to heat the crystal. More likely there's not enough power to warm it up in the specified warmup time. If the ambient is above +70, even by a few degrees, the crystal can overheat. Exactly what temperature it starts to overheat is a function of several things, mainly the tolerances on cutting process of the crystal. When it does overheat, it's likely to move 10 to 20 ppb per degree for that first degree. That's *way* more than you would ever want to see.

It would be nice if control loops had no noise, and if the crystal heated and cooled at the same rate. In fact, you have enough heater power to run things up at a very high rate (you want to warm the thing up very quickly). The cooling rate at the hot end is "not so fast" (small delta T = low heat flow) .  A noise spike that gets the crystal too hot, will hang around for a while. You can have short term stability trouble before the oven is fully shut off.

Simply put - OCXO's don't play nice beyond their upper ambient spec. 

-------------------------

In most gear, there's a fan, and it runs all the time. It may speed up a bit when things get real hot, but that's about it. With a fan like that, when the ambient moves a degree or two, the inside of the box moves a degree or two. There is no catastrophe when the fan suddenly comes on. If anything most variable speed fan setups result in the innards of the box seeing less temperature change rather than more. Thermally, fans are a good idea. 

------------------------

Now, taking those numbers (and a few assumptions, and the numbers from the previous post):

Let's say your gizmo does have a case that's 22C above ambient. Let's also say there is no fan. Let's also assume that there's something else in the box. It's pretty common for the "stuff" in a box to heat it up by 20C or more with no fan (The "ambient" inside the box is 20C above the outside air). If the gizmo is in the box, it's case is now 44C above the outside air. 

deep breath….

If the outside air is at 30C, your gizmo is at 74C. If it's an OCXO with a 70C upper end rating, it could be in trouble. The rest of the box is perfectly happy. It's designed with 70C rated parts and the spec on the box is 50C max.  

Yes, that's a lot of this and that. Yes there are some assumptions sprinkled here and there. Without assuming a few things there's not much way to get to a rational conclusion. Are OCXO's a bit strange thermally - yes of course they are. An eBay surplus rubidium is even stranger. Power transistors have issues as well. Stuff that has power associated with it needs some attention as part of the thermal design….

---------------------------

Are big / noisy / shaky / power hungry fans a good thing - of course not. They are simply the most common way to solve thermal problems in equipment. These days you can get fans that are less big/noisy/shaky than they once were. It's still better to not use them *if* you have all your thermal ducks in a row. That means you have the full specification on the OCXO (and likely a few other things). 

If we're talking about surplus gear or parts - you are not going to have a lot of information. The only reasonable assumption you can make is that the eBay OCXO is rated for moving air. Assuming you know the upper ambient of the OCXO, check it's actual case temperature. Keep it below the specified max ambient. Ideally, keep it 10 C or more below the rated upper ambient. The short term stability likely will be better ….

Bob 


On Dec 16, 2012, at 2:22 PM, Said Jackson <saidjack at aol.com> wrote:

> Hi Volker,
> 
> You are correct on that.
> 
> Bob is right in that thermal sensitivity is measured in a thermal chamber with constant airflow at a very constant rate and temperature. While this works, it is also somewhat unrealistic because who is going to set up a thermal chamber for their ocxo in the field?
> 
> Because of that many companies put airflow shields on top of the ocxo in the test chamber to simulate more real life conditions where massive airflow is not happening, and self heating needs to be taken into consideration.
> 
> Moving-air skews the operating power of the ocxo as it removes heat from the ocxo. Thus a manufacturer could claim +75C max operating temp inside a test chamber, but if you try to operate in still air at 75C in a small enclosure your ocxo would overheat quickly due to the internal power consumption adding to the ambient temperature. Thus the test chamber actually works to "cool" the ocxo by removing excess heat and keeping its case at 75C no matter how much power is consumed inside the ocxo.
> 
> As an example of this consider that a typical DOCXO runs at 55C case temp in still air with 22C ambient. Now say that Docxo has 1E-010 per degree C thermal sensitivity (not a great docxo.. But thats actually better than the spec of the ocxo used on the Mini-T: that one is 10ppb from 0C to 60 C as far as I know)
> 
> What happens when I turn on a strong fan right next to the Docxo? The fan will throw ambient air at the same 22C temperature at the unit, and immediately start cooling off the 55C case of the ocxo due to the temp difference between the ocxo case and ambient air.
> 
> The result? If the fan can cool off the Docxo to say 30C, we have had a massive 25C temperature change shock on the docxo without a single C temp change in ambient air!
> 
> Now 25C * 1E-010/C = 2.5E-09 change in frequency which gives a 2.5ns/s drift rate just because the fan switched on!
> 
> Typical single oven ocxos will have about 1ppb per C sensitivity, so the above example would result in 25ns/s drift just because the fan came on. Thats really bad for Gpsdo type performance expectations, and will certainly ruin your ADEV performance which us time-nuts expect to be around 1E-012 not 2E-08 for a good gpsdo :)
> 
> Hope this shows why a fan on an ocxo is not a good idea.
> 
> Bye,
> Said
> 
> 
> 
> 
> 
> 
> 
> On Dec 16, 2012, at 9:47, Volker Esper <ailer2 at t-online.de> wrote:
> 
>> 
>> Indeed? I didn't expect that. There are people who say, that the control loop of OCXOs is optimized for still air and no additional cooling at all.
>> 
>> Said told us, that...
>> 
>>> ...a fan is about the worst thing you can do for your Z3805 it will
>> significantly worsen the stability of the output frequency...
>> 
>> Since the main task of the OCXO-oven is to stabilize the internal temperature, I can't imagine, that it get's into trouble when not externally cooled!?
>> 
>> If I'd ventilate the air around the OCXO case the heater had to work more and the power dissipation would be greater.
>> 
>> Am I wrong with taht?
>> 
>> Volker
>> 
>> 
>> 
>> Am 16.12.2012 15:35, schrieb Bob Camp:
>>> Hi
>>> 
>>> By far the most common way to test and certify OCXO's is in moving air. It's rare to see one get in trouble from to much ventilation. The more common problem is thermal runaway due to inadequate ventilation.
>>> 
>>> Bob
>>> 
>>> On Dec 16, 2012, at 7:57 AM, Volker Esper<ailer2 at t-online.de>  wrote:
>>> 
>>>> 
>>>> Steward,
>>>> 
>>>> What is the intended and what is the actual supply voltage? Which current is the unit consuming?
>>>> When we know that, we can compare the power consumption with our units. If it is in the same range, it should - with a little luck - be working properly.
>>>> 
>>>> My two units are intended to be supplied by 19.5 to 30 V. I use 24 V and the Z3805s draw 0.9 A each. If I increase voltage the current decreases (typical for the switching supplies inside the Z3805).
>>>> 
>>>> I don't cool the units, they just lie on an old electronics magazine (for not to scratch the case of my signal generator, lying below the magazine), so they can freely convect their heat.
>>>> 
>>>> Cheers
>>>> 
>>>> Volker
>>>> 
>>>> 
>>>> 
>>>> Am 12.12.2012 01:21, schrieb Stewart Cobb:
>>>>> This may be a newbie question, but I'm a newbie, so:
>>>>> 
>>>>> Do the HP telecom GPSDOs (Z38xx) require external airflow for cooling?
>>>>> They don't have built-in fans, but they sorta look like they depend on a
>>>>> rack-level cooling fan, which a telecom rack would almost certainly have.
>>>>> 
>>>>> I ask because I bought a Z3816 awhile back which worked for about a week
>>>>> and then failed. I traced the failure to an internal power supply brick,
>>>>> which had a big finned heat-sink attached but nevertheless smelled
>>>>> overheated and was shorted internally.
>>>>> 
>>>>> I never found a replacement power brick, and I don't have time to mess with
>>>>> it right now, so I recently bought a Z3805A. It, too, looks like it's
>>>>> working, but it started to feel awfully warm after a few hours, so I
>>>>> unplugged it for now.
>>>>> 
>>>>> It probably wouldn't take much of a fan to bring the internal temperature
>>>>> down close to ambient, and the fan could be powered easily enough from the
>>>>> supply rails. But that might create a temperature gradient where the
>>>>> designers didn't intend one. Or it might cause problems I don't even know
>>>>> about yet.
>>>>> 
>>>>> At the moment, the Z3805A is in a fan-less 19-inch rack with a bunch of
>>>>> other equipment, in a lab environment. Should it have its own fan?
>>>>> 
>>>>> Cheers!
>>>>> --Stu
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at 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 at 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 at 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 at 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 at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.




More information about the time-nuts mailing list