[time-nuts] Re: Oscilloquartz 8663-XS stability problem

[Brian K]

"The phenomenon that you see may be due to moisture absorbed/adsorbed by one or more of the components. While in storage the unit equilibrates with maybe 60% RH. When the oven is operated the RH is reduced to near zero. (Each degree C reduces RH by about 10%)
It may take weeks to settle."

The 8663-XS can is hermetically sealed so I doubt if it is a humidity problem.
As I pointed out in the eevblog thread, there is a solder blob on the side of the can.
I suspect this seals a hole used to fill the inside with an inert gas such as dry Nitrogen.

Many years ago I worked for Marconi Instruments.
One project I worked on was a Selective Level Meter used to test analogue FDM Telecomms systems.
During environmental testing they discovered problems with crystal filter frequency response changing at high humidity.
The solution was to fit the filter inside a sealed container filled with Nitrogen.
I suspect Oscilloquartz and other OCXO manufacturers do a similar thing.
Some MTI OCXOs have a similar solder blob.

________________________________
From: Neville Michie <namichie at gmail.com>
Sent: 28 October 2023 00:38
To: Discussion of precise time and frequency measurement <time-nuts at lists.febo.com>
Cc: Brian K <brian_mk at hotmail.co.uk>
Subject: Re: [time-nuts] Oscilloquartz 8663-XS stability problem

The phenomenon that you see may be due to moisture absorbed/adsorbed by one or more of the components. While in storage the unit equilibrates with maybe 60% RH. When the oven is operated the RH is reduced to near zero. (Each degree C reduces RH by about 10%)
It may take weeks to settle.
Cheers
Neville Michie


Sent from my iPhone

> On 28 Oct 2023, at 10:06 am, Brian K via time-nuts <time-nuts at lists.febo.com> wrote:
>
> I recently constructed a homebrew 10MHz distribution amp.
> It has provision for both an external 10MHz source such as a GPSDO or an internal OCXO.
> For the internal OCXO I chose a used Oscilloquartz 8663-XS from China.
>
> When I began testing the unit I discovered an issue with frequency / temperature stability.
>
> I plotted a graph that shows the external temperature of the OCXO can as the device warms up.
> It stabilises at 60degC after around 45 minutes.
>
> I also plotted the frequency error measured using a Racal 1998 with an external 10MHz clock from a GPSDO.
> The frequency also appears to stabilise after around 45 minutes.
>
> [cid:0acb1b9d-a986-47a3-85de-53b95bb8479e]
>
> So far so good.
>
> However I find that if I leave the unit running for a few hours, the frequency gradually drifts upwards by up to 0.5Hz.
> That is well outside the 8663-XS specification. It is supposed to remain within 4x10^-9 from -20 to +70degC (I assume that is external ambient temp).
> The OCXO temperature (as measured on the outer can) remains constant at 60 degC.
> The tuning voltage and the supply voltage (12V) also remain constant.
> The tuning voltage comes from a LH0070 10V precision reference and a 10 turn 10K trimmer.
> I expected the tuning voltage required for 10MHz to be around 5V. The spec indicates a range of 0->10V.
> I found that the tuning voltage needed to be set to just below 10V to get 10MHz.
> Initially I thought that perhaps that's because it's an old used device.
> Subsequent tests however make me think it's related to temperature.
>
> I find that if I remove the equipment case lid, the frequency drifts slowly downwards.
> When I replace the lid, the frequency error slowly recovers.
> I assume the OCXO temperature drops slightly when the lid is removed despite it having ventilation slots directly above the OCXO.
> I expected the OCXO internal temperature to be maintained to within around +/- 1degC by the heater control circuit such that it would compensate for small external temperature variations.
>
> I am at a loss to explain what is going on and am running out of ideas.
> It appears that the OCXO internal temperature regulation is not working as well as it should.
> Is it just my particular device that behaves in this way or is it a common issue with this OCXO?
>
> I don't want to buy another used device from China (around 40 GBP) only to find the same problem.
>
> I would welcome any thoughts/suggestions from the experts here at time-nuts.
>
> For the full story, please refer to my recent thread in the eevblog forum:-
> https://www.eevblog.com/forum/projects/oscilloquartz-8663-xs-ocxo/
>
>
> [cid:18711b70-068e-474b-8ab5-74e886152aa1]
>
> Regards,
>
> Brian.
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com
> To unsubscribe send an email to time-nuts-leave at lists.febo.com