[time-nuts] hp10811 OCXO shipping jolt/drop survival?

Bruce Griffiths bruce.griffiths at xtra.co.nz
Sun Oct 11 03:26:47 UTC 2009


swingbyte wrote:
> Bruce Griffiths wrote:
>> swingbyte wrote:
>>  
>>> Bruce Griffiths wrote:
>>>    
>>>> swingbyte wrote:
>>>>  
>>>>      
>>>>> HI all ,
>>>>> I had a 5370 A shipped to me that suffered enough of a drop/roll to
>>>>> break a handle and bend the chassis such that the bottom panel had to
>>>>> be modified to get it back on.  When I connected the timebase output
>>>>> to the start input the frequency display is 10.0000000xx with the
>>>>> last
>>>>> two numbers moving about quite a lot.  Could this indicate damage to
>>>>> the oscillator due to the drop?  I haven't tuned up the inputs yet so
>>>>> this may be an artefact of that.
>>>>> Thanks
>>>>>
>>>>> Tim
>>>>>             
>>>> Tim
>>>>
>>>> What gate time was used to measure the frequency?
>>>>
>>>> Bruce
>>>>
>>>>
>>>> _______________________________________________
>>>> 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.
>>>>
>>>>         
>>> Hi Bruce,
>>> With a 1s gate time it ranges from 9.999 999 996 to 10.000 000 012
>>> MHz  The 5335 with a 1s gate time gives 10.00000101 with the last
>>> digit flickering 0.   Of course what's really bad is I have two clocks
>>> to check now.  The std deviation  as measured by the 5370 of the 5370
>>> clock is ~1.8 kHz while the std dev of the 5335 clock is 7.8 kHz -
>>> both have 10811 OCXOs.. The 5335 measures the 5370 clock  std dev as
>>> 4.35mHz.  I think I'll have to check these against the Rb as I don't
>>> have a GPSDO yet.
>>> Thanks
>>>
>>> Tim
>>>
>>>
>>>     
>> Tim
>>
>> What was the gate time used for the STD deviation measurements?
>>
>> Bruce
>>
>>   
> I used a 1s gate time on the 5335 and 100ksamples to get a std
> deviation of 4.35mHz  (thats milli Hertz)  so I hope that means they
> are both (5335 & 5370) fairly stable.  With a  1s gate time on the
> 5370  the reading is 10.000 000 000  3 to 10.000 000 001 2 MHz.  which
> I read as 1 part in 10E10.   On the issue of 5370 gate time for the
> std dev is 1 period and 100 samples.  The 5335 measures the clock of
> the 5370 as 10.000 001 06 so maybe one of the 10811s is 1Hz high?  The
> problem of two clocks.
> For the performance tests I get the following results:
> (13)  period 100.00 +-1ns  =>101.17ns
> (14) => 101.1843ns
> (15) jitter test should be <100ps => 38ps
> Trigger levels range from -1.34V to +0.61V
>
> My unit is a 2036A ser # so the first to have the 10811 as standard-
> still  only 29 years old!
> What is the "ic rot"  John was talking about and what solutions are
> out there?
>
> Thanks for your help
> Tim
>
The IC rot sounds like a corrosion or contamination problem that
(gradually?) changes the IC characteristics.
The only cure is to replace the affected ICs and/or input board(s) with
versions that don't suffer from the effect.

My 5370A is a similar vintage (first owner Motorola - they last
calibrated it in 2002) without an internal 10811A.
Jitter is about 15ps when measuring the period of its own 10MHz output.

One of the HP counters (5345A) of a similar vintage used virtually
identical input amplifiers.


Bruce




More information about the time-nuts mailing list