[time-nuts] Re: HP 5065A repair

[Matt Huszagh]

I'm back to troubleshooting the A3 multiplier assembly because if the
60 + 5.315... MHz signal to the SRD isn't correct, I wouldn't expect to
know if and when de-flooding succeeds.

The 5.315... MHz component is a bit weak - in the neighborhood of 5mVrms
instead of the expected 20mVrms. Additionally, the AGC DC voltage (TP2)
is 70 mV, but should be around 4.5 V.

I've got A3 hooked up in isolation on my bench with 5MHz at 1Vrms to J2
and 600mVrms to J5 at 5.315MHz. Any suggestions for how to troubleshoot
this? I measured Q4 and C12 to ensure they're not shorted - they're
fine. I also checked different points in the circuit and compared them
with the voltage annotations in the schematic. Unfortunately, I'm having
some trouble understanding this circuit, and some of the annotations
confuse me.

According to the annotations, the base of Q4 (the AGC amplifier), is at
+0.7Vdc for no AGC and +6Vdc for full AGC. But there's no emitter
resistor so how does this get to +6Vdc? Moreover, I don't see a DC path
to supply current to raise the DC voltage. In case it's helpful, Q4 is
apparently a 2N2484.

Another example is Q9 base is labelled with 2.4V. I think that's dc
since I see Vrms in a number of other places. But I don't see how that
could be true. This is fed via a +20V supply, and the base is biased by
a resistive dividier with 18k and 2k, so it should be at most +2V, but
then there's the parallel path of the BE junction and emitter resistor,
so it would be even lower than this. So, where are they getting 2.4V?

I checked a number of points in the main signal path and some looked
more or less in line with amplitude annotations, but others didn't. I'm
not totally sure how precise these annotations are supposed to be,
though.

Finally, I wanted to clarify how the RVFR SRD bias works. The only DC
path to the bias test point is through J4 (which provides the
60 + 5.315... MHz signal to the SRD). But, the RVFR schematic does not
show any DC bias applied to the SRD. I presume that's there but just not
shown. Is that right? That would make sense, since the pot there (R40)
would divide down the bias.

Any suggestions would be immensely appreciated.

Thank you!
Matt

Matt Huszagh <huszaghmatt at gmail.com> writes:

> I've decide to start by addressing the cell flooding. Despite the
> incorrect AGC DC voltage, the 60 MHz signal looks roughly ok and from a
> quick look I don't think I can really perform the loop alignment without
> some sort of RVFR signal.
>
> I've attached a current-limited supply at 5V through a 5ohm power
> resistor to the TED wires and have the unit powered with the meter set
> to 2nd harmonic. The meter indication almost immediately started
> deflecting off 0. It's been just a few minutes and the readings are
> getting almost to 10 at max deflection, though deflecting back and
> forth. I'm hopeful this is a good sign. I'll update as this unit has
> more time to de-flood.
>
> Please feel free to make suggestions if I should be following a
> different approach at the moment.
>
> Thanks!
> Matt
>
> Matt Huszagh <huszaghmatt at gmail.com> writes:
>
>> There are a few other things I wanted to mention.
>>
>> I checked the DC voltages around the cell and lamp heaters. J15P3 (cell
>> heater) measured 15.4V and J15P5 measured 14.4V. P4 measured 28.6V. The
>> specified nominal voltages for P3 and P5 are 18V, so these are a bit
>> off, but since the heater currents look ok as measured by the front
>> panel meter, I got the sense that this is probably ok. Thoughts?
>>
>> Also, as part of the pre-emptive maintenance I replaced the lamp
>> assembly film resistor and, in the process, disassembled part of A12
>> including the lamp assembly. Everything looked fine to my eye and I have
>> pictures I can share if desired. In particular, the lamp assembly was
>> not charred.
>>
>> Matt
>>
>> Matt Huszagh <huszaghmatt at gmail.com> writes:
>>
>>> Hi,
>>>
>>> I have an HP 5065A received in unknown condition. After performing some
>>> pre-emptive maintenance (basically Corby's recommendations here:
>>> https://www.eevblog.com/forum/metrology/inspecting-aligning-and-testing-an-hp-5065a-rubidium/
>>> and the first aid advice from Luciano here:
>>> http://www.timeok.it/hp5065a-corner-3/) I turned the unit on. +/- 20V
>>> supplies look good and cell and lamp ovens drop down from pegged in
>>> under 45 min to about 32 each.
>>>
>>> However, the 2nd harmonic reading is 0 and of course the unit does not
>>> lock.
>>>
>>> This unit was left unpowered for some time (I got the sense it was at
>>> least several years) before I received it, and therefore cell flooding
>>> is likely. However, I have noticed a few other discrepencies.
>>>
>>> The first is that the A3 TP2 AGC DC voltage measures 71 mV, instead of
>>> the expected >4.5 V. I also checked the 60 MHz signal into 50 ohms and
>>> see about 30 mW. One place in the manual says this should be >1V (which
>>> it is), but another place says 300 mW. Is loading this with 50 ohms
>>> correct? If so, maybe 300 mW is a typo and it should read 30 mW? In any
>>> event the AGC DC voltage is wrong and I suppose needs to be addressed.
>>>
>>> The second issue, and probably a more serious one, is that the PHOTO I
>>> meter reading is 20 (less than the specified 25) and the actual current
>>> measured out of A12 is -10 uA (should be 50 uA or greater). There are
>>> two issues here: (1) the actual current disagrees with the meter
>>> indication and (2) the current is low. Is the low current possibly
>>> explained by cell flooding and or incorrect/insufficient 60 MHz RF
>>> signal application?
>>>
>>> Thoughts on next steps? I believe I need to address both issues
>>> above. Addressing cell flooding is straightforward, but will take some
>>> time. The troubleshooting section for A3 says to perform the loop
>>> adjustment procedure if the AGC DC voltage is wrong. Which should I do
>>> first? Or, should I start with something else?
>>>
>>> Thanks!
>>> Matt