Inspecting, aligning, and testing an HP 5065A Rubidium

[rubidium]

Look at these old prices!
http://www.barrytech.com/hewlett-packard/standards/hp5065a.html

Nice old gear site where he listed many old auction prices on multiple units.  I'd love to get some of the 2002-2004 prices he listed today! ;^}

Bill

Ah ... looking in the rear-view mirror can indeed be depressing.  At those prices any one of us would buy them all today.

Hi Jim,

As Corby said, these units are hard to find, in late June I've casually spotted an offer on a national radio amateur forum... was very hard to resist knowing these particular devices, finally decided to contact the owner, declared as operative (!), he has accepted a bit lower offer by me.

When arrived was clear that the unit isn't operative, the output 5 MHz signal was nice and very stable, but the Rb lamp was inoperative, for the rarity of these devices I decided to repair this unit obviously !

I haven't a primary standard and don't want to go for the hard "time-nuts" way, too expensive, at the moment I'm using another Rubidium unit (EFRATOM FRK-H) got some years ago in NOS conditions for only 30 euros, I do not trust the frequency calibration of these units, I will try to compare with a GPSDO (OSA Star-4) in next months.

GOOD Luck to find one !

Stephano

I've learned to be a (mostly) well-disciplined and patient guy when it comes to this stuff - largely as a result of a few failures in the past. About 5 years ago I gambled on a CS4000 that turned out to be little more than a "tire chock" in that it had an exhausted CBT. Of course it was advertised with one of several luring statements like "Working", or "Pulled from a working environment", etc.  Fortunately, I was able to get a refund from the seller. Recently I lucked out on a PRS-50 at low cost after convincing myself that it probably had a bad PSU board and perhaps held the promise that the CBT still had some decent life once the PSU was repaired. Quite honestly, some of the more favorable gambles I've made on surplus high-end gear involved bad power supplies - in particular, failed electrolytics.

[HighPrecision]

Now my 5065A unit is constantly powered ON from 2 weeks, after various adjustments and temperatures checking the output signal at A7TP2 is increased, @137Hz (offset) is 1 Vpp and @274Hz (centered) is 280 mVpp, all very stable!

New update:

Found another problem, in the C-Field coil there is no current flowing despite the presence of a voltage on A15 R10 and R11 current sense resistors, about 2V DC variable with the C-Field control.

The RVFR "C" coil is about 12 Ohms, found the dual transistor A15 Q6 (1854-0221) broken, either B-E junctions are shorted; ordered a NOS original replacement and changed also the C6 tantalum capacitor slightly leaking (47µF).

Now the C-Field adjustemnt acts on output frequency for about 12 milliHertz. The major surprise is the 2nd Harmonic level on the A7TP2 test point is increased to 1.32Vpp @ 274 Hz (centered) and over 4Vpp @ 137 Hz (offset) !

The 2nd Harmonic signal now is clean, noise is greatly reduced.

Stephano

[notfaded1]

The major surprise is the 2nd Harmonic level on the A7TP2 test point is increased to 1.32Vpp @ 274 Hz (centered) and over 4Vpp @ 137 Hz (offset) !

The 2nd Harmonic signal now is clean, noise is greatly reduced.

Stephano

Sounds like transistor fix made a nice difference.  Now I'm wondering if I should buy some 1854-0221 while they can be had still.  WTH I just bought a couple new ones.

Bill

[5065AGuru]

Stephano,

With no C-field current you were probably locking on the wrong line.

One of the lines a Rubidium can lock on at low C-field currents is HIGHLY sensitive to magnetic fields.

The HP 5065A when working properly is designed to provide enough current even at the minimum pot setting to avoid this line.

Great that you are making progress!

Cheers,

Corby

[HighPrecision]

Corby,
thank you for the explanation on C-Field !

Actually I'm trying to calibrate the 5065A output frequency with the only method available for me, using the Oscilloquartz GPSDO Star-4, TC setted to 3600 sec. and using a FA-2 counter statistics to average 10MHz frequency variations, approx. 260 microHertz P-P, after some weeks and adjustment of C-Field, the averaged difference is stable around -4.4 microHertz @ 5 MHz.

I think this small difference is too difficult to compensate with C-Field control.

I have found a replacement connector for mains input, same MIL serie and pin number (3), but different rotation and reversed socket/plug sides, not very safe to have live mains on plug side, but I will avoid to modify the rear panel of the 5065A 

The RVFR signal at A7TP2 is very strong, the A7R29 trimmer for the meter indication is somewhat difficult to set, from full CCW position the meter needle jumps immediately over full scale with a very small rotation of this trimmer in the '2nd Harmonic' position of meter switch.

That is a very interesting experience with this unit !

Best, Stephano

[5065AGuru]

Stephano,

That (the R29 pot you mentioned) is a very good sign.

If the instrument has a strong signal the pot does indeed end up being almost max CCW and hard to set.

Cheers,

Corby

[testpoint1]

today I also have time to fix my current 5065A, just follow Corby's procedure, may need several weeks since I need to replace many components and its battery, I have some idea for this:
1. +20V or -20V output, according its manual, it is +-0.2V tolerance;
2. for the Tantalum capacitor replacement, if the pin/terminal is OK, the testing result seems good, I try to replace them all, but when I purchased some "new" capacitor from Ebay, the testing KPI is worse than old one, but you can use the cemeric capacitor instead of them (100uF 30V using the 220uF 63V instead of it); the attached photo is for capacitor testing;
3. for the Rubidium lamp, it is quite good, I can say, till now,  its technology better than current small Rubidium standard, just like LPRO, SRO-100 or PRS10 etc.

[bnz]

I recently got a 5065A and replaced the suspect caps and the 1k3 resistor. As it immediately locked and had good meter readings I did not do any alignments up to now. I just kept it running like that for that rubidium balance gets established again.
I am however uncertain about the development of the second harmonic reading of the meter, it kept increasing and is now at 50 (according to the manual the expected range is 20-40). Is that good or a problem? I.e. should I stop warming up and start trouble shooting 

[KE5FX]

I recently got a 5065A and replaced the suspect caps and the 1k3 resistor. As it immediately locked and had good meter readings I did not do any alignments up to now. I just kept it running like that for that rubidium balance gets established again.
I am however uncertain about the development of the second harmonic reading of the meter, it kept increasing and is now at 50 (according to the manual the expected range is 20-40). Is that good or a problem? I.e. should I stop warming up and start trouble shooting 

It's normal for the 2H reading to go up over the first several days of operation.  Given the dramatic increase you're seeing, it may have already had some cell flooding the last time someone aligned it. 

The best approach, now that the caps have been replaced, is to let it sit there for a week or two before you make any adjustments. Any tweaking you perform prior to burn-in will likely need to be repeated once things stabilize.

[pquadrat]

There seems to be a russian/ukrainian/soviet era clone of the HP 5065A:

https://www.ebay.com/itm/235042380910?itmmeta=01J2RHPDED6FRZYP97PKYS4MM4&hash=item36b99e1c6e:g:60IAAOSw5Y1dd0FU&itmprp=enc%3AAQAJAAAA8CKDKQNDVCtCDVBDql01YbjfMIvU1%2Btq5xs%2FaRdNi6qNFAWoZ%2BcOuOXXU7AiSaHIstm7okhMSpMZ1R49LA87UphmzeBprV80FrDFGW7MKOwfktdmPkkIpKTEcuStBVgosfp3yYp411ut1VVhmT7K89rqrqFevkq3y6fDWQ0hAeAAV3zi%2FIcb%2BphI1iKxLmpfUw5pOlckuhyP5KifmsSVOBeJsqYjhMQRi6IjU%2B4kogoNWCPlrpagy27jsk3su8VoVLEIP%2FTGt0vF%2B8cJUiBpuF2h366EE%2Fw7TPAXygtFBEpCL%2Ftv8e1fQp1vtZ%2Fh4unGzg%3D%3D%7Ctkp%3ABk9SR6LX2ZGWZA

[KE5FX]

Interesting piece of hardware.  Can't tell from the photos if it is just a repackaged miniature telecom-grade standard, or a full-fledged 5065A-class design.  I asked the seller for more info.  If it's really a custom-engineered product from the ground up, I'll buy one and see how it performs.

[dipsocket]

Interesting piece of hardware.  Can't tell from the photos if it is just a repackaged miniature telecom-grade standard, or a full-fledged 5065A-class design.  I asked the seller for more info.  If it's really a custom-engineered product from the ground up, I'll buy one and see how it performs.

That listing first appeared a few years ago. I found a few things about it:
In the pdf
https://inis.iaea.org/collection/NCLCollectionStore/_Public/23/012/23012324.pdf
it is stated:

The technology for the production of rubidium devices was first reported in the Soviet Union in 1983, some 10 years after it was reported in (West) Germany. Good original work is also reported, and it is clear that the Soviet technology is now well advanced with the 1984 production of Chl-69 devices which are considerable improvements over contemporary HP-5065A and Efratom XSRM devices. The recently developed Chl-74 mobile unit resembles the HP 5065A and it appears to be competitive with any existing rubidium gas cell products available in the West.

There is a picture of the unit that is captioned "SCHV-74" on https://ruknar.com/en/assets/gallery/15/124.png
which is mentioned on page 7(192) in a pdf on your site: http://www.ke5fx.com/vch1006/Demidov_Design_and_industrial_production_of_frequency_standards_masers_in_USSR.pdf

and on https://ilrs.gsfc.nasa.gov/network/stations/inactive/LVIL_sitelog.html with specifications:
  9.01.01 Frequency Standard Type : RUBIDIUM
          Model                   : SChV-74
          Manufacturer            : ex-USSR
          Short Term Stab.  [e-12]: 5
          Long Term Stab.   [e-12]: 10

I also found a very poor scan of what seemed like a page of the user manual in an attachment on the below message board which shows a separate filter cell, but I was not able to determine if the image was from the -69 unit or the -74 unit:
http://www.cqham.ru/forum/showthread.php?13839-%D1%D7%C2-74&s=2e9e5bb2588cf11398029cb4623143c2

  I remember reading another document which mentioned the -74 is a combined-cell design, but I can't find it now.  I did contact the seller and he seemed to be willing to scan the user manual for a reasonable fee, however I did not follow through.

[coppercone2]

I love how they copied the bezel, no shame

wanted that star trek look

[KE5FX]

That listing first appeared a few years ago. I found a few things about it...

Good stuff, thanks.  It appears that most or all of the schematics are present in that forum thread, and that it is indeed an HP 5065A "homage" -- or at least, the one they are discussing in the thread is.

Even if it's an integrated-cell design it should still be interesting to tinker with. I went ahead and ordered one... we'll see what it looks like.  The last time someone randomly dropped a link on here to an auction of unusual Russian time & frequency hardware, it turned out to be very worthwhile.

[pquadrat]

I got a 5065A some time ago, the unit was not used for quite some time, and with unknown working status. But it was nearly for free.

Befor switching it on, I replaced all the tantaliums, most of the did not leak at all, and there was no damage on any of the PCBs. The unit is a late one, serial 2740A01540. It has already the 10811 crystal oscillator.

Then I switched it on (yes, the special power cable came with the unit), and let it warm up for 2 hours.

I could not get it to CO, so I did the cell flooding procedure from the manual for a week, Nothing changed.


Next I  checked the lamp assembly, and the 1.33KOhm resistor was open circuit. After replacing it with a 1,3kOhm, feeding it with 20V, the lamp started after some seconds. and the assembly took 137-140mA. I did polish the mirror.

But still nothing changed.

Readings were now:
Supply 40
Lamp Oven 38
Cell Oven 40
Osc Oven 36
Photo I 14  (was the same with the non-working lamp)
5MHz 16
Control 0
Error 0
2nd Harmonic 6
1 MHz 38
100KHz 40

But I do not see anything on A7 TP2, just very little noise. and it does not go into continuous operation.

Assuming it did work until the 1,33KOhn resistor went open, I think it should run.

Any suggestions?

Thanks,
Peter
from Hamburg, Germany

[pquadrat]

It is an overall very clean unit.

[pquadrat]

It's working now! Two wires at the bottom of A7 were swapped. Nothing destroyed.

But it is running 4mHz below my GPSDO (HP Symmetricom Z3801A). Is this normal? Counter is a HP 5370B.
And Second Harmonic is above 50.

[KE5FX]

2nd harmonic > 50 is too high.  Sounds like you haven't gone through the alignment process, as it's normally between 20 and 40 IIRC.  I would try that after letting it run for an hour or two, then plan to repeat the alignment process if needed after it's been operating for a few weeks.

4 mHz at 10 MHz is a fractional error of 4E-10, which is >10x more than you'd like to see.  If no alarms are active on the GPSDO, the 5065A's C field may be (rather severely) misadjusted. I wouldn't adjust that until getting the 2nd harmonic in range, though.  Suggest following the alignment process in the manual but for now, skip any adjustments that are hard to reach, such as the ones on the side of the A3 assembly.  They are unlikely to be a real problem. 

You might also look at the simplified procedure attached, from notes by Corby Dawson.  (I think it's the latest version, hopefully he will correct me if not.)

[pquadrat]

Now i did the alignment according to Corby's procedure, and all is fine.

I have about 2Vpp at A3 TP2, and all readings of the meter are fine. For the freqency to match the GPSDO, I had to increase the C-field.
Readings on the HP 5370B jump arount +-0.2mHz maximum, which should be normal.

Now I will watch stability. The OCXO (HP 10811) itself aged at only 0.2 to 0,3 mHz per day, which is well in spec.

[5065AGuru]

The 2nd harmonic reading is adjustable and has no effect on operation unless you set it to low, and then the lock light circuit will not work.
Basically set it to around 40 and then it is an indicator of the units health. On a good unit it should stay stable for a long time.
Glad you got it running!

Cheers,

Corby

[pquadrat]

Thanks KE5FX and Corby!

It is now running for more than two weeks, and it is very stable, did not drift anything I could detect with my GPS and HP 5070B.

But I really do not need it. Is anyone interested? What is it worth? I have no clue.

-Peter

[KE5FX]

It's worth quite a bit.  I think you'd see at least USD $3000 to $5000 with a good ad on eBay.  That's the best non-primary frequency standard you can buy for less than six figures.

[pquadrat]

Thanks, I will put it on ebay soon.