Author Topic: HP 6920B Meter Calibrator  (Read 15482 times)

0 Members and 1 Guest are viewing this topic.

Offline gilbenl

  • Regular Contributor
  • *
  • Posts: 170
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #25 on: September 07, 2016, 02:11:56 pm »
Holy thread resurrection, Batman!!  I dug deep for this one.   :D

I just bought one of these on the 'bay today, so I'll be joining the club too.  It should be here in a week or so, what with the holiday on Monday. I just looked at the slide show of your restoration, and it appears to me that you did a fine job of it.  I don't think I'll go quite so far as to reshuffle caps on this, but we'll see when it arrives.

Did you ever get around to tweaking and increasing the accuracy?  I will likely be picking your brain for info once I get into working on it.

- Pat

Replace the caps. You don't have to go for the nostalgic approach of the restuff, but those old spragues cannot be good--even if the meter seems to function, they're not good.

As far as tweaking the accuracy, I ultimately decided it wouldn't be worth the effort. You'd have to replace pretty much the whole unit.
What doesn't kill you, probably hurts a lot.
 

Online Cubdriver

  • Supporter
  • ****
  • Posts: 2829
  • Country: us
  • Nixie addict
    • Photos of electronic gear
Re: HP 6920B Meter Calibrator
« Reply #26 on: September 07, 2016, 04:38:26 pm »
Holy thread resurrection, Batman!!  I dug deep for this one.   :D

I just bought one of these on the 'bay today, so I'll be joining the club too.  It should be here in a week or so, what with the holiday on Monday. I just looked at the slide show of your restoration, and it appears to me that you did a fine job of it.  I don't think I'll go quite so far as to reshuffle caps on this, but we'll see when it arrives.

Did you ever get around to tweaking and increasing the accuracy?  I will likely be picking your brain for info once I get into working on it.

- Pat

Replace the caps. You don't have to go for the nostalgic approach of the restuff, but those old spragues cannot be good--even if the meter seems to function, they're not good.

As far as tweaking the accuracy, I ultimately decided it wouldn't be worth the effort. You'd have to replace pretty much the whole unit.

Dachh!!  I just re-read my post in your quote - that was supposed to be restuff, not reshuffle.  I typed that on the ipad, and between touch screens not liking my fingers and autocorrect there are occasionally some odd results.  I usually catch and correct them.   :-[

I managed to snag an original manual for it as well; both that and the unit itself are currently enroute.  I figure it's a safe bet that most if not all of the electrolytics will likely be toast by now, and will probably be shotgunned when I go through it.

-Pat
If it jams, force it.  If it breaks, you needed a new one anyway...
 

Offline MudMan

  • Contributor
  • Posts: 36
  • Country: de
Re: HP 6920B Meter Calibrator
« Reply #27 on: September 07, 2016, 06:14:54 pm »
I got a 6920B a few days ago from ebay. All caps are original and I've checked all of them, they were good. Also the unit is working fine, so why recap?
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 8460
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: HP 6920B Meter Calibrator
« Reply #28 on: September 08, 2016, 01:35:40 am »
Gee, you guys buy everything I look at on eBay. :-DD

Nice scores, everyone. Thanks for resurrecting the thread. I hadn't gotten around to looking for prior teardowns and was curious what this calibrator was like.
I TEA.
 

Online Cubdriver

  • Supporter
  • ****
  • Posts: 2829
  • Country: us
  • Nixie addict
    • Photos of electronic gear
Re: HP 6920B Meter Calibrator
« Reply #29 on: September 08, 2016, 05:56:07 am »
Gee, you guys buy everything I look at on eBay. :-DD

Nice scores, everyone. Thanks for resurrecting the thread. I hadn't gotten around to looking for prior teardowns and was curious what this calibrator was like.

 :-DD  I'm always afraid that I'm going to wind up bidding against someone here.  Funny that several of is have similar interests and are eyeballing the same things. 

I'd had a search for a 6920B in my feed for a long time, but never seemed to find one for less than a small fortune.  This one popped up as a BIN-OBO, and I figured it would have been gone by the time I looked - didn't check email till late afternoon that day.  Based on the rarity of calibrators in that price range, I decided to simply jump on the BIN price rather than making an offer and possibly missing out on it to someone else.  I'll know perhaps as soon as tomorrow if I've bought a pig in a poke - it's departed Keasby, NJ and should be arriving shortly.

-Pat
If it jams, force it.  If it breaks, you needed a new one anyway...
 

Offline gilbenl

  • Regular Contributor
  • *
  • Posts: 170
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #30 on: September 22, 2016, 01:43:17 pm »
I got a 6920B a few days ago from ebay. All caps are original and I've checked all of them, they were good. Also the unit is working fine, so why recap?

The capacitance may be reasonable, but the ESR will certainly be very high. The ESR, tested out of circuit, will be high. Even if they did happen to pass, they'll change rapidly with use. I've heard time and time again that if you see old sprague caps, always replace--don't even test.
What doesn't kill you, probably hurts a lot.
 

Offline cbarbos2

  • Newbie
  • Posts: 3
  • Country: br
Re: HP 6920B Meter Calibrator
« Reply #31 on: November 08, 2016, 12:05:22 pm »
Hello, guys.
I am new in this forum but I would like to receive any help with the HP 6920B.
I have one that the DC option is working fine but in the AC mode I have a distorced signal at the output.
I have the service manual and I have doubt about how the demodulator works and how the signal is in the series regulator. The manual said that is a positive full retificated wave.
All of transistors are good but they are not original ones.

Thanks for any help,

Carlos
 

Offline PA4TIM

  • Super Contributor
  • ***
  • Posts: 1123
  • Country: nl
  • instruments are like rabbits, they multiply fast
    • PA4TIMs shelter for orphan measurement stuff
Re: HP 6920B Meter Calibrator
« Reply #32 on: November 08, 2016, 07:59:18 pm »
I got a 6920B a few days ago from ebay. All caps are original and I've checked all of them, they were good. Also the unit is working fine, so why recap?

The capacitance may be reasonable, but the ESR will certainly be very high. The ESR, tested out of circuit, will be high. Even if they did happen to pass, they'll change rapidly with use. I've heard time and time again that if you see old sprague caps, always replace--don't even test.

I have not heard it ever  before but  I measured a lot of Spraques and often they are better as the new replacements. And for replacing them because the ESR measurements. Let me guess,  at 100 kHz ?  Show me the ESR value in the datasheets, Oh, you can not find that value ?  But how you then know they are that bad  >:D
www.pa4tim.nl my collection measurement gear and experiments Also lots of info about network analyse
www.schneiderelectronicsrepair.nl  repair of test and calibration equipment
https://www.youtube.com/user/pa4tim my youtube channel
 

Offline Vacuuminded

  • Contributor
  • Posts: 15
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #33 on: September 05, 2017, 02:48:11 am »
Hi Gilbenl (and Arkham), I'm certainly glad to see I'm not the only one nutz enough to rebuild one of these!  Found mine at a hamfest several years ago, determined it had some issues but was mostly functional, and there it sat.  Finally tearing back into it now, and in the process of recapping.  Some were definitely bad, others check out to be perfectly fine after reforming.  Still replacing all with good quality electrolytics in the name of preventative maintenance because I plan on using this instrument for many years! 

Kind of stuck on that 4400uF 2V cap myself..  I am of the old mindset that aluminum electrolytics should be run at around 80% of their rated voltage, for maximum life yet still remaining mostly, or fully "formed".  Just can't bring myself to put a 4700uF 35V electrolytic in place of this 2V unit, for I firmly believe it will fail prematurely.  My thought was to replace it with a few amply sized solid tantalum caps in parallel, equaling the 4400uF or so of the original, voltage rating not important so long as it's over 2V.  But...  With that inductor in series, I wonder if C or ESR is in any way critical here to prevent oscillation with load variation??  I assume they used those big germanium transistors as diodes due to the lower Vfd, but beyond that am far from understanding several of the circuits in this instrument..  If any of you can lend any insight here, or speak as to how your rebuilt units perform, I would certainly appreciate it! 

Also just a question..  How clean is the AC and DC output waveforms from your units?  Prior to recapping and part way through, I still have a decent spike visible, and a fairly distorted AC waveform.

Regards,
Chris
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #34 on: September 14, 2018, 07:15:59 pm »
Seems like this thread gets hit about once a year.   >:D  I have found a 6920B in good condition for what I think is reasonable $$, and probably will be bringing it home soon.  I just wanted to thank everyone who has contributed to the thread so far, and hope to be able to share some data points soon.  I should probably go ahead and pre-order the caps, right?   :-DD

WRT the choice of capacitors: I can't imagine an electrolytic of any kind being used in a circuit where its parameters are critical to proper circuit operation.  When we were designing with them, the important thing was to be sure the voltage rating and ESR are appropriate; unless you hand-sort them, capacitance is bound to have a large std. dev. sample to sample, and since large values are primarily used for coupling or filtering, I would never find a high ESR acceptable.   That's just my take on it - but then I worked mostly in analog and linear power supply design, so there may be edge cases.
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 8460
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: HP 6920B Meter Calibrator
« Reply #35 on: September 14, 2018, 11:56:51 pm »
WRT the choice of capacitors: I can't imagine an electrolytic of any kind being used in a circuit where its parameters are critical to proper circuit operation.

As with many things, "it depends." The ESR of electrolytics have been used to eliminate oscillations in circuits where other capacitor types, including low-ESR electrolytics, would exacerbate the problem. So, in general, be sure the voltage and ESR are appropriate replacements, as you said. The capacitance value isn't as critical (as can be seen in the tolerance of most general-purpose electrolytic caps). In some applications, using too high a capacitance can introduce problems with in-rush current. A little higher than original should be just fine and may be necessary with old equipment when you can't get the same value.
I TEA.
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #36 on: September 15, 2018, 03:06:35 am »
I get that.  To be clear, I was talking about precise values, which you're just not going to get with an aluminum electrolytic.   ;) 

The large value Spragues which apparently are the issue with this unit appear to all be used for basic low pass filtering, where low ESR is probably desirable.  If I find out differently on specific caps, I'll be more careful of the replacement characteristics!  I don't expect they will be anything special, though, because in looking at the BOM, the unique parts tend to not have nominal values listed (hand-selected, presumably).
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 8460
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: HP 6920B Meter Calibrator
« Reply #37 on: September 15, 2018, 04:37:45 pm »
Sounds good. I look forward to seeing your calibrator!
I TEA.
 
The following users thanked this post: GregDunn

Offline factory

  • Frequent Contributor
  • **
  • Posts: 363
  • Country: gb
Re: HP 6920B Meter Calibrator
« Reply #38 on: September 21, 2018, 10:19:26 pm »

I'm unable to measure the output at L1 given the current state of the unit. Correction, there's a math opportunity here...
Max current output is 10A
Inductor is 100uH
Ton=1ms (no clue, parts list doesn't specify the inductor)
Ipk=10A ("maximum" output of unit)


Note the manual & front panel specifies the maximum output on the 10A ranges is only 5A.

I've only changed one capacitor in mine as the seal on the end was cracked too, it was C10 (1200uF 20V) and was replaced with a higher 40V rated Kemet PEG225 axial type. I also had to make a replacement for the missing output indicator, it could do with being a bit brighter but it does work.

David
« Last Edit: September 21, 2018, 10:28:22 pm by factory »
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #39 on: September 22, 2018, 07:30:21 pm »
OK, here we are.  It's a bit worn on the front, meaning it's apparently been used a lot.  Everywhere fingers could have touched, or stuff could have rubbed against it, it probably did.   :o  The rest of the unit is surprisingly free from scrapes, nicks, etc.  though, and all the controls seem to work well.  One thing which caught my interest is the sticker which indicates it spent the latter part of its life at NASA Lewis Research Center (now NASA Glenn).  Someone bolted a sturdy handle on the top panel, so I guess it was the roving metrology department for the Center's lab benches.  Or maybe they used it for the big things like the wind tunnel data acquisition bays which couldn't easily be disassembled and carried to the met lab. ;D

It has the standard IEC power inlet on the back, which I take it means it's a later model (sn116A02301, not marked "Harrison Labs") and the fuse holder is cracked and barely holding together (I managed a further $25 off the already low sticker price).  I took a look online when I was thinking about buying it, and saw them going for 3-4x what this guy was asking, so I took the risk.

It does power up and emits a noticeable hum, but the voltage output works in AC and DC mode on the ranges up to 100V - The dial changes voltage smoothly and precisely, albeit it's noticeably out of cal on the low side.  That, plus the hum, speaks to me of bad power supply caps and maybe other things yet unseen.  It is going to need a thorough cleaning; the top board has a layer of ancient greasy dust on it and the vinyl coverings are tacky.  I'm thinking about removing the handle too; the top cover is not designed to carry the entire weight of the meter and has already bowed from the stress.  And of course installing a proper screw-on fuse holder.

I've got a couple of pieces of gear in the queue at the moment, so I'm not sure when this one will get to the front, but I think I'm just going to add the caps to my upcoming parts order so I won't need to panic-order any replacements when I do pull the boards out.

 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 8460
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: HP 6920B Meter Calibrator
« Reply #40 on: September 22, 2018, 09:32:14 pm »
Cool! Those panels often end up sticky over time. They'll clean up nicely, though.
I TEA.
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #41 on: September 22, 2018, 09:54:08 pm »
I just wish there was a simple way to replace all the paint worn off the front panel without re-doing all the lettering too.   :(  Oh well, if I wanted a pretty one I could have spent $400, right?   :-DD
 

Offline bitseeker

  • Super Contributor
  • ***
  • Posts: 8460
  • Country: us
  • Lots of engineer-tweakable parts inside!
Re: HP 6920B Meter Calibrator
« Reply #42 on: September 23, 2018, 04:39:47 am »
Right! ;D
I TEA.
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #43 on: October 01, 2018, 09:10:57 pm »
Phase 1 repair done.  Fuse holder replaced, and the Sprague electrolytics changed out to new 105C rated parts.  (Except for one - I haven't found a 5µF 150V that will fit in the space of the old one)

The power-on hum is greatly reduced, though not absent; it's likely coming from the power transformer so maybe the updated circuitry is better behaved due to lower leakage currents.  Interestingly, the new caps took a few moments to form, and the voltage was pretty unstable on several of the ranges until they settled.  I know some of them are on the outputs (there are separate secondaries and circuits for the AC, DC and current modes), so they don't automatically have voltage applied till you choose that range and mode.  The good news is that all the outputs now illuminate the output lamp, whereas before it wasn't behaving on some of them.

The voltages aren't terribly close to the dial setting, and by different amounts on different modes/ranges.  I think after letting the unit heat up for a while (and turning on the outputs for each range several times) I'm going to check the reference and supply voltages with my most accurate meters to see if there's anything out of spec there.  If all looks good, I suppose I'll have to think about how I really want to calibrate this unit.  If they are made at all, it needs separate adjustments on each range and I certainly don't have voltage and current standards which will work on more than one range.  I guess since its accuracy spec is significantly worse than my Keithley and Fluke bench meters, I could use them as rough checks, but I know that's just a circular reference and not really accurate.
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #44 on: October 01, 2018, 11:07:03 pm »
Aaaaand... the problem deepens.  For one thing, the manuals are very coy about what the actual "ground" reference is on this unit.  It has multiple secondaries on the power transformer, and there is no "earth ground" shown on the schematic - just a chassis ground which is further isolated by the output transformer.  The closest thing to a ground is the center tap for the transformer feeding the series regulator and I assume that's what voltages are referred to.  It's also called "-S" on the reference supply (is this terminology that I should know?).  If this is the case, then the reference PSU has some problems and there are a couple of caps being run very close to their rated voltage.  There may be another component(s) damaged.  I guess it's time to throw it on the variac and do more testing.

Also, I did the first troubleshooting steps which involve measuring resistances at the output terminals, and those are way off.  I think cleaning the rotary switch contacts is indicated as soon as I get the power supply voltages sorted.

Edit: the AC reference is not well.  I would expect the two main filter caps (C16-17) should have approximately the same voltage on them, but they don't - C16 has about 53VDC on it while C17 measures about 38V (50V rated caps).  Looking at the AC reference's bridge rectifiers (CR24-27) on a scope seems to indicate that they're all OK so the voltage across the bridge's output ought to be balanced and it's not even close. 

Line voltage is ~120V.  I wish more of the circuit voltages were marked on the schematic.  The DC reference PSU voltages are all a few tenths of a volt out of spec high but the voltage across C13 is reasonable (38VDC).  I need to measure a few more of the internal voltages to see if I can figure out where this thing is going awry.
 
« Last Edit: October 02, 2018, 06:03:53 am by GregDunn »
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #45 on: October 04, 2018, 08:22:48 pm »
More details...  Again, assuming that "-S" is the reference ground for the circuitry, it looks like the 15VDC to the optoisolator is OK, and so are the reference DC voltages.  All are a bit out of spec (0.2VDC or so), but they're not grossly wrong.  I decided to use differential mode on my scope as well as a couple of portable DMMs to check some of the internal voltages; the 15VAC and 1VAC references also OK.  The 15VAC is a little high but it generates a 1VAC output that is very close to correct.  My suspicion is that earlier my probe was not making good contact, because when I moved it slightly I got very different readings.  I pulled the tip off and re-seated it, and all seems to be well now.

The only thing which looks really wrong is the rectified DC for the AC reference stage power.  One of the caps (C16) is being subjected to 53VDC, that's certain.  Since the internal voltages are not in error by much, I'm steadily being guided toward the conclusion that either the cap is under-rated or something downstream is not drawing enough current to pull the voltage down - and both seem unlikely given that the outputs from the stage are reasonable.  The actual voltage at that winding's CT  is only about 2VDC different from the "-S" reference and there's no direct connection between them.

Still scratching my head to figure out what's next.
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #46 on: October 12, 2018, 06:35:04 am »
Just wanted to give an update: I temporarily bodged in an 80V cap in place of the 50V C14 so I didn't have to worry about it blowing up while I did further troubleshooting.  C14 (or actually its replacement) still reads about 51-52VDC, and as before the ±6.2VDC, +12.6VDC and 15VDC test points are pretty close.  The 1VDC and 1VAC reference voltages also look good. 

Now, the surprise.  It looks like the output switch S3 has some oxidation on the internal contacts.  Since I could leave it powered on now, I started checking outputs from the calibrator but I was intermittently seeing odd variance of the outputs - not only in AC mode but also in DC mode - when I pressed the output switch.  Cycling the switch or pressing it harder usually gave me a good output.  I exercised the switch a few dozen cycles back and forth, and... I'm now getting very accurate outputs on AC and DC, volts and amps on every range I tested (I didn't try the full 10A range; it's current-limited anyway).  I had already cleaned the power/mode switch S1 and the range switch S2 contacts, but since the guts of S3 are not accessible to the world  I may think about trying to get some de-oxit inside.  At any rate, if I'm careful to make sure the switch is properly closed, the calibrator seems to work properly and gives impressively accurate values.  Pretty sure the real problem was the error amplifier signal not being routed properly through the contacts.

Still not sure about what to do with the fact that C14 is seeing too high a voltage.  Since the unit is well within spec now and all the listed internal voltages are very close to what they should be, I'm not sure if there's anything I can or should do.   :-/O   :-DMM
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #47 on: October 14, 2018, 10:41:41 pm »
Hopefully no more "troubleshooting" updates, but I note that the DC output of this unit seems to have a significant amount of AC present - around 40 mVpp which causes the apparent DC to seem unstable.  I postulate that the average level changes periodically due to the interaction of the DMM sampling with the ripple waveform (obviously they're not synced).  Anyway, this causes a periodic drift of the apparent DC value on the order of several mV even though the actual average level is pretty steady.  It still stays within that 0.2-0.5% spec, but the fact that it wanders makes it hard to use it to compare with multiple DMMs which record sub-millivolt changes pretty well.  AC doesn't really have a "ripple" problem, but it also is pretty unstable in the lower digits - still within spec, but again, hard to use for comparison.  And it looks more like a trapezoid than a sine wave.

The caps on the output of this device are brand new and is good shape; every voltage range seems to have a similar behavior, and of course they all use separate secondaries / bridges for the DC outputs.

a) Is this the way the DC output looks to others who have one of these?

b) I know there are compromises, but why can't I put some more filtering on the output to reduce the ripple even further as long as I allow for the differing impedances and of course surge currents?  Might work for AC, too, since it's all 60 Hz.

It would be nice to have a more pure DC waveform on the output if this thing is to fulfill its primary purpose of being a reference.   >:D
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 18098
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: HP 6920B Meter Calibrator
« Reply #48 on: October 15, 2018, 12:02:24 am »
Just check you don't have a faulty diode in one on the PSU rails, it's not uncommon and then your smoothing caps are dealing with higher ripple currents @ 60 Hz instead of 120 Hz.
Avid Rabid Hobbyist
 

Offline GregDunn

  • Frequent Contributor
  • **
  • Posts: 728
  • Country: us
Re: HP 6920B Meter Calibrator
« Reply #49 on: October 15, 2018, 01:15:36 am »
All those bad boys are on the bottom board, so I'll have to rotate the top board out of the way.  Good project for tomorrow!
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf