looking for HP3456A (6.5 DMM) calibration service in Italy or continental Europe

• looking for HP3456A (6.5 DMM) calibration service in Italy or continental Europe
  Posted by gmac34 on 24 Sep, 2023 20:27
• I recently got a HP 3456A bench meter to repair, unfortunately fresh capacitors only cured half of its problems and it turns out I need to replace one of the fine-line resistor networks.
  So it will definitely need to be calibrated.
  I'm looking for recommendation on "hobbyist friendly" cal labs in Italy, I do not need traceability, just to get it sorted, and I'm on a student budget...
  I would preferably bring the item in person, so driving distance from Bologna would be best, but recommendations In Europe are also welcome.
  Thanks
  Grazie
• #1 Reply
  Posted by alm on 24 Sep, 2023 21:20
• Do you need just calibration (a report of how much it is off) or also adjustment (twiddle trimmers until it reads all zero?). The latter is a manual adjustment, so cal labs will hate it and likely charge extra compared to more modern meters like the HP 34401A where they just press a button and the whole thing gets adjusted via computer control.
  
  If you have another 6.5 digit or better DMM or know someone who does, and can improvise sources that are sufficiently stable short term, then I would adjust the 3456A as well as possible to the other meter, and then send it out for calibration without adjustment. DCV and resistance are relatively easy. ACV is much harder to generate all the voltages and frequencies it wants. Depending on what functions are affected by that resistor network, you may not have to adjust everything.
• #2 Reply
  Posted by Vgkid on 24 Sep, 2023 22:31
• If the U200 resistor network is replaced , then all ranges(voltage/resistance) will need to be fully calibrated. That will even go down to selecting the proper select on test link resistors  . I think it would be better to use another working 6.5dig meter.
• #3 Reply
  Posted by gmac34 on 25 Sep, 2023 06:06
• I assumed it would need a full adjustment but the manual is not too clear on what happens after replacing U200. There are mentions of what to adjust if the other fineline (U500) is replaced (that one is matched to the voltage Reference Assembly), and it that case there are a bunch ion jumpers to move around, but I cannot find a procedure to replace U200.
  On the manual I find:
  "Thin-film resistors (U200 fineline resistor package) with .010% tolerance are used in the feedback paths for the gains of 100, 10, and 2. Gain of 2 is used for Ohms measurements only."
  Gain of 100 and 10 are used in all ranges apart from 10V and 1000V
  
  To do a DIY adjustment I have a 3245A, that would cover most voltages needed for the adjustment, apart form 100v dc.
  For resistance I don't have much that is stable, maybe I can use the faulty U200...
  But I only have a keithley 2000 as possible reference.
  
  Off course the alternative is to keep the meter as is and accept that it cannot measure 100V and 1000V (it's the 100:1 voltage divider in U200 that seems faulty)
• #4 Reply
  Posted by NoisyBoy on 25 Sep, 2023 22:32
• Unfortunately, I cannot help you with your original question of a cal lab recommendation in Italy.  So hopefully someone local or in the EU can help.
  
  I see the direction has moved towards a self-cal, it all depends on what you are using the DMM for, if you get it for the purpose of doing accurate 6.5 digit measurements, then any home cal or using another DMM as a transfer standard is not going to be good enough for such meter.  Challenge being the difficulty in having the stable sources at various test points that you need, in addition to a transfer standard with recent calibration.  You really need a calibrator for such work. 
  
  However, if accuracy is not needed, or you are just having the 3456A as a TE collection, then a self-cal may be adequate so you don't have to ship the vintage equipment around. 
  
  
• #5 Reply
  Posted by gmac34 on 26 Sep, 2023 07:14
• Hi, I'm quite aware that calibrating a 6.5 digits dmm with another 6.5digits dmm is quite meaningless.
  And it’s is good to make it clear before the metrology experts start shouting at me.
  
  Last time I asked for a quote to a cal lab, when I was back in Norway, they asked around 500€ to do a keithley 2000, something much easier to adjust that the 3456A, that would be outside my budget…
  
  The reasoning for the self cal is that I should anyway get the instrument in a “usable” state to make sure it does work properly before investing in a proper adjustment (if ever that occasion comes up).
  
  At the moment the 3456A is not a functioning instrument, the DCV cal seems ok as it agrees very well with the  keithley 2000, but ACV and resistance are a bit wonky and anyway when I got the unit the seals were broken so, there is not much to preserve.
  
  As per calibration sources, wouldn’t the 3245A be a good source (up to 10V)? It is a lm399 based instrument, short therm stability should be quite good.
  
  Anyway, if Vgkid is correct I might not able to repair this unit at all, still cannot find references on the service manual on how to replace U200, which is odd given that many troubleshooting procedures end in “replace U200”
• #6 Reply
  Posted by Kleinstein on 26 Sep, 2023 07:58
• I don't see very special adjustment procedures needed for the adjustement after changing U200. It should be relatively straight forward. I would start with the voltage ranges and only after this adjust the resistance ranges. This second step may need to set / cut links for the coarse part. The jumpers suggest that the scattering between U200 samples can be quite large.  As a test source the 3245 should be OK with good enough short time stability and noise, at least for a first test.
  The procedure should more or less follow the normal calibration / adjustment, except for the links that are normally not changed that often.
  
  It absolutely makes sense to do a DIY adjustment first and than also use and test the meter for some time to give it a burn in, before eventually getting a real calibration. Adjustment should be faster when starting from a relatively good starting point (e.g. no need to change the links).  A new U200, even if NOS or claimed from a even more broken meter may shown some initial settling / drift.
  
  If there is still something bad with the resistance, one may want to find that error (could be as little as dirt or a bad connector) before a real calibration anyway.
  The AC part should not be effected by U200, but the former owner may have touched it too and the damaging even may have caused more damage than just U200.
• #7 Reply
  Posted by gmac34 on 26 Sep, 2023 08:16
• Quote from: Kleinstein on 26 Sep, 2023 07:58

  I don't see very special adjustment procedures needed for the adjustement after changing U200. It should be relatively straight forward. I would start with the voltage ranges and only after this adjust the resistance ranges. This second step may need to set / cut links for the coarse part. The jumpers suggest that the scattering between U200 samples can be quite large.  As a test source the 3245 should be OK with good enough short time stability and noise, at least for a first test.
  The procedure should more or less follow the normal calibration / adjustment, except for the links that are normally not changed that often.
  
  It absolutely makes sense to do a DIY adjustment first and than also use and test the meter for some time to give it a burn in, before eventually getting a real calibration. Adjustment should be faster when starting from a relatively good starting point (e.g. no need to change the links).  A new U200, even if NOS or claimed from a even more broken meter may shown some initial settling / drift.
  
  If there is still something bad with the resistance, one may want to find that error (could be as little as dirt or a bad connector) before a real calibration anyway.
  The AC part should not be effected by U200, but the former owner may have touched it too and the damaging even may have caused more damage than just U200.
  

  
  If you know it, could you point me to the section of the SM that talks about cutting links? I could not find it (I only find the section about the jumpers to move to match U500 to the reference board)
  Thanks
• #8 Reply
  Posted by Kleinstein on 26 Sep, 2023 09:22
• I have just seen the jumpers JMPR201-JMPR204 in the schematics. They are for the "coarse" (e.g. 0.2 %) setting for all the current source.
  With a wrong setting there the adjustment range of the trimmers for the resistance ranges may not be large enough.
  
  I would first keep them as is and try the adjustment of the resistance ranges likely start with the 10 K range. If one than runs out of adjustment range with the trimmers one can use the jumpers to shift the whole range. Probably a try and error for the direction and than some guessing on the size.
• #9 Reply
  Posted by alm on 26 Sep, 2023 16:02
• Quote from: NoisyBoy on 25 Sep, 2023 22:32

  I see the direction has moved towards a self-cal, it all depends on what you are using the DMM for, if you get it for the purpose of doing accurate 6.5 digit measurements, then any home cal or using another DMM as a transfer standard is not going to be good enough for such meter.  Challenge being the difficulty in having the stable sources at various test points that you need, in addition to a transfer standard with recent calibration.  You really need a calibrator for such work.
  

  From what I read it was more about self-adjustment, because most cal labs would not like to change solder links or do a lot of trimmer twiddling. No one is saying you couldn't send it to a cal lab afterwards for traceable data. You do know that calibration in general does not include adjustment, right? Particularly on instruments where adjustment is more than sending commands from a computer.
  
• #10 Reply
  Posted by gmac34 on 28 Sep, 2023 17:05
• Ok, maybe there is some confusion with adjustment and calibration, I think in Italian "calibrazione" is fiddling with pots and "taratura" is comparing with a known standard. Sorry
  Anyway what I was looking is both, if needed.
  I received the replacement fine-line, measured to see if it was ok, installed, and the issue was still there, I measured the original, it seems good to, so I put it back and the fault was gone. there might have been some oxidation on the golden fingers. So I cleaned everything and reinstalled the original.
  
  I guess it's fixed now and didn't need a new U200 after all.
  
  I will not touch the DCV calibration, it is quite consistently less than 5-6 ppms away from my other meter.
  
  ACV and resistance might still require some fiddling, but I need some better resistors to do something about that