EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: nedrysoft on March 16, 2023, 02:29:40 am
-
Hi,
I have a 34401A with a fault and I haven’t found any reference to people also having the same fault, so I figured I’d ask here.
Essentially what I am seeing is that after some amount of time while measuring a 5mV signal the meter stops reading real results and start throwing up random readings.
This happens regardless of whether the front or rear inputs are selected.
If I pull the ground probe off and reattach the fault appears to go away for a little while.
I’m trying to figure out whether other functions are affected, i.e whether this manifests in say resistance or current measuring modes.
I realise this is currently a hit of a vague description, but I thought I would ask here while I figure out whether anything else is affected in case this is a fault that somebody else has seen and remedied themselves.
Thanks.
-
can you run the sleft test see some youtube video about it
https://www.youtube.com/watch?v=cPUt9p-pdZQ (https://www.youtube.com/watch?v=cPUt9p-pdZQ)
-
Self test passes even when starting it while the meter is in this fault mode.
I’ve never seen any logged error codes either.
I’m currently running a 4W resistance check, but will run a V range test after.
-
It seems like some of these 34401A units have a mind of their own. With mine, if I leave it set to measure DCV (leads shorted) and come back an hour or two later, it sometimes has switched itself to ACV. It has never thrown any kind of error, nor has mine ever made an inaccurate measurement. Nobody (including me) seems to have the solution in my case. Perhaps yours is just another manifestation of the 34401 going haywire for no good reason.
-
Check the front/rear switch for damage or oxidation/ contamination.
-
I haven't looked at the schematic, are there optocouplers used between earth referenced and "floating" sides?
-
While it certainly could be an analog problem, the fact that it resets when you reset the probe means it also might be a digital problem.
I recently repaired an HPAK 3648a power supply that would randomly go to different modes and had a drifting readback. Turned out to be a bad eprom that was intermittently providing bad program data to the MPU.
-
I have a somewhat similar problem. On one of my 34401As, when taking a DCV measurement (typically 10V, but I've no reason to assume that it is particularly limited to this range), it will sometimes read garbage i.e. random voltages, however there is no random mode switching and I don't recall any autorange 'hunting'. In my case it usually seems to happen when you first switch to DCV or at initial power up. Forcing manual range and cycling through the ranges doesn't seem to help, nor switching from front to back inputs, but flicking to resistance mode and back to DCV does seem to reliably clear the problem. My current suspect is a bad relay so I have purchased replacements but haven't got around to fitting them yet. The problem is also somewhat intermittent so it may take a while after I get around to changing the relay(s) to confirm the fix.
-
With mine, if I leave it set to measure DCV (leads shorted) and come back an hour or two later, it sometimes has switched itself to ACV.
dismantle front panel and take apart the push-button rubber switch assembly. clean out crud (that is causing a false 'button press') and then reassemble.
cheers,
rob :-)
-
I've done the front panel. There was a whole thread about this arbitrary switching a while back. Despite all of the suggestions back then, I never got to the bottom of it.
Anyway, I don't want to hijack the current thread which is about a different 34401A problem.
-
Just to exclude some 'impossible' scenarios, could that be a probe fault ? I.e. with a short instead of a probe, does this behavior repeat ?
-
Many thanks for all the replies.
I am definitely thinking it’s related to the switch, I wondered if this was the cause of my issues earlier on and I managed to press it slightly (not enough for it to click, barely any pressure), the switch returned to its original position but it did start showing incorrect readings after this.
I will reply properly, I really appreciate the replies here and I will go through them all once I am back home and have the meter in front of me.
-
Many thanks for all the replies.
I am definitely thinking it’s related to the switch, I wondered if this was the cause of my issues earlier on and I managed to press it slightly (not enough for it to click, barely any pressure), the switch returned to its original position but it did start showing incorrect readings after this.
I will reply properly, I really appreciate the replies here and I will go through them all once I am back home and have the meter in front of me.
A few years ago, we discussed problems with the fron/rear switch, how to identify and how to repair (clean) it. We encountered most problems on 2W Ohm mode. Please search for that thread.
Please actuate the switch several times, and use the rear jacks for testing.
HP Replacement parts should not be available any more, so it would be possible to de-solder the whole switch, carefully disassemble it - attention many small contact springs inside - clean the innards with IPA from outgassed plasticizer and re-assemble the whole switch.
I've done that already on my 30 y/o unit.
Frank
-
In case anyone wants a NOS switch, there's at least one on the auction site right now. I bought one for my stock just in case I get a meter with a physically broken one. Which.... HAS happened.
-
So, armed with the schematic I was able to pick a few points where I could take a look at the DC bout signal to work out where it was going wrong.
I traced a 6mV input signal through the switch and into K101, where on one contact of the relay I could see 6mV and on the other contact with was seeing a very erratic signal when the meter went into the error condition.
Because the nature of the fault made it very intermittent it took quite a while to find where the good signal went bad.
I need to do a couple more checks to be absolutely sure, but it looks like K101 is the problem.
I’ve read some threads on here about it.
-
It’s K101.
After tracing the fault, I confirm that on pin 4 the incoming DC signal looks good, on pin 3 when the unit is reading erratically I get a suitably erratic reading.
Searching around I discovered a number of discussions about K101, including talk of reflowing the contacts, since I cant get a new relay, I tried that.
After doing this I could see that it made an immediate change, the DC measurement had been a bit noisy and after reflowing it looked a lot better, so I left it in MIN/MAX mode for a day and no more erratic readings.
After another 12 hours it started reading erratic again, some times reading ok, other times reading erratically.
When it’s in erratic mode I can’t short 3&4 and the reading goes normal, so it’s definitely K101.
Not sure where to go from here since I can’t purchase a new replacement part as they’re not available.
There’s some talk of other lower rated replacements, but the 500V replacement isn’t available any more and only a 200V part:
-
Assuming that you are referring to Coto 3540-05 as the replacement that is no longer available, then I guess that
us 34401A enthusiasts are all doomed you will have to improvise. One possibility is to salvage the part from a scrap device (34401A or E1412A, maybe others?). What I did for another HP device was to make a little adapter board to fit a modern relay. You could also do this kind of substitution with some flying wires, but I can't imagine either of these 'fixes' doing the "low thermal" characteristic any favours.
-
small thread about this who link another
https://www.eevblog.com/forum/repair/agilent-34401a-k101-relay-replacement/ (https://www.eevblog.com/forum/repair/agilent-34401a-k101-relay-replacement/)
-
FWIW I got an ebay 34401A courtesy of @cellularmytosis with the offer that he'd buy it from me if I didn't want it. I had seen it but passed. He bought it and very kindly offered it to me if it worked properly as he was buying it to hack.
When it arrived it just displayed random digits in all modes. Naturally, I opened it up for a look. There was a 1 cm brown spot of solder flux where an op amp for the ADC had been replaced. Once I cleaned it with isopropyl and a soft toothbrush it worked flawlessly and is more accurate and stable than my other 34401A which is in NOS condition. The wonky one was used a lot and had been cal'd many times. The closet queen had never had a cal done despite advanced age. But it does look like new. I learned the war horse beat the closet queen during USA Cal Club round 2.
I have now done almost a dozen "repairs" on a variety of mostly consumer kit that required nothing more than removal of the flux residue. Flux is hygroscopic. It absorbs moisture from the air and forms high impedance leakage paths. This raises the noise level in the typical low voltage MCUs that run the UI and perform other tasks.
Good luck and
HaveFun!
Reg
-
.. There was a 1 cm brown spot of solder flux where an op amp for the ADC had been replaced. Once I cleaned it with isopropyl and a soft toothbrush it worked flawlessly and is more accurate and stable than my other 34401A which is in NOS condition..
Which one opamp?
-
IIRC It was one of the pair of LM357s (i.e. unobtainium), but this was several years ago and I don't want tear it down to check. In any case, my point is not about the op amp, but the pernicious effect of flux residues. It's merely a relevant example. Most often the problem is residue around the MCU logic lines rather than an analog circuit.
My stereo started reporting "No disk" with CD-Rs and not long after with stamped CDs. Culprit was a very small bit of residue (3-4 mm). Much of which was underneath the chip. I didn't manage to get that on the first try. But a 2nd pass in which I saturated the area with IPA did the trick and 4-5 years later it still works fine and has *never* reported "No disk" again.
My first such "repair" was Mother's DVD/VHS remote. Since then I've revived about $2000 worth of kit at actual value (e.g. ~$335 for the 34401A). I've now done this so often my first step if something gets wonky is teardown, clean and reassemble. I've been amazed and not little angered by how often failure to clean the PCB causes a fault. I tumbled to the issue when I was living in Houston,TX and had my second VCR fail after 12-18 months despite very limited use. When I took them apart to see what might be scavenged I saw that the PCB was covered with flux. Houston is humid. I saw airplanes spewing fog on *many* planes during the 18 months I commuted between Dallas and Houston every week.
-
This particular 34401a looks like it has come right off the production line, the inside is immaculate with no evidence that anything has ever been reworked, once I saw the error I opened it up and spent quite a while inspecting for anything obvious, nothing looked like it gone pop and no rework or residue of any kind anywhere.
I have since applied heat to the contact pins on K101 as per the other thread, I did this originally and it worked for 48 hours until I powered off for a few hours and when I powered it up again, it had gone faulty again. The second time of apply heat I did it at the temperature and with the timing proposed in the other thread, since then it’s continued working even after powering off for a number of hours.
In the absence of a better solution, I’ve gone with that again, it could well just be placebo and just luck that for the moment the problem has gone away. It’s going for a 120 mile trip in the car tomorrow, so it will be interesting to see how it behaves once I get to the other end.
I have purchased a used replacement, with it being used it could have its own set of issues or not work at all, but it was worth a punt.
-
small thread about this who link another
https://www.eevblog.com/forum/repair/agilent-34401a-k101-relay-replacement/ (https://www.eevblog.com/forum/repair/agilent-34401a-k101-relay-replacement/)
Actually K101 is not specifically related to 2W resistance measurement. This relay is used for all measurements that go through the main "High" jack I.e. everything except current. Every time you switch mode or range it temporarily disconnects the jack then reconnects it again. Arguably it's not needed and could be shorted out, except that would cause a self test failure as it is also used to isolated the jack during internal diagnostics. It would also make a lot of sense if remains open during current measurements but I've never checked to see if this is the case.
-
This particular 34401a looks like it has come right off the production line, the inside is immaculate with no evidence that anything has ever been reworked, once I saw the error I opened it up and spent quite a while inspecting for anything obvious, nothing looked like it gone pop and no rework or residue of any kind anywhere.
I have since applied heat to the contact pins on K101 as per the other thread, I did this originally and it worked for 48 hours until I powered off for a few hours and when I powered it up again, it had gone faulty again. The second time of apply heat I did it at the temperature and with the timing proposed in the other thread, since then it’s continued working even after powering off for a number of hours.
In the absence of a better solution, I’ve gone with that again, it could well just be placebo and just luck that for the moment the problem has gone away. It’s going for a 120 mile trip in the car tomorrow, so it will be interesting to see how it behaves once I get to the other end.
I have purchased a used replacement, with it being used it could have its own set of issues or not work at all, but it was worth a punt.
My first serious scope, a used USAF dual trace 60 MHz Dumont 1062 died a few days after the 30 day warranty expired. After many hours of work I traced the problem to cracked solder joints.
Good luck,
Reg
-
I have had it back under test here since we arrived home earlier today.
So far so good, it’s been powered up for 6 or so hours and it looks good, I could sense when it was likely to go bad because the measurements were quite noisy, and they look pretty good.
So the journey in the back of the car doesn’t appear to have sent K101 back into a fault condition, 7 hours isn’t anywhere near enough time to make a judgement though, but initial signs look ok for the moment, I will need to do a few thermal cycles.
-
Just a passing thought... my nice old 33120a had what appeared to be relay issues that actually turned out to be traces going bad somewhere inside.
-
Just a passing thought... my nice old 33120a had what appeared to be relay issues that actually turned out to be traces going bad somewhere inside.
Always worth mentioning stuff like that in case somebody comes across the same issue.
In this case it’s definitely K101, eventually I had it in the fault mode and was able to probe the input contact and output contact, input was perfect, output was very wrong, shorting the pins with tweezers restored the correct reading, releasing the tweezers the reading went bad immediately.
I couldn’t find any posts relating to the exact problem I was seeing, although plenty of posts about K101 going bad but mainly people seeing a issue in resistance mode, I hadn’t been using resistance mode though.
I read the post about restoring K101 functionality by apply an iron to the contacts, I originally just gave them a couple of seconds and I didn’t see the issue for another day and a half, I powered it down for quite a few hours for a thermal cycle, the issue came back.
At this point I figured I may as well just follow the thread advice exactly, since then I haven’t seen it reoccur, it has been on a long car journey, had a number of thermal cycles and range changes and so far it’s behaving, of course now I’ve said that…..
-
You should ask on the groups.io HPAK list. The full name is HPAK spelled out in full. It's the best support available for old HPAK KIT. Lot's of grey beards. It's not unusual to grt a response from a member of the orinal design team.
-
The used K101 relay I grabbed off eBay arrived a week ago, I figured it was worth grabbing since it was only 35 GBP including shipping from Poland to the UK.
I was of course prepared for it being dead on arrival or itself being faulty, but I opened up my unit and replaced it.
Powered up and things seemed ok, the relay was working which was the first worry out of the way.
I had had the unit powered up for a week now in min/max mode and there has been no further reoccurrence of the original problem, which was that after some about of time being powered up, the measurements would suddenly start going wrong in DC mode (and I assume other modes that use the HI input).
I would have seen this problem happen many times over the period I’ve now had it powered up for, so I’m happy to now conclude that the problem has been resolved.
I am not sure how lucky I got that somebody had one of these relays on eBay at the exact time I needed one, I haven’t seen one listed once then and doing a search for completed items doesn’t throw up and recent past sales either.
-
I picked up another 34401a a while back, that one was exhibiting a very similar issue, although it was showing random readings after a short amount of time powered on, I hadn't bothered to get to the bottom of it since it was a spare, I was more interested in getting my original unit back working reliably.
A couple of days back I decided to look at this other unit after deeming the unit mentioned in this thread, repaired.
There are a couple of things with this other unit.
1.) It shows random readings.
2.) Switching to AC while feeding in my DC source shows an AC reading of 0.362V (ish, moving about). The signal I'm feeding it definitely doesn't have that much ripple on it, I am feeding it into my original 34401a at the same time and that it showing an effective ripple of 0 compared to what I'm seeing on this, however, this ripple only seems to affect AC measurements so there's something not right in the AC mode, since in DC I'm measuring a 5mV signal, I'm not seeing that move at all (and certainly not 360mV) so that's less of an problem for me currently, I've marked up on my PDF schematic places to probe to figure out what is wrong with AC mode when I get the time.
So, I opened the unit up to see what was going on. I put it in resistance mode and shorted the terminals on the front and I was seeing 1 ohm (ish), it was rising. I shorted the terminals on the switch hoping that was the cause but that made no difference.
I flipped the unit over and shorted the terminals on K101 and immediately the the resistance measurement dropped down to zero, K101, again. :(
Since I don't have another K101 replacement, I had purchased the COTO 3501-05-511 relay as per the K101 thread here, this has vastly lower voltage switching capabilities, but I had no other options.
I swapped out the relay and powered up the unit and the resistance was fixed and it's been running for a couple of days since the swap in min/max mode and there have been no bad readings, this unit was throwing bad readings constantly, so swapping the relay has resolved the major issues that prevented me using the unit.
I got lucky with a relay I got from eBay that had been pulled from a unit, but I don't think I'm likely to get another one that way.