EEVblog Electronics Community Forum
Electronics => Repair => Topic started by: dohyun on June 23, 2016, 12:05:28 am
-
Howdy all,
I recently got a 34401a on ebay marked as "for parts" but thankfully all the self tests pass. However, I noticed a somewhat odd issue. When testing DC voltage with a 10V ref (from voltagestandard.com) I get a fairly large out of tolerance (though it is close enough to 10V to not be completely wonky). The odd part is, when I switch out of auto ranging and go to manual and step up to the 100V range, it is dead on accurate.
I am trying to figure out if this is a calibration only issue, or if something is off and I can fix it (kinda hoping for the latter)
I've measured as far as U102 input, and from the looks of it on U102 Pin 5, there is the still-solid 10V signal.
Has anyone ever run into something like this before? This is my first attempt at a DMM repair.
-
Err, I meant U101 there, not U102
-
Additional info:
It looks like it is not just the 10V range, when testing against 1V and 100mv (using an additional borrowed 34401a) the DCV measurements in my 34401a are off for the 100mv, 1V, and 10V range, but a dead on match with the other 34401a on the 100V and 1000V ranges.
-
Not an answer, but gathering more evidence: There is an option to operate the meter in either high-impedance mode, or 10M\$\Omega\$ input impedance on voltage ranges of 10V or less - does changing this option in the system menu make any difference?
-
Thanks for the reply Andy.
I just tried that and the results didn't change between the 10M and high-impedance mode.
-
Since you have a 10V reference, I would just calibrate the 10V only and see the result.
-
Makes sense. I've got the Fluke 4 terminal shorting plug on order to zero adjust for the calibration and then will give that a shot. Will update with the result!
-
Makes sense. I've got the Fluke 4 terminal shorting plug on order to zero adjust for the calibration and then will give that a shot. Will update with the result!
Even better.
Do the zero adjust first and then the 10V DC calibration only.
I had a corrupted storage chip before and it would not write the new calibration data.
-
I was about the ask for some actual values for those "fairly large out of tolerance" measurements (quantitative vs. qualitative). It sound like the meter just needs adjustment, but I'll wait to see the outcome of the zero and 10V calibrations.
-
Since it will be a few days before the shorting plug gets here, here are some numbers:
10V ref calibrated (by an in-cal 3458A) at 10.00000V
When measuring the 10V ref in the 10V range I get 10.06887V however when I move up to the 100V range I get 10.0002V (which looks to be in tolerance for the 100V range)
-
That's out alright, but less than I was imagining for "fairly large". Looking forward to your post adjustment update.
-
I think that did the trick. After zero calibration and calibrating the 10V range to the 10V reference it is looking consistent and solid.
Since it looked to be a calibration adjustment issue, I might take advantage of the $75 calibration w/ data deal from Global Test Equipment for some of the other functions which I am not as able to verify their performance.
All in all I think it was a worthwhile score, $157 (including shipping) for the "broken" meter and $35 for the fluke 4 wire shorting block.
But for now, things are looking great, thank you all for the help!
-
Great news,
Before you send it out for calibration, keep it running for a few days to see the variation of your 10V standard and make sure you don't have a too large drift ind the 10V range. There must have been a reason for the range to be so far off.
-
I think that did the trick. After zero calibration and calibrating the 10V range to the 10V reference it is looking consistent and solid.
Since it looked to be a calibration adjustment issue, I might take advantage of the $75 calibration w/ data deal from Global Test Equipment for some of the other functions which I am not as able to verify their performance.
All in all I think it was a worthwhile score, $157 (including shipping) for the "broken" meter and $35 for the fluke 4 wire shorting block.
But for now, things are looking great, thank you all for the help!
Nice work, and a great price. Thanks for the info on GTE. I need to have a (repaired) 34970A calibrated that's had it's FRAMs replaced. Last time I had one calibrated it cost me $125 + return shipping. GTE covers the return shipping! :-+
Jay
-
Good call, I will keep it running for a few days and see if there is any more drift.
-
Congrats on that adjustment working out. Looks like you got yourself a great deal. Definitely see how well it holds over time before shelling out money for the full adjustment.
-
So far the measurements (from spot checking) have only varied from 9.99999 to 10.00003 over the past couple of days. However I am going to set up some automated measuring to watch it a bit longer before feeling confident enough to send it out for cal.
-
That is already good news.
It seems you got very lucky with that instrument.
The variation you saw so far are very normal and can be contributed to temperature drift of the 34401A and the 10V reference
-
Since it looked to be a calibration adjustment issue, I might take advantage of the $75 calibration w/ data deal from Global Test Equipment for some of the other functions which I am not as able to verify their performance.
Hmm. They're not returning my emails. Did you try contacting them via email by any chance?
Jay
-
Jay,
I did, I e-mailed tech-help@4gte.com asking if the promotion for $75 was still on, it took a few days but someone got back to me asking if I wanted to go ahead (and that yes the $75 deal was still on).
I will send you a message with the contact details of the person who directly replied to me.
-
Yep, the special is still on their site: http://4gte.com/calibration-and-repair/ (http://4gte.com/calibration-and-repair/)
-
Yep, the special is still on their site: http://4gte.com/calibration-and-repair/ (http://4gte.com/calibration-and-repair/)
Hi bitseeker,
Thanks for that. The weirdest thing is that I didn't have to email again. He emailed me and said that I was caught in their spam filter, after buying a TDS5054B from me four minutes earlier. :-DD
Jay
-
Oh, that's a good one. At least they periodically check their spam folder. So, now you're mutual customers.
-
Meter is on its way to the calibration folks. Thanks again for all the help. BTW, GTE gave me a heads up that one of their standards is going out for cal on 7/14 so anything arriving on or after that might be delayed for a few weeks.
-
Hmm, might be good to wait. Get a fresh cal from a fresh cal.
-
I was contemplating that (waiting) but I was also figuring that if they cal my meter and then the standard goes off to be cal'd then I would find out reasonably soon that the cal job they did on mine was in spec (as the end point of the two time points of calibration)
Though I am new to the precision and calibration world so I could be totally off base :)
-
The freshly checked cal instruments can be slightly better than just before the end of calibration cycle. So calibration just after the check might be slightly more accurate. However for the 34401 the calibration instrument should be really much better than the meter. So this should not matter that much. Even at the end of the cycle there instrument should be way better than the 34401. The weak point there is more the disturbance / drift of the 34401 due to shipping and warmup time. It would be a different if you send in an 8 digit meter, where the quality of the reference could be an factor.
Having the calibration just at the beginning also has some downsides: there instrument may still recover from shipping disturbance and they may be in a hurry because of a backlog of waiting instruments.
If at the end they find out there instrument was way off (which is very unlikely), they should in principle give you a note and maybe offer corrections (new calibration). So you have timely check for the unlikely case that there instruments where out of spec.
-
I heard back from the cal lab. Sounds like initially the zero offset was out of tolerance on both the front and rear terminals as well as DCI.
They made adjustments and re-ran the tests this time passing.
The lab noted that there is still some drift with the unit, and that it is not uncommon for the age of the DMM to have that happen. They also mentioned if I get anymore 34401A's to avoid ones with serial numbers starting in 3146A as those are starting to show more signs of issues.
That is a bit worrisome, in part because I had done the zero offset adjust with the Fluke shorting plug prior to sending it in and it was dead on. Hopefully it won't drift outside of tolerances for a good while.
If it does, I guess I am back to trying to find out what is causing the drift and fixing it.
-
Also note that the early 34401A (panel assembly 34401-66502) have unobtanium VFD driver ICs and cannot be replaced by the newer assembly due to firmware incompatibility, which is another reason to get the later units (firmware 06-04-02 or higher), if you buy another.
On the other hand, the VFD itself is the same for both assemblies and new VFDs are available from China, which is a nice benefit of having a 34401A vs. some other model.
-
Thanks for the info on the VFDs! Definitely good to know.
I got the unit back today, will be checking it out and seeing if the things which were out of tol before start drifting at all (zero offsets and DCI).
-
Check the 2W ohms ranges, especially the lowest range. There have been a few 34401A owners that have found issues with the front/rear switch . Typically the 2W ohms tests will show a higher than normal offset and can be seen by exercising the switch a few times.
-
If drift is a possible problem, I would check the input bias current, as one possible source of drift could be excess leakage (e.g contamination) which will show as excess leakage. Transport could have caused condensation / humidity build up.
-
So far so good.
I've gone over the calibration as-received and as-left data and it looks like the main issues were with zero offset on front and rear 100mV range, zero adjust front and rear on 2W 100R and 1KR ranges. DCI gain verification was also out of tol in the 10mA and 100mA ranges.
As left everything was back in tolerance. As far as I've been able to test on my side things are still looking good after the journey from the lab to me.
Thank you again for the help on this! I will keep an eye on the ranges which were out of tolerance in case anything starts drifting heavily. (Though I need to figure out something for a precision current source :)
Also before the next calibration I probably will swap out the VFD itself as it is on the dim side.
-
Sounds good. Regarding the VFD, plenty to choose from:
http://www.ebay.com/sch/i.html?_sacat=0&_nkw=34401a+%28vfd%2Cdisplay%29&_sop=15 (http://www.ebay.com/sch/i.html?_sacat=0&_nkw=34401a+%28vfd%2Cdisplay%29&_sop=15)