Author Topic: Fluke 8840A Capacitor Replacement  (Read 2001 times)

0 Members and 1 Guest are viewing this topic.

Offline pinpassionTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Fluke 8840A Capacitor Replacement
« on: June 01, 2020, 09:52:35 pm »
Hi Gang,

I just picked up a Fluke 8840A from the auction place. I liked the 8842 tear-down Dave did, and that inspired me to pick this one up. I got lucky, because it is better then I expected it would be. I looked under the hood before I powered it up, and to my surprise found it had the AC option board installed. That was not listed in the auction, and it was not seen as a factory installed option on the back panel.

I wanted to look at the board for any obvious problems before juicing it up. It looked perfect. All the PS caps looked good, no sign of any previous repairs or modifications. The fuses were all the correct values, and were aok. No sign of any thermal problems were seen. It passed the sniff test. Leaving the cover off, I turned it on and the display sprang to life. Nice and bright, no dim segments. So far, so good.

I don't have any precision voltage or resistance sources, but I do have other multi-meters to check it against. I ran it through some AC and DC voltage checks, and some resistance measurements. Looks good. I checked to see if anything on the board was getting hot. Nope..cool as a cucumber. I did the self test based on the service manual, and all passed.

Now, the date codes on the chips tell me this was made in mid-1990. So, this puppy is 30 years old. It seems to have been treated well in a lab environment, and is cosmetically nice. A keeper! Ok, now I want to keep it happy for years to come. That brings me to the point of my post.

I want to replace the PS electrolytic caps, and the tantalum caps I hear everybody looks at suspiciously. I want to go with high quality replacements. Based on the schematic and parts callout in the service manual, this identifies the following:

Electrolytic s
--------------
C601, electrolytic 6800 uf, 16 volts

C603, electrolytic 330 uf, 100 volts

C605 & C607, electrolytic 470 uf, 50 volts  > (SOLV PR) Not sure what that means?? It's in the call-out.

C611, electrolytic 100 uf, 50 volts

Tants
------
C204,602,608, C612,  1uf, 35 volts

That's what I see, 5 electrolytics and 4 tantalum's. I figure Nichicon for the electrolytics. I can search out Digikey or Mouser for them. I don't know what to use for the tantalums..? If anyone has done this before on a Fluke 8840A, and has part numbers to share..BY ALL MEANS, PLEASE DO. As it stands, there does not appear to be any issues at present, and I want to keep it that way. I can look at all the PS rails and measure voltages and ripple, but I would like to change them even if all looks good due to their age. I'm open to suggestion on what to use for the tants or the replacements in general!!

Thanks for reading, and any advice offered!

Mike
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14072
  • Country: de
Re: Fluke 8840A Capacitor Replacement
« Reply #1 on: June 02, 2020, 06:54:52 am »
If the tantalums are still the old drop shape ones, one should replace them. If there are the more modern brick shape SMD form, they are probably OK. The manufacturers learned to build them better and the circuit designers also learned not to use them in places where failure (short) results in fireworks. I don't think a tantalum cap from 2000 is less reliable than a new one - the old one is already tested for the early drop outs.
 

Offline pinpassionTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Re: Fluke 8840A Capacitor Replacement
« Reply #2 on: June 02, 2020, 09:35:49 am »
Thanks Kleinstein. Unfortunately, the Tants are the drop shaped parts. I didn't see any surface mount components on the board..Old school, but so am I..LOL. These 62 year old eyeballs of mine appreciate the old components when it's time to make a replacement.

Do the new tantalum caps still retain the drop shape? I don't do electronics for a living. It's just been a hobby and passion of mine since grade school. I'm not aware of what the failure mode is on them, but I have heard they do go short circuit and cause fireworks. I want to avoid that! I suppose I can find some diagrams online of what the new components look like. As long as they are compatible with the PC board, I don't care what they look like..providing they don't go bang.

Mike
 

Online tautech

  • Super Contributor
  • ***
  • Posts: 28136
  • Country: nz
  • Taupaki Technologies Ltd. Siglent Distributor NZ.
    • Taupaki Technologies Ltd.
Re: Fluke 8840A Capacitor Replacement
« Reply #3 on: June 02, 2020, 09:51:53 am »
Modern bead tants are fine, just be sure to voltage derate them 50% or more.

Avid Rabid Hobbyist
Siglent Youtube channel: https://www.youtube.com/@SiglentVideo/videos
 

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Fluke 8840A Capacitor Replacement
« Reply #4 on: June 02, 2020, 12:38:55 pm »
Agreed, modern beaded tants are much more reliable than the earlier versions. But it is best not to run them close to their specified voltage. If those 35V tants are in a circuit of 25V or less then 35V replacements are OK. If over 25V go with 50V as an extra measure.
An old gray beard with an attitude.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4643
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Fluke 8840A Capacitor Replacement
« Reply #5 on: June 02, 2020, 04:49:06 pm »
When I recapped mine, I didn't bother with the tants. Med will probably never speak to me ever again...   :-DD

C204 and 602 are on the 5V rail, C608 on the -15V rail, but C612 is on the 30V rail, so it wouldn't be unreasonable to uprate the voltage on that when you replace it.

The only electro that was out of spec in mine was C601, which is the main pre-reg reservoir cap on the 5V rail. It measured something like 8500uF, way above 6800, and possibly indicative of degraded dielectric.

Given they're all just being used as decoupling caps, with the exception of 204 which is part of the RC for the reset on the CPU, if you're really anti-tant, there should be no reason you can't use electros.
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Online Kleinstein

  • Super Contributor
  • ***
  • Posts: 14072
  • Country: de
Re: Fluke 8840A Capacitor Replacement
« Reply #6 on: June 02, 2020, 05:38:17 pm »
The only electro that was out of spec in mine was C601, which is the main pre-reg reservoir cap on the 5V rail. It measured something like 8500uF, way above 6800, and possibly indicative of degraded dielectric.
The electrolytic caps often have an asymmetric tolerance (allowing more positive than negative deviation, like +50%/-20%). So the 8.5 mF instead of 6.8 mF is probably still inside the tolerance range. If not used for a long time the capacitance may drift up, but would come back after some use and not formation, growing a thicker oxide again.

The tantalum caps had lower ESR, this may still be important a low temperature. Otherwise low ESR normal caps may be OK.
 

Offline med6753

  • Super Contributor
  • ***
  • Posts: 11313
  • Country: us
  • Tek nut
Re: Fluke 8840A Capacitor Replacement
« Reply #7 on: June 02, 2020, 05:50:05 pm »
When I recapped mine, I didn't bother with the tants. Med will probably never speak to me ever again...   :-DD



I'm no lover of tants. But if the circuit calls for them I try to use them.  ;D
An old gray beard with an attitude.
 

Online bdunham7

  • Super Contributor
  • ***
  • Posts: 7727
  • Country: us
Re: Fluke 8840A Capacitor Replacement
« Reply #8 on: June 02, 2020, 06:47:47 pm »
Just did a 8842A, here are the p/n for the electrolytics from Mouser.  These were selected to physically fit nicely and install easily in addition to their electrical characteristics. 

UPW2A331MHD1CA 330u/100V

UPW1J471MHD3 470u/63V (qty 2)

UPW1J101MPD 100u/63V

LGY1E682MELZ30 6800u/25V

I carefully tested the power supply outputs for ripple and noise both before and after, and I tested the removed capacitors after removal.  The power supply readings were unchanged and the capacitors all tested completely good and had no signs of physical leakage, reverse current leakage was very low, ESR was low and all of the capacities were over nominal except the 330uF one was reading 265.  I had tested that one in circuit and that slight reduction in capacity is what prompted me to change them all.  I think it is pretty clear that this was all entirely unnecessary, so take that for what it's worth. 

I did not replace the tantalums.  They looked good and measured perfectly in-circuit.  If you do replace them, I'd go up a voltage rating to 50V--C612 is exposed to 30 volts, pretty close to it's max rating per the parts list.  I would suggest Kemet T356B105K050AT from Mouser.  I believe there are also a couple of the same tantalum caps on the AC board, although I don't know what voltage they are exposed to.

If I were you I would consider leaving it alone and maybe concentrate on making a nice set of 4-wire Kelvin leads for it or something.

And if you didn't know, getting a 'free' Option 09 AC board is like winning the lottery. 



« Last Edit: June 02, 2020, 06:54:17 pm by bdunham7 »
A 3.5 digit 4.5 digit 5 digit 5.5 digit 6.5 digit 7.5 digit DMM is good enough for most people.
 

Offline AVGresponding

  • Super Contributor
  • ***
  • Posts: 4643
  • Country: england
  • Exploring Rabbit Holes Since The 1970s
Re: Fluke 8840A Capacitor Replacement
« Reply #9 on: June 03, 2020, 05:24:18 am »
The only electro that was out of spec in mine was C601, which is the main pre-reg reservoir cap on the 5V rail. It measured something like 8500uF, way above 6800, and possibly indicative of degraded dielectric.
The electrolytic caps often have an asymmetric tolerance (allowing more positive than negative deviation, like +50%/-20%). So the 8.5 mF instead of 6.8 mF is probably still inside the tolerance range. If not used for a long time the capacitance may drift up, but would come back after some use and not formation, growing a thicker oxide again.

The tantalum caps had lower ESR, this may still be important a low temperature. Otherwise low ESR normal caps may be OK.

Yes, I'm aware of the tolerance situation. This mostly refers to lesser quality caps, some have +80%!

C601 calls for +/-20%, so it was definitely out of spec.

I don't think ESR is the primary consideration here, given the application. If I had to guess, it would be package size, and possibly tolerance.
nuqDaq yuch Dapol?
Addiction count: Agilent-AVO-BlackStar-Brymen-Chauvin Arnoux-Fluke-GenRad-Hameg-HP-Keithley-IsoTech-Mastech-Megger-Metrix-Micronta-Racal-RFL-Siglent-Solartron-Tektronix-Thurlby-Time Electronics-TTi-UniT
 

Offline pinpassionTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Re: Fluke 8840A Capacitor Replacement
« Reply #10 on: June 03, 2020, 01:11:28 pm »
Hi Gang,

Thanks to all that responded and offered thier help. I appreciate it.

Before I saw the list of parts used to replace the caps in the Fluke 8842, I went to Mouser and ordered the following:


647-UVZ2A331MHD1TO - 330uF 100 Volts 20%

647-URY1H471MRD - 50volts 470uF

647-UVZ1H101MPD1TD - 100uF 50V 20%

647-UVR1C682MHD - 16volts 6800uF

80-T354A105J35AT-TR - 35volts 1uF 5%

I wish I saw the list before I ordered. It was a dizzying array of parts and choices for just a few caps!! Jezz!! I am still concerned if the choices I made will fit on the board..that was the reason I had hoped someone that did it before might still have available the choices they made..I jumped the gun too quickly..my bad. I did go with Nichicon electrolytics, and the tantalums are Kemet. Some of the choices I made were because they had them in stock, whereas others were not. I went for general purpose caps, nothing fancy. I tended to give some weight to parts they had in larger quantities thinking they were popular replacement and more "universal". I don't know how flawed that thinking may have been. Anyway, they come tomorrow, so we will see.

After reading the replies, I'm not sure I really need to bother replacing them..LOL  Well, I just got the meter this week, so maybe I'll play around with it for a little while before I do. In a way, it's kind of nice to have something that has never seen a soldering gun since leaving the factory.

The AC option board WAS like winning the lottery!! It was totally unexpected, and since I had seen previous posts on the 8840A in the blog, I knew where the board was located, and what it looked like. As soon as I slid the cabinet off, I saw it and was wearing a big grin. It works too, no issues. One thing I did see was that mine does not have the small sheild on the AC board bottom like others I have seen. I don't know if that is due to it's revision level or what..I have not removed the board yet to have a look at the component side. I don't trust the little nylon pins that hold it to the chassis mount and I know they get brittle and easily break. One is already broken. I am hesitant to push my luck with them.

Since I am a test equipment junkie, I am expecting an Agilent 34401A (another Ebay purchase) this week. It looks near new, and is tested working just out of calibraion a few months ago. I hope I am as lucky with that as I was with this!!

Thanks again guys!! You all are such a great resource!!

Mike


 

Offline pinpassionTopic starter

  • Contributor
  • Posts: 13
  • Country: us
Re: Fluke 8840A Capacitor Replacement
« Reply #11 on: June 05, 2020, 12:21:07 pm »
Hi Gang,

As suggested by Bdunham7, I'm interested in getting a nice set of 4-wire Kelvin leads. i would prefer to get commercially made cables rather then rolling my own. Does anyone have a recommendation on what cables are available that are are good quality at a decent price. I think a set that bundles both the current and sense leads per lead would be easier to use, so I would still have two leads lying on the bench for the 4-wire measurement. I know they would not be as flexible as the individual wires alone, but that seems like an acceptable trade-off. Good silicon jackets would help I think. Clips at the test end, and proper jacks that would mate with most common multi meters are what I have in mind.

Any ideas?

Thanks.

Mike
 

Offline GigaJoe

  • Frequent Contributor
  • **
  • Posts: 484
  • Country: ca
Re: Fluke 8840A Capacitor Replacement
« Reply #12 on: August 15, 2021, 04:34:42 pm »
I decided to replace my , 

470uF\50V  ->  1800 /50V
330 \ 100V ->  1000\50V (VFD display , 38V )
6800 \ 16V ->  3X 6800 \ 16V
dont bother tantalum

became a bit better , on  M  speed , LSD wobbling ±1

 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf