Author Topic: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert  (Read 1587 times)

0 Members and 1 Guest are viewing this topic.

Offline jewelie

  • Regular Contributor
  • *
  • Posts: 75
  • Country: gb
  • Everyone's gotta start somewhere
    • Personal website, of sorts, sometimes with a Pi-driven bird-feeder webcam (currently down but will be putting it up again in the near future with IR capability.)
Hiya

This is probably a bit of an arrogant thing for a newbie to say, but I think I spotted either a design flaw or a manufacturing defect on an old Fluke 87 I've very happily inherited off a family member.

Although very old it was barely used, so it didn't deserve to have any technical problems, but it did have one: the input alert function, when a lead is plugged into the 10A or the mA/uA socket and the function switch isn't set to A/mA/uA, only worked for the mA socket, not the 10A socket!

Was very nerve wracking for a very skint newb otherwise stuck with the kind of meters that would make Dave *criiinge* but I had a go at diagnosing it after reviewing the circuit and the service manual and did eventually fix it-



(BTW the jump wire does genuinely miss the contact inbetween by a significant margin, about 2mm, it's just the angle of the photo.)

That seems pretty poor design or manufacturing to me - a trace carrying a safety-related so close to the edge of the PCB that it could, in essence, fall off!

So anyway, if anyone else comes across the same problem on an old/original Fluke 87, I wouldn't be surprised if it wasn't the same issue.

(Yeah, feeling rather pleased with myself, had to post this somewhere. :D )

Julie
x
Anxious newbie to EEVblog.  Resuming an interest in basic electronics after a close shave with a joint EE/Computing degree decades ago
 
The following users thanked this post: saturation

Offline retiredcaps

  • Super Contributor
  • ***
  • Posts: 3318
  • Country: ca
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #1 on: June 04, 2016, 12:22:25 am »
Good job troubleshooting and diagnosing the problem and then sharing it with us.
 
The following users thanked this post: jewelie

Online tautech

  • Super Contributor
  • ***
  • Posts: 16024
  • Country: nz
  • Taupaki Technologies Ltd. NZ Siglent Distributor
    • Taupaki Technologies Ltd.
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #2 on: June 04, 2016, 01:07:49 am »
Nice find and fix. Maybe you could've used some sleeving or insulated wire for the bodge.
Looks like abrasion was the cause of the *lost* trace.
Avid Rabid Hobbyist
 
The following users thanked this post: jewelie

Offline retiredcaps

  • Super Contributor
  • ***
  • Posts: 3318
  • Country: ca
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #3 on: June 04, 2016, 01:23:15 am »
Modemhead's blog contains a few cases of corroded traces on Fluke meters fixed with 30 awg wire.  He doesn't use any insulation on some of the repairs because the wire follows the original pcb layout path exactly where the copper traces are eaten away.

http://mrmodemhead.com/blog/all-posts/

The problem with the original Fluke 80 series is that the user has to remove the case to change the battery and fuses which leads to possible contamination of the pcb board due to dirt, greasy hands, environment, etc.  Handling the pcb carelessly and improper taking apart and putting back together could also lead to damage for those traces very close to the edge of the pcb.
 
The following users thanked this post: jewelie

Offline jewelie

  • Regular Contributor
  • *
  • Posts: 75
  • Country: gb
  • Everyone's gotta start somewhere
    • Personal website, of sorts, sometimes with a Pi-driven bird-feeder webcam (currently down but will be putting it up again in the near future with IR capability.)
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #4 on: June 04, 2016, 05:53:59 pm »
Nice find and fix. Maybe you could've used some sleeving or insulated wire for the bodge.
Looks like abrasion was the cause of the *lost* trace.

Yeah, I really don't like not having insulation around it, but I had to be a bit pragmatic on this occasion.  It was a single strand from a 22AWG multistranded cable and I didn't have any kind of insulator to put around it that was small enough and wouldn't have got in the way of trying to do the soldering.  On the pragmatic side it's pretty firm, mostly covered up even when the meter covers are off (so no accidental knocks), and it sticks up a good 2mm out of the way.  Blurgh.

A drop of epoxy resin perhaps?  I'm a bit worried about using *anything* in case it's mildly conducting!
Anxious newbie to EEVblog.  Resuming an interest in basic electronics after a close shave with a joint EE/Computing degree decades ago
 

Online Monkeh

  • Super Contributor
  • ***
  • Posts: 5945
  • Country: gb
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #5 on: June 04, 2016, 06:31:50 pm »
Epoxy won't be conductive. Certainly not conductive enough to matter.

You could've taken that wire over the trace instead of the via, and made the connection at the via on the other side of the TO92 component. Much easier to avoid short risks and secure the wire.
 
The following users thanked this post: jewelie

Offline jewelie

  • Regular Contributor
  • *
  • Posts: 75
  • Country: gb
  • Everyone's gotta start somewhere
    • Personal website, of sorts, sometimes with a Pi-driven bird-feeder webcam (currently down but will be putting it up again in the near future with IR capability.)
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #6 on: June 04, 2016, 06:36:16 pm »
Modemhead's blog contains a few cases of corroded traces on Fluke meters fixed with 30 awg wire.  He doesn't use any insulation on some of the repairs because the wire follows the original pcb layout path exactly where the copper traces are eaten away.

http://mrmodemhead.com/blog/all-posts/

Thanks for that!  Yeah, I think I'll get some 30AWG wire-wrap wire for such jobs in future.  :)  (Update: ordered. :) )

I had it following the original trace to begin with then realised that was probably pretty stupid given what became of the original trace and bent up upwards, out of the way, instead, lol. ;)

The problem with the original Fluke 80 series is that the user has to remove the case to change the battery and fuses which leads to possible contamination of the pcb board due to dirt, greasy hands, environment, etc.  Handling the pcb carelessly and improper taking apart and putting back together could also lead to damage for those traces very close to the edge of the pcb.

Yeah.  I don't know if it had happened before arriving at my doorstep, but it is quite possible that I did in taking it apart and cleaning it up (silly me).... but to be fair to me, in that picture, you can see how even the text on edge of the PCB is uncovered and "falling off", so to speak - that's how far out the PCB trace went too it would appear.  Fortunately it doesn't look like anything else on the PCB went as far out.

I do think I've become a convert to old 2nd hand stuff versus cheap new stuff now for this kind of equipment.  :)  The time spent reading the schematic, appreciating how it all works, diagnosing the fault, and fixing it was absolutely priceless.  I have a terrible problem with lack of self-confidence and always convince myself that I'll just break things or not succeed before I've even started, but actually having a service manual and schematics really made the process much more comfortable to attempt. 

Additionally, getting at that trace properly involved taking the internal cover & screen off, which meant an opportunity to clean up the zebra strips, which is something I'd been avoiding doing and turned out to be ridiculously easy to do.  The contrast, or rather the viewing angle is still a bit iffy though when viewed from above as opposed to below.  Don't know if this is still the zebra strips, or a fault, or just either normal for this device or even just a factor of age on LCDs?




Everything I've been able to test on the meter is now reading spot on.  :) 

All in all it's been great buzz of an experience and confidence boost with a nice bounty at the end!    Ultimately Dad sent to me one very beat up and thoroughly broken original Fluke 73 and a less mangled Fluke 87; I ended up with what appears to be (touch wood) now a fully working 87 and he now has an equally fully working 73 (one that looked so different, he said he couldn't begin to recognize it when it arrived.)  Best of all, I was even able to re-"calibrate" the 73 as it was from an era where doing so involved a single pot and I had a MAX6225 2.5V reference (it was *well* out, which is unsurprising given what it had gone through.)

So he's got a nice, clean, accurate working meter and I've got both a confidence boost and a dream meter that I'd *never* have been able to afford otherwise and all it cost was time, effort and postage.



I'm getting to that point now where I *crave* free broken things, just for the opportunity to try to repair and recover something that would otherwise have ended in a bin somewhere.  :)

Julie
xxx
Anxious newbie to EEVblog.  Resuming an interest in basic electronics after a close shave with a joint EE/Computing degree decades ago
 

Offline jewelie

  • Regular Contributor
  • *
  • Posts: 75
  • Country: gb
  • Everyone's gotta start somewhere
    • Personal website, of sorts, sometimes with a Pi-driven bird-feeder webcam (currently down but will be putting it up again in the near future with IR capability.)
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #7 on: June 04, 2016, 06:40:14 pm »
Epoxy won't be conductive. Certainly not conductive enough to matter.

You could've taken that wire over the trace instead of the via, and made the connection at the via on the other side of the TO92 component. Much easier to avoid short risks and secure the wire.

Agreed.  When the wire-wrap wire arrives, I think I might very well re-do it.
Anxious newbie to EEVblog.  Resuming an interest in basic electronics after a close shave with a joint EE/Computing degree decades ago
 

Offline saturation

  • Super Contributor
  • ***
  • Posts: 4788
  • Country: us
  • Doveryai, no proveryai
    • NIST
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #8 on: June 04, 2016, 07:31:54 pm »
Good job.  Be wary using older Flukes on CAT III level circuits as the conformal coating seems to be falling, aka delaminating, away.  I've seen many an 80s V1 with rippling, bubbling, and peeling off coating on the PCB as can be seen in your photo.  I wouldn't be surprised that that trace came off were the coat had delaminated.  Just check the integrity of the traces particularly along the inputs; you probably have given the sleuthing to find that fault, and then, enjoy.
Best Wishes,

 Saturation
 
The following users thanked this post: jewelie

Offline jewelie

  • Regular Contributor
  • *
  • Posts: 75
  • Country: gb
  • Everyone's gotta start somewhere
    • Personal website, of sorts, sometimes with a Pi-driven bird-feeder webcam (currently down but will be putting it up again in the near future with IR capability.)
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #9 on: June 04, 2016, 08:05:47 pm »
Good job.  Be wary using older Flukes on CAT III level circuits as the conformal coating seems to be falling, aka delaminating, away.  I've seen many an 80s V1 with rippling, bubbling, and peeling off coating on the PCB as can be seen in your photo.  I wouldn't be surprised that that trace came off were the coat had delaminated.  Just check the integrity of the traces particularly along the inputs; you probably have given the sleuthing to find that fault, and then, enjoy.

Anything over about 20V or 0.5A scares me silly and I don't go there generally; I take extreme precaution on the few occasions I attempt to measure mains voltage etc.  Yes, I had a good look in general around the board when I got it, treating it as an unknown quantity, and definitely a good look around the input traces.

I didn't known about the conformal coating issues on old Flukes though, thanks for pointing that out!
Anxious newbie to EEVblog.  Resuming an interest in basic electronics after a close shave with a joint EE/Computing degree decades ago
 

Offline retiredcaps

  • Super Contributor
  • ***
  • Posts: 3318
  • Country: ca
Re: Fluke 87 (original) Product Flaw? - l10A Socket Detect / Input Alert
« Reply #10 on: June 04, 2016, 10:50:47 pm »
I do think I've become a convert to old 2nd hand stuff versus cheap new stuff now for this kind of equipment.  :)  The time spent reading the schematic, appreciating how it all works, diagnosing the fault, and fixing it was absolutely priceless.
I buy used/abused non working meters for the sole purpose of learning.

Quote
I have a terrible problem with lack of self-confidence and always convince myself that I'll just break things or not succeed before I've even started, but actually having a service manual and schematics really made the process much more comfortable to attempt.
The days of getting a good service manual with schematics from Fluke and other manufacturers like Keysight, etc are over.  Only the old stuff has them.

Quote
Additionally, getting at that trace properly involved taking the internal cover & screen off, which meant an opportunity to clean up the zebra strips, which is something I'd been avoiding doing and turned out to be ridiculously easy to do.
As you probably found, modemhead has an entry for the above

http://mrmodemhead.com/blog/fluke-8x-faded-lcd/

Quote
The contrast, or rather the viewing angle is still a bit iffy though when viewed from above as opposed to below.  Don't know if this is still the zebra strips, or a fault, or just either normal for this device or even just a factor of age on LCDs?
It looks okay to me.  The Fluke 70 series does indeed have better contrast. I noticed that with my own meters.

Quote
I'm getting to that point now where I *crave* free broken things, just for the opportunity to try to repair and recover something that would otherwise have ended in a bin somewhere.  :)
Yes, I pickup/get broken things to try and fix as well.
 
The following users thanked this post: jewelie


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf