Author Topic: Fluke 87V displaying dashes only  (Read 4433 times)

0 Members and 2 Guests are viewing this topic.

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Fluke 87V displaying dashes only
« on: July 03, 2021, 05:01:30 pm »
Hello everyone!  ;D

I have recently gotten my hands on a broken Fluke 87V (rev.9)
Unfortunately it's currently unusable, as selecting any position with the dial results in the multimeter booting up and displaying the F87 logo, however after that it's only 4 dashes. (I'm linking a video showcasing the fault here -> https://streamable.com/0c5k3f   as well as attaching some pics of the board, just in case you spot something I didn't.)
Holding the AutoHOLD button while turning the DMM turns all LCD segments on just fine, so I assume it isn't an issue with the LCD display itself.

Of course I have tried checking the fuses, replacing the battery, cleaning the PCB with isopropyl, different probes, all to no avail.
I can't see any corrosion(under the rotary dial as well), cracked/cold joints(at least to naked eye), nor any obviously blown components anywhere on the board either - no visual clues so far  :(

Has anyone ever seen anything like this with a F87V? I Would like to fix it, but no idea where to even start troubleshooting this other than the stuff I've already tried.
Of course I have contacted Fluke about this, and was told to contact my local Fluke repair point, but I'm not quite ready to give up (and pay someone else to repair it instead) just yet.
Oh, and I haven't found anything in the manual about this sort of behavior either.

Any ideas what causes this? Or how to fix it up? I'm out of ideas  :-[
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #1 on: July 03, 2021, 06:52:25 pm »
Just a guess, maybe ADC is not responding and firmware is stuck waiting forever for conversion to finish? ADC is LTC2435-1. You may scope the SPI to see what is going on there.

Also, what is that on the traces in the middle?
« Last Edit: July 03, 2021, 07:18:11 pm by Manul »
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #2 on: July 03, 2021, 09:02:37 pm »
Heya! Thanks for the reply and tip!
Furthermore, you might have (not 100% sure yet) caught the issue on your first guess.

After reading your comment I've probed the three SPI pins on the ADC with my logic analyzer & another multimeter and found that:


state | SCK      | SDO      | CS     
-------|------------|------------|----------
off :  | floating  | floating  | floating
on  : | LOW      | floating | HIGH     

Some explanation:
State is what the rotary dial is currently set to, off being the off position, and on being any (it was same for all) other.
The LOW and HIGH (3v3 logic) states i feel like are self-explanatory. The "floating" means that the pin was sitting near 0V, but not quite pulled down - my other (working) multimeter was picking up slight voltage fluctuations within milivolt ranges that usually appear when the probes are left floating in the air, which makes me assume those pins weren't in proper "LOW" states. This would also make sense with the logic analyzer channels not being pulled down by those pins. But that's just my speculation, do correct me if I'm wrong about this.
The state of the pins did not change with time.

Hmm, fried ADC? What do you think?
I feel like this is already a huge leap forward  ^-^ Thanks!


P.S. Oh, and regarding the anomaly on the traces in the picture - it's not there anymore, must have been a dust particle.
« Last Edit: July 03, 2021, 09:04:33 pm by Multiple Cheese Slices »
 
The following users thanked this post: bd139

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #3 on: July 03, 2021, 09:35:09 pm »
Looks ok, but MCU is not reading the ADC at all. No CS low, no clock. ADC is probably fine. The question is what causes the MCU to get stuck. Multimeter often experience some mechanical stress during use, maybe some trace, via or solder joint is broken.

Found service manual on the internet. Schematics down in the end.
 
The following users thanked this post: Multiple Cheese Slices

Offline Miti

  • Super Contributor
  • ***
  • Posts: 1088
  • Country: ca
Re: Fluke 87V displaying dashes only
« Reply #4 on: July 04, 2021, 01:52:48 am »
Post a picture with the switch and it’s contacts. Is there any contact missing or misplaced (if that’s even possible with this meter)?
Fear does not stop death, it stops life.
 

Offline bdunham7

  • Super Contributor
  • ***
  • Posts: 3422
  • Country: us
Re: Fluke 87V displaying dashes only
« Reply #5 on: July 04, 2021, 02:44:25 am »
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 Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #6 on: July 04, 2021, 09:23:30 am »
The contacts under the dial seem good.
 

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #7 on: July 04, 2021, 09:30:53 am »
It does in fact behave like the DMM in the vid at 5 minutes in, however (unfortunately) it seems to be permanently stuck instead of recovering. Any way to troubleshoot this further?
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #8 on: July 04, 2021, 01:44:11 pm »
Check rotary switch circuit and its output. Check that the voltages supplied are correct ( -2.5, +0.8 ) and then check how is SWPOS value at different positions. Does it match calculated value of the resistor dicider in that position?

Quote
S1, R22, R53, R54, R55, R58, R59, R60, R61, R84 & C20 allow the position of the
rotary switch S1 to be read by the microprocessor U2 as an analog signal over the signal
path SWPOS. U2 reads the position of S1 by pulling the SWPWR signal to logic high
(+0.8V) and reading the signal at SWPOS. The voltage divider formed by R61, R60,
R59, R58, R55, R54, R53 & R84 creates a signature voltage for each switch position of
S1. R22 will pull this voltage down slightly when the divider taps are connected, but will
pull the SWPOS voltage to -2.5 V when S1 is between switch positions. C20 suppresses
noise.
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #9 on: July 04, 2021, 08:17:00 pm »
Here are the voltages of TP20/pin 5 of the microcontroller at different switch positions:
off      0v
Vac    336mV
Vdc    0.6V
temp. 0.9V
ohms 1.1V
diode 1.4V
mA    1.8V
uA     2.2V


The voltage at "-2.5V" is 0V on my board, and SWPWR is 3.3V. I suppose that's not good(?)  ???
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #10 on: July 04, 2021, 08:54:41 pm »
The voltage at "-2.5V" is 0V on my board, and SWPWR is 3.3V. I suppose that's not good(?)  ???

Are you sure that you used circuit ground as a ground and not -2.5V?
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #11 on: July 04, 2021, 09:15:10 pm »
I have put the black probe on the negative battery terminal and the red probe on the via left of R22/ & solder joint left of R61. Is this the correct way?
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #12 on: July 04, 2021, 09:30:06 pm »
I have put the black probe on the negative battery terminal and the red probe on the via left of R22/ & solder joint left of R61. Is this the correct way?

Well, no, it is not correct, because if you observe schematic of power supplies, you will see that battery negative is actually "-2.5V" in the circuit. And "ground" is a point 2.5V above the battery negative. But that does not change nothing, your measurement results seem ok, except that they are shifted by 2.5V.

If we assume that MCU is ok (which is of course not guaranteed), then logically the problem should be some signal which is input to that MCU. So it waits for something to be ready and it is not ready, something like that. You could use schematic, determine the inputs, determine their function and look for something suspicious. I'm getting out of ideas here, except slow, methodical testing.

By the way, have you tested, that SWPOS signal actually reaches the MCU?
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #13 on: July 04, 2021, 09:53:35 pm »
Ah, I see, in that case my bad about the measurements. Good thing it's only shifted.

Yes, the SWPOS does reach the pin5 on the MCU.

I'll do as You've said. I'll post updates if I spot anything unusual.
Oh, and just in case the fault was caused by GSM (since this fluke is pre-rev11), what would be the actual part getting damaged? the MCU? Would there be any way of making sure whether that is the case or not?

Thanks!
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #14 on: July 04, 2021, 10:08:12 pm »
Can they get permament damage from GSM? I think no, but I don't know for sure. Can not help on this.

One final idea is to check all power supplies according to schematic (+2.5, +0.8, -2.5). Also make sure that MCU ADC is powered and has correct reference voltage (check AVCC, VEREF-, VEREF+).
 
The following users thanked this post: Multiple Cheese Slices

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #15 on: July 04, 2021, 10:30:44 pm »
If you are a patient person, you could actually do a full list of every MCU pin voltage/signal (1, 2, 3 ... and so on), except maybe LCD SEGx and COMx pins. Better do with scope, so you don't confuse varying signals with some average voltage. Use circuit ground as ground. If you would do such a list then forum members can think about it. Also it could help others too, who will find this thread later looking to solve their problem.
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #16 on: July 05, 2021, 12:59:08 pm »
Hello!

Before I move on to checking each MCU pin one by one, I've decided to check the voltage regulator outputs. My ground reference was the battery (-).

I've found that the output of:
U8 is 0-250mV (is bad - the output seems to be random), expected 2.5V, as shown in the schematic. The input is 5V, as it should be. The output has no continuity to ground, so I don't think it's a short (?)

U6 is 3V3 (good)
U10 is 5V (also good)

I suppose this would indicate a faulty LT1790 (U8, 2.5V) voltage regulator. What are your thoughts?
Thanks in advance for your help
 

Offline giosif

  • Frequent Contributor
  • **
  • Posts: 750
  • Country: gb
Re: Fluke 87V displaying dashes only
« Reply #17 on: July 05, 2021, 01:21:50 pm »
Reading the data sheet for LT1790, it seems to be a voltage reference and not a voltage regulator.
Based on that, I expect the reference to be able to source much smaller current than a voltage regulator would (reading the data sheet, it appears it could source about 5-10 mA).
Where I'm going with this is maybe, even is there is no dead short on the output, the load of the circuit is high enough that it brings the output of the reference down.
Probably the quickest way to confirm if U8 is good is to de-solder it and check it out of circuit.

HTH
 
The following users thanked this post: Multiple Cheese Slices

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #18 on: July 05, 2021, 01:45:06 pm »
That is a bit ironic, because when you were checking rotary switch, you did not choose the circuit ground as reference. If you did choose it, then you would spot this power supply issue.

I aggree that it is likely LT1790 is just loaded too much. Might be also solder crack or something.
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #19 on: July 05, 2021, 02:17:25 pm »
I've tested the voltage reference by itself like you've suggested, giosif. When supplying it with 5V, the output is still just a few milivolts and varies, I suppose that's one issue found.

Also, yes, my bad about picking the wrong circuit ground earlier. I admit it, Manul  :D


I'll order a new one and see if the DMM works after replacing it. I'll update you as soon as I can, of course!

« Last Edit: July 05, 2021, 02:32:05 pm by Multiple Cheese Slices »
 

Offline Manul

  • Frequent Contributor
  • **
  • Posts: 754
  • Country: lt
Re: Fluke 87V displaying dashes only
« Reply #20 on: July 05, 2021, 04:41:07 pm »
Maybe that LT1790 is bad, but frankly speaking this is not a test, because there are no bypass capacitors and the behaviour is unpredictable. LT1790 hooked like this will likely oscillate like crazy in the region of Mhz, and not produce any meaningful output as a regulator.
 
The following users thanked this post: Multiple Cheese Slices

Offline giosif

  • Frequent Contributor
  • **
  • Posts: 750
  • Country: gb
Re: Fluke 87V displaying dashes only
« Reply #21 on: July 05, 2021, 06:56:08 pm »
Would you have a power supply capable of providing 2.5V limited at a few mA?
If so, maybe you can hook it up in place of the voltage reference and try powering on the meter.
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #22 on: July 06, 2021, 06:21:39 pm »
Giosif, I have a few LM317 's (adjustable linear voltage regulators) which could be set to 2.5V. However I have no way of limiting the maximum current they could supply to a range of few mA; Would hooking one up risk frying anything else? The brand new LT1790 voltage references should be arriving tomorrow if everything goes right, so if a regular voltage regulator shouldn't be in there for whatever reason, we won't wait long for the proper replacement.


Manul, I have re-done the tests with capacitors of capacities close to those shown here: https://www.analog.com/en/products/lt1790.html , and the output is still the same. Thanks for pointing that out tho!

 

Offline giosif

  • Frequent Contributor
  • **
  • Posts: 750
  • Country: gb
Re: Fluke 87V displaying dashes only
« Reply #23 on: July 06, 2021, 07:07:51 pm »
If the replacement part arrives tomorrow, I don't think it is worth the effort (and maybe some risk as well) of using an LM317.
 
The following users thanked this post: Multiple Cheese Slices

Offline Multiple Cheese Slices

  • Contributor
  • Posts: 29
  • Country: pl
Re: Fluke 87V displaying dashes only
« Reply #24 on: July 07, 2021, 06:26:28 pm »
Heya!

The replacement voltage reference has arrived! I have soldered it in, and another issue came up.

After setting the rotary switch from off to any other position, the LT1790 seems to be working properly, however only for (about (i don't have a scope atm to check the exact time)) 1/4 of a second.
Looking at the input voltage, U10 seems to be the cause - the 5V regulator powering the LT1790 (among other stuff) is only running for ~1/4 of a second before dropping it's voltage from 5V to ~0.6 - 0.4V. The voltage supplies to the ENABLE pin is constant 3V3.

Assembling the DMM back together and powering it on results in a "low battery" message, which makes sense to me, since I assume that this voltage regulator would be one of the first things to stop working with low supply voltage.

What are your thoughts? Should I solder the U10 out and test it by itself, or does that sound like an issue caused by something else?

There don't seem to be any shorts caused by my attachment of the new voltage reference. Also, U10 isn't getting hot (hot enough to boil IPA).  :-\

Here's a vid showing the behavior of U10: https://streamable.com/bgudk8
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf