Fluke 787 repair, any idea?

[kamcm]

Fluke 787 repair, any idea?

Symptom: 

1) When the 787 idle for a few days, It's able to measure low voltage correctly. (Both AC and DC)
2) After measured 110V ac or 100V DC. The 787 will not able to display a correct reading anymore. Sourcing 30V DC to 787 will display a reading of 28V
3) Then I try to measure 220V AC, the 787 display "OL". Sourcing a 30V DC to Fluke 787 again and the reading drop to 24V.


I realized the R63 was burned. I used another multimeter to measure the resistance of R63, the reading is 0.4ohm

I don't know what's the exact model of R63.
I tried to googling the service manual but no luck.

Please advice,

Thanks

[saturation]

Some thoughts.  The fault sounds like its in the input protection.  Idle for a few days improves performance suggests its the PTC.  The photo looks like a PTC, but I don't know the 787 PCB, but the input circuits of Fluke HH appear similar.

Try forcefully cooling the 787 with freeze spray to R63 or putting the whole unit in the freezer.  If it improves, it again points to something thermally resettable.  If its truly the PTC, Fluke PTC parts are typically 1 k 2 W but I don't know the exact part.  If you substitute it with any 1k 2W resistor and the meter functions correctly, you have the right part to replace.  Older Fluke's used a fusible resistor instead of a PTC, and you can use that until you find the right part.

Note if the PTC isn't the only thing faulted, look then also at the MOVs, as they tend to get damaged together by some issue.

[armandas]

Looking at the second image, the markins are DALE R301. If you look at this document, the code is for a 301mOHM resistor, which matches your measurement.

[retiredcaps]

I attached a picture of the 787 from a while back.

As saturation mentioned, the PTC should measure around 1k ohm.  The black colored PTC is labeled RT2 (right hand side of the picture).  Just below it is a green 1k ohm fusible resistor.  The blue MOVs are labeled RV1, RV2 and RV3.  These MOVs should measure 0L (open circuit) ohms.  All these measurements can be done on the 787 "in circuit".

The power resistor next to the green fusible resistor is likely a current limiting resistor.

PS. Are resistance measurements on the 787 accurate?  For example, does a 1k ohm resistor measure 1k?  How about a 10k ohm?

[retiredcaps]

If its truly the PTC, Fluke PTC parts are typically 1 k 2 W but I don't know the exact part.

A while back when doing research on the above, I think I found a suitable replacement, but no one stocks it?

YS4020. 1.1k ohm, 1000V rated.

http://datasheet.octopart.com/YS4020-Thermometrics-datasheet-11237.pdf

[retiredcaps]

Regarding the PTC, Dave's ohms overload #252 is potentially relevant to the OP's problem.

http://www.eevblog.com/2012/03/04/eevblog-252-multimeter-ohms-overload/

[kamcm]

Some thoughts.  The fault sounds like its in the input protection.  Idle for a few days improves performance suggests its the PTC.  The photo looks like a PTC, but I don't know the 787 PCB, but the input circuits of Fluke HH appear similar.

Try forcefully cooling the 787 with freeze spray to R63 or putting the whole unit in the freezer.  If it improves, it again points to something thermally resettable.  If its truly the PTC, Fluke PTC parts are typically 1 k 2 W but I don't know the exact part.  If you substitute it with any 1k 2W resistor and the meter functions correctly, you have the right part to replace.  Older Fluke's used a fusible resistor instead of a PTC, and you can use that until you find the right part.

Note if the PTC isn't the only thing faulted, look then also at the MOVs, as they tend to get damaged together by some issue.

Looking at the second image, the markins are DALE R301. If you look at this document, the code is for a 301mOHM resistor, which matches your measurement.

I attached a picture of the 787 from a while back.

As saturation mentioned, the PTC should measure around 1k ohm.  The black colored PTC is labeled RT2 (right hand side of the picture).  Just below it is a green 1k ohm fusible resistor.  The blue MOVs are labeled RV1, RV2 and RV3.  These MOVs should measure 0L (open circuit) ohms.  All these measurements can be done on the 787 "in circuit".

The power resistor next to the green fusible resistor is likely a current limiting resistor.

PS. Are resistance measurements on the 787 accurate?  For example, does a 1k ohm resistor measure 1k?  How about a 10k ohm?

Thanks for helping me,

I did some measurement and here is the result.

The black PTC : 1.146k ohm
Green fusible resistor: 1.011k ohm
Brown color current limiting resistor: 0.910M ohm
Blue Mosfets: All of them are fine, OL ohm

-------------------
Something really odd happened when I used the 787 to measuring resistance

I used Fluke 702 sourcing 2ohm to Fluke787 and the reading is 0.3ohm
Sourcing 0 ohm to Fluke787 and the reading is -1.6ohm
Sourcing 10k ohm to Fluke787 and the reading is 9.76k ohm
Sourcing 1k ohm to Fluke787 and the reading is 1.004k ohm


Finally, I used fluke 1857 to sourcing 50V DC to 787. I don't know why but the sourcing voltage drop to 17V and 787 showing a reading of 12.3V I used a fluke 179 to verified the fluke 1857 is working fine.


I used Flir E4 (hacked to E8) to inspect the pcb at the time fluke 1857 sourcing 50V dc to 787. Nothing getting warm!


Please advice,

Thanks

[retiredcaps]

Finally, I used fluke 1857 to sourcing 50V DC to 787. I don't know why but the sourcing voltage drop to 17V and 787 showing a reading of 12.3V I used a fluke 179 to verified the fluke 1857 is working fine.

Can you use the Fluke 179 to measure the input impedance of the 787 (set to DCV mode)? I'm going to guess that it should be around 10M ohm or 11.11M ohm?

I ask because the 1587 sourcing 50V to the 787 shows 0.01M ohm or 10k ohm where as it shows 10.1M ohm for the 179.

[kamcm]

Finally, I used fluke 1857 to sourcing 50V DC to 787. I don't know why but the sourcing voltage drop to 17V and 787 showing a reading of 12.3V I used a fluke 179 to verified the fluke 1857 is working fine.

Can you use the Fluke 179 to measure the input impedance of the 787 (set to DCV mode)? I'm going to guess that it should be around 10M ohm or 11.11M ohm?

I ask because the 1587 sourcing 50V to the 787 shows 0.01M ohm or 10k ohm where as it shows 10.1M ohm for the 179.

When I used Fluke 179 to measure the input impedance of the 787, the reading is 11.24k ohm.
When I used Fluke 179 to measure the input impedance of a 87V, the reading is  11.11M ohm.

But I still have no idea which component make this happened

Any advice?

Thanks

[retiredcaps]

When I used Fluke 179 to measure the input impedance of the 787, the reading is 11.24k ohm.

Well, there is something definitely wrong with the 787 since it only shows 11.24k ohm.

Take apart your 787 and post clear focused pictures of both sides of the pcb.  Maybe we can spot something like carbon tracks like this blog by modemhead?

http://mrmodemhead.com/blog/ideal-carbon-footprint/

Can you retest those MOVs again using the Fluke 179?  Again, they should all read 0L (open circuit) ohms.

[kamcm]

When I used Fluke 179 to measure the input impedance of the 787, the reading is 11.24k ohm.

Well, there is something definitely wrong with the input protection circuitry on the 787 since it only shows 11.24k ohm.

Take apart your 787 and post clear focused pictures of both sides of the pcb.  Maybe we can spot something like carbon tracks like this blog by modemhead?

http://mrmodemhead.com/blog/ideal-carbon-footprint/

Can you retest those MOVs again using the Fluke 179?  Again, they should all read 0L (open circuit) ohms.

Thanks, I just take it apart.

I retested those three blue movs with 179, they are fine.

[retiredcaps]

Another set of tests.

Can you measure the input impedance of the 787 for the 4V, 40V, 400V and 1000V DC ranges?  That is, we are taking the 787 out of autorange mode into manual range for each measurement. 

[retiredcaps]

Thanks, I just take it apart.

Can we get a picture with the black rotary knob removed so we can see the pcb underneath it?

[retiredcaps]

Can you measure the input impedance of the 787 when it is set to millivolt DCV?  It should be around 10M ohm.

[kamcm]

Thanks, I just take it apart.

Can we get a picture with the black rotary knob removed so we can see the pcb underneath it?

BINGO , you're right

Something really bad happened under the black rotary knob

It seem some kind of overload made it happened...  :bullshit:Is it blow away the trace?  Is it still repairable?

I just clean it with some alcohol pad. Then, I measure the input impedance when the black rotary knob is point to DCV, It's 11.20M ohm. 

[retiredcaps]

Something really bad happened under the black rotary knob

Did you lend out this meter to someone?

It seem some kind of overload made it happened...  :bullshit:Is it blow away the trace?  Is it still repairable?

Can we get another picture of the damaged area up close?  Both sides?

Do the DCV readings work now after cleaning it with alcohol?

BTW, while you are taking pictures, for yourself and others in the future, can we get some high resolution photos on the board in sections?  This will help you and others identify components on the pcb since there is no schematic.  Maybe take both sides of the pcb in 3 separate sections so we can see the writing on the smd chips.  If possible, try to get over 2000x2000 resolution for each section?  You can do this part later on.  There is no rush for this.

[retiredcaps]

Can we get a picture with the black rotary knob removed so we can see the pcb underneath it?

BINGO , you're right

Modemhead gets the credit here.  I just remembered that carbon blog problem and after you measured the MOVs twice, I thought the rotary knob might be hiding an oops moment.

[kamcm]

Something really bad happened under the black rotary knob

Did you lend out this meter to someone?

It seem some kind of overload made it happened...  :bullshit:Is it blow away the trace?  Is it still repairable?

Can we get another picture of the damaged area up close?  Both sides?

Do the DCV readings work now after cleaning it with alcohol?

BTW, while you are taking pictures, for yourself and others in the future, can we get some high resolution photos on the board in sections?  This will help you and others identify components on the pcb since there is no schematic.  Maybe take both sides of the pcb in 3 separate sections so we can see the writing on the smd chips.  If possible, try to get over 2000x2000 resolution for each section?  You can do this part later on.  There is no rush for this.

I will never lend out  my meter

=====
I did some testing again.

The ACV reading: Correct
The DCV reading: Correct
The DC mV reading: Correct
The ohm reading: OL


After the cleaning. the voltage reading is correct, but it will showing up OL when it should be 0 ohm.
I cleaned the carbon residue and I killed it  

I took it apart again and took some high resolution image for the damage area.
But I am not able to upload it to the forum. "Your file is too large. The maximum attachment size allowed is 2000 KB."
I will upload it elsewhere.

Thanks,

[kamcm]

High resolution image for the damage area

[retiredcaps]

I have a busy day tomorrow so I will have to look at this tomorrow PM.  I'm off to bed soon.

I will have some suggestions and thoughts.  In the meantime, you may want to look at some of the solder repair videos at

https://www.youtube.com/user/paceworldwide/videos

I can't remember which one or ones is the repair series?

[Smith]

You might want to get rid of some of the carbon deposits between these two points first. Some delicate Dremel work, or using a sharp knife or something like that should do the job. Maybe after that some reconstruction using silver paint?

[kamcm]

You might want to get rid of some of the carbon deposits between these two points first. Some delicate Dremel work, or using a sharp knife or something like that should do the job. Maybe after that some reconstruction using silver paint?

yeah, I will use Dremel to clean the remaining carbon residue.

But I had no idea how to reconstruct it. What materials should be use?


I had a silver paint pan like this one. But I knew the silver paint is in an extremely thin layer.
I am afraid it's not hard enough to undertake the rotation of the switch and the silver paint will  fall to pieces finally.

Please advice,

Thanks

[SeanB]

Grind out the carbon, which will leave a depression in the board. Then fill in the hole with some 2 part white epoxy and level it off. Then you wait for it to cure, and cut some thin brass shimstock ( available from hobby stores) to match the pads. Then grind a depression the thickness of the shimstock and use some epoxy to stick them down, holding flat with some waxed paper and a block to make it flat. Polish the new contact area with 1200 grit water paper and finally some brass polish to get it smooth. Contacts will have to be a thin tab going through the board ( make sure there are no traces under or in an inner layer)  and connected with jumpers the other side.

With a little care it will work as well as the original. If you can find some gold plated PCB material larger than the contacts you can simply use paint stripper ( and soak for a week) to strip the board from under the thin copper and use that instead of the shimstock.

[retiredcaps]

The ohm reading: OL

After the cleaning. the voltage reading is correct, but it will showing up OL when it should be 0 ohm.

Hmm, it is strange that it shows "Auto" ranging mode, but it shows 0L k ohm.  Since the 787 measures all the way to 40M ohm, I would have expected the annunciator to show M ohm, not k ohm?

I attached two old pictures of the rotary switch to give you and idea of the size, shape and location of the pads.  They are obviously not to scale, but if you manipulate the picture, you might be able to get a very close approximation.

[kamcm]

Grind out the carbon, which will leave a depression in the board. Then fill in the hole with some 2 part white epoxy and level it off. Then you wait for it to cure, and cut some thin brass shimstock ( available from hobby stores) to match the pads. Then grind a depression the thickness of the shimstock and use some epoxy to stick them down, holding flat with some waxed paper and a block to make it flat. Polish the new contact area with 1200 grit water paper and finally some brass polish to get it smooth. Contacts will have to be a thin tab going through the board ( make sure there are no traces under or in an inner layer)  and connected with jumpers the other side.

With a little care it will work as well as the original. If you can find some gold plated PCB material larger than the contacts you can simply use paint stripper ( and soak for a week) to strip the board from under the thin copper and use that instead of the shimstock.

The ohm reading: OL

After the cleaning. the voltage reading is correct, but it will showing up OL when it should be 0 ohm.

Hmm, it is strange that it shows "Auto" ranging mode, but it shows 0L k ohm.  Since the 787 measures all the way to 40M ohm, I would have expected the annunciator to show M ohm, not k ohm?

I attached two old pictures of the rotary switch to give you and idea of the size, shape and location of the pads.  They are obviously not to scale, but if you manipulate the picture, you might be able to get a very close approximation.

Thanks for your help,

It is not a easy task. Anyway, I will try it out when I have time.