Fluke 87-I repair

[staze]

All,

Got a Fluke 87 off fleabay for $20 as broken. Indeed, on firing it up, it behaves rather strangely.

So, knowing MrModemhead did some work with these, I immediately tested the 1K fusible resistor. Indeed, it's open. Great, so wired a 1K 1% 1/4w resistor in across it (with some leads), and now the ohms and diode read correctly. But, the ACV and DCV both read all over the place. ACV reads random mv's, and DCV reads about -34V. DCmV reads correctly. Pretty sure current measurements are correct too, but haven't checked. Range switch test checks out as well.

Interestingly, ACV reads random on all ranges (about -80mV to -100mV on 400mV range, about the same on the 4V range, .33V on 40V range, slowly decreasing to 0.5V on the 400V range, and 4V on the 4000V range (last couple seem to just slowly be decreasing to 0V)), and does nothing when you short the leads together. DCV reads -34.xx V on the 40V range, 0V on the 400V and 4000V range. OL on the 4V range. Makes me wonder if C1 might be bad (though, that shouldn't have anything to do with DCV since it's shorted out for those readings). RT1 seems to be good, as it's at 1.48K. I do only get 850K out of R2 (that's in circuit).

I'm really hoping there isn't something blown with the resistor network. I'll fully admit, I haven't tested the power supply(s) yet, but that's mainly because I haven't mapped out where the test points are yet (the service manual for these is great, albeit, a bit difficult to jump back and forth on).

Was mainly hoping someone had an idea to point at. Obviously if that fusible resistor is bad, then someone shocked this thing. It reminds me a bit of high input impedance, but it would zero out if the leads were shorted, and that isn't happening. Guess it also reminds me a bit of a digital issue (bad logic signals), but haven't gotten that far.

So, anyone seen anything like this? got any ideas? I thought at first the spark gap (E1) (guessing that's what it is) was bad since it had a big split in it, but that seems to be normal from pictures I've seen. =P Big hope is I don't have to order a replacement R1 until I figure out the issue, and can order whatever else at the same time. =)

Thanks!

[staze]

I just noticed R29 looks blown. There's some black residue around it, and the numbers are partially gone. It's reading 176K in circuit. Also looks like there might be some scorching around the solder blob between R1 and RT1. Looks like it arched toward the trace going to pin 1 (I think) of Z1, which is probably what blew out R29. I can see R29 in the schematics, and it's definitely in the AC and DC volts path, but it seems to also be in the millivolts, so I'm not sure why it wouldn't matter there (other than maybe 100K and 170K look pretty similar when you're only dealing with 400mV).

Thoughts?

[retiredcaps]

Got a Fluke 87 off fleabay for $20 as broken. Indeed, on firing it up, it behaves rather strangely.

I was going to bid on that and then it disappeared.  At least I know who now. 

I'm really hoping there isn't something blown with the resistor network.

That should be easy to measure to verify its values.

I thought at first the spark gap (E1) (guessing that's what it is) was bad since it had a big split in it, but that seems to be normal from pictures I've seen.

Yes, there should be a split in the spark gap.

Did you check the varistors (RV1 and RV2)?  They should measure infinite resistance with another multimeter.  You do the measurement in circuit.  A bad varistor will measure shorted.

[retiredcaps]

Thoughts?

A clear focused picture will help.

Black residue might be carbon which conducts.  From modemhead's site

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

[retiredcaps]

Another thing to check.

Put the 87 into DCV.  With another multimeter, measure the input impedance.  You should get around 11.11Mohm.  The input impedance for the mV range should be around 10.00Mohm.

[staze]

Did you check the varistors (RV1 and RV2)?  They should measure infinite resistance with another multimeter.  You do the measurement in circuit.  A bad varistor will measure shorted.

RV1 in circuit shows infinite, RV2 (one that bridges the cutout) shows 100meg, decreasing. Actually, reversing the leads results in infinite... testing RV1 with reversed leads results in a non-infinite reading as well. Cap in series?

Pictures of R29 attached, as well as the arch marking on the back where RT1 and R1 connect.

[staze]

Another thing to check.

Put the 87 into DCV.  With another multimeter, measure the input impedance.  You should get around 11.11Mohm.  The input impedance for the mV range should be around 10.00Mohm.

DCV I get 8.74MR ±4KR (fluctuating). DCmV I get 9.945MR ± 1KR.

[retiredcaps]

I just noticed R29 looks blown.  It's reading 176K in circuit.

My Fluke 87 R29 reads 99.5kohm in circuit.

[retiredcaps]

Pictures of R29 attached, as well as the arch marking on the back where RT1 and R1 connect.

I would clean up those areas with some IPA and scrub away all that black residue.

[retiredcaps]

Another thing to check.

Put the 87 into DCV.  With another multimeter, measure the input impedance.  You should get around 11.11Mohm.  The input impedance for the mV range should be around 10.00Mohm.

DCV I get 8.74MR ±4KR (fluctuating). DCmV I get 9.945MR ± 1KR.

I get a rock solid 11.11M and 10.00M ohm.

[staze]

Hmm... so resistor network, or those varistors, or.... Could you check those, and see if you get infinite in both directions?

Waiting for the end of the daily show before I clean up that soot. But I would think if it were conducting, I'd get lower rather than higher resistance.

[staze]

yup, no difference.

Okay, so, RV1, RV2, R29, or the resistor network. I'm guessing Pin 1 of the network is the one closest to the input jacks (all by itself). So, which pins should I be measuring between for what values? I'm not sure the service manual is very clear on this (it says pin 1, but not pin 1 to what).

Though, guessing RV1 and RV2 are fine if they're in the 100+ megaohm range. I don't have a SMT replacement for R29, but I could probably colder on a replacement axial one. The part I don't get is the changing value... open resistor acting as a capacitor?

Oh, and what does R2 measure in circuit for you?

[staze]

Pictures of R29 attached, as well as the arch marking on the back where RT1 and R1 connect.

I would clean up those areas with some IPA and scrub away all that black residue.

Done, no change. There's actually a physical divot in the PCB on the backside where that arc occurred. =/ 

[staze]

okay, something is clearly wacko... measuring R2 one way, I get 850K, the other way (leads reversed), I get 1.6M. This is with the unit off. Something is clearly acting as a cap, or I'm being dumb about this...

[staze]

So, on a whim, I set my second meter (right now, a Fluke 189) to voltage, and interestingly, I'm getting a voltage out of the 87 when it's set to DCV or DCmV proportional to the reading on the 87. In the case of the DCV, it's about -170mV when the 87 shows about -34V. And -7mV when the 87 shows -7mV. So maybe those resistors are all fine, I'm just getting voltage leaking into the circuit from somewhere and throwing off my resistance readings. Also explains the polarity issue.

So... where would voltage be leaking in from? My thinking would be in section 2-4 of the manual... the voltage clamping network of Q1, Q2, Q6, CR7, CR8, and R58. Maybe one of those transistors shorted from an overvolt? (someone really blew the ass outta this thing)... but the MOVs didn't blow... RT1 did... so... what the hell?

[retiredcaps]

Hmm... so resistor network, or those varistors, or.... Could you check those, and see if you get infinite in both directions?

As you probably guessed, after I took the photo, I went to sleep.

I have to run a bunch of errands today, but I took some quick measurements using a Fluke 187 (500Mohm capacity).  I can take more measurements later tonight and read your replies more fully.

For now (all readings in circuit - both directions),

R2 = 909k
E1 = infinite
RV1 = infinite
RV2 = infinite
R29 = 99.5k
R1 = 0.9971k

Z1 = to come later, but a quick measurement shows they all match values shown in Figure 5-3 (I didn't measure the 9.996M as it wasn't "obvious" to me which pins to measure)

Since your RV1 and RV2 are reading in the 100Mohm area, I suggest you remove (desolder) RV1 and RV2 and see if that fixes the problem.  If you are correct about the 87 getting blown out, it is likely the varistors are defective even though they don't measure shorted.

[staze]

Cool. I'll remove those MOVs and try that. I did clean everything up, and not seeing anything indicating the damage to the PCB is causing issues. R29 still looks physically damaged.

One other question... could you take a picture of the top and bottom rotary switch in the "off" position? I'm not 100% sure I didn't screw something up there. =P

Those MOV's shouldn't be leading to voltage on the input side of things, would they? Anyway, at this point, guess I'm looking at at least replacing R1, and R29. But also figure it's RV1 and RV2.

[retiredcaps]

R29 still looks physically damaged.

Since my R29 measures close to the 100kohm in the schematic, yours is definitely bad.

Those MOV's shouldn't be leading to voltage on the input side of things, would they? Anyway, at this point, guess I'm looking at at least replacing R1, and R29. But also figure it's RV1 and RV2.

Here is a thread where bad varistors was causing Robomeds to have incorrect voltage readings.

https://www.eevblog.com/forum/testgear/fluke-87-v-mystery-low-impedance-input/

PS. Did you measure E1?  If you did, I do not see the measured value?

[retiredcaps]

Z1 = to come later,

Pin Z1 is closest to input jacks.  Measured with Fluke, unknown calibration status, 187 and in both directions. Figures in the brackets are from Figure 5-3 of service manual.

Z1 to Z2 = 10.012M (9.996)
Z3 to Z7 = 1.1117M (1.1111)
Z4 to Z7 = 101.11k (101.01)
Z5 to Z7 = 10.014k (10.010)
Z6 to Z7 = 1.0008k (1.0001)

edit: typos

[staze]

Was this with the 87 on or off?

[retiredcaps]

could you take a picture of the top and bottom rotary switch in the "off" position?

The sun is gone here so I can't use natural light and get a clear focused shot. This is the best I can do with a flash.

The top and bottom pic is shot at angle to get rid of the flash.

[retiredcaps]

Was this with the 87 on or off?

All readings I have taken is with the 87 off with no battery installed.  Rotary position is in the off position.

[staze]

Perfect! I'll check my switch, and those values when I get home. Maybe also remove those MOVs and check with them out.

Also printed up the schematics today on 11x17 (A3) so I can take notes on values. Will post back later tonight. Thanks!

[staze]

Okay, checking Z1:

Z1 to Z2: 9.984M.
Z3 to Z7: 1.1091M
Z4 to Z7: 100.87K
Z5 to Z7: 9.994K
Z6 to Z7: 0.9986K

This is with, as you, an unknown cal 189 with switch off, and no battery. So, Z1 seems to be okay, if a little lower than it should be.

Also with the unit off, and no battery, the two MOVs register as OL both directions (getting somewhere).

E1 also open both ways. R2 still shows as 0.8517M (which seems wrong), and R29 definitely has physical damage... just going to be a bitch to replace with Z1 in the way. C1 checks out (just incase it was shorted). C20, on the other hand, in circuit appears to be bad, and I might spy some cracking around a leg (which may or may not be normal for a ceramic cap of this age). Could you test C20 in circuit and let me know what you get? I get about 0.136nF. I'll lift if I need to, but curious what it shows for you.

At least I'm fairly confident Z1 is okay at this point. =) So, one thing down. Leakage could be coming from one of the clamping transistors (Q1, Q2, Q6)...

Interestingly, it is drawing a very small amount of current when turned off... about 9.80uA. Not sure if that's abnormal or not, but from what I read on Mr Modemhead's site, it should draw nothing when off. So... leakage somewhere?

[lowimpedance]

Interestingly, it is drawing a very small amount of current when turned off... about 9.80uA. Not sure if that's abnormal or not, but from what I read on Mr Modemhead's site, it should draw nothing when off. So... leakage somewhere?

Give the PCB another good clean with IPA and check again.

[retiredcaps]

Also with the unit off, and no battery, the two MOVs register as OL both directions (getting somewhere).

I take the battery out because I might commit an oops and short something out.  It looks like removing the battery made change in the readings of the varistors.

R2 still shows as 0.8517M (which seems wrong),

My R2 definitely reads 909K in circuit (both ways).  R2 is supposed to be +/- 1%.

Could you test C20 in circuit and let me know what you get? I get about 0.136nF.

I get 0.115nF (no relative mode used) in circuit both ways.

Interestingly, it is drawing a very small amount of current when turned off... about 9.80uA.

This is consistent with modemhead's finding on the Fluke 83 (8.85uA)

http://mrmodemhead.com/blog/fluke-83-dmm-repair/

Same for this Fluke 87 post (at very bottom - 7.5uA)

http://www.edaboard.co.uk/fluke-87-flat-batteries-t268153.html

My own test also shows 7.5uA when in the off position.

[retiredcaps]

Give the PCB another good clean with IPA and check again.

Modemhead doesn't always document all his repairs, but I have firsthand knowledge that he sometimes gives multiple IPA baths to clean the PCB.  Sometimes a stiff clean toothbrush is required and lots of drying time.

I, too, have sometimes needed to give pcbs multiple IPA baths to get rid of contamination that was causing leakage current.

[staze]

okay, just cleaned it again, and again... same result. ~9.8uA when switched off. Interestingly, and maybe I'm being oblivious here, but I'm seeing 9V between the negative side of the battery (VDD/GND?) and the COM jack when the unit is switched off. That, to me, seems... wrong.

[staze]

Modemhead doesn't always document all his repairs, but I have firsthand knowledge that he sometimes gives multiple IPA baths to clean the PCB.  Sometimes a stiff clean toothbrush is required and lots of drying time.

I, too, have sometimes needed to give pcbs multiple IPA baths to get rid of contamination that was causing leakage current.

Okay, I'll give it another one tomorrow... but yeah, obviously something is leaking somewhere. Either that, or the switch is toast...

[retiredcaps]

Okay, I'll give it another one tomorrow... but yeah, obviously something is leaking somewhere. Either that, or the switch is toast...

If you read the Fluke 83 repair blog, it goes into great detail how the code switch works including diagrams on how to take the code switch off and check for contamination.

[retiredcaps]

Just a couple of other measurement suggestions?

1) What is the resistance between Z1 pin 1 and the left side of R29?  Mine is 0.2 ohms.  I'm curious if the pcb track is damaged?

2) What is the resistance of R2 out of circuit?

[staze]

I got the back switch all clean... I don't see anything dirty anymore (there wasn't really before). I'll just have to dunk the whole thing again tomorrow (once I pick up another bottle of 99% IPA) and let it soak a bit. But from what I see in the description for the 83, it sounds like when the switch is off, it's just shorting Vss to ground... so shouldn't there be SOME draw when off (only 9uA, or about 1/100th of what it draws when on).

That said, Vss and Vdd appear to be out of spec when the unit is on... which is probably due to whatever has failed. But I really don't know at this point.

[retiredcaps]

I'll just have to dunk the whole thing again tomorrow (once I pick up another bottle of 99% IPA) and let it soak a bit.

A bit of a long shot, but I also take a sharp multimeter probe or dental pick and carefully drag it between the pins of the main IC during the IPA bath.  The alternative is to use a stiff toothbrush and give the pins a good scrub.  After it dries thoroughly, I then inspect it with a 10x jewelers loupe.

Right now, I'm out of ideas, but if you are taking voltage readings, please post them so we can see.

[staze]

Just a couple of other measurement suggestions?

1) What is the resistance between Z1 pin 1 and the left side of R29?  Mine is 0.2 ohms.  I'm curious if the pcb track is damaged?

2) What is the resistance of R2 out of circuit?

1. 0.03R after zero'ing out the leads/probes.
2. 850K. Damn. Means I have to find a replacement. Gar.

So just tested Vss and Vdd again. Vss is perfect at -3.20V. Vdd is at 2.99V. So, Those are dandy. So it's not power supply, seemingly. Vref checks out at 1.2324V. VBT+ is 3.5948, VBT- is -5.794V. So I think all the various PS measurements are right on. My leakage is somewhere else, it would seem.

Doing the calculation, to get only 9.805uA of leakage (on a brand new, 9V battery at 9.473V) there's a good >1Mohm resistance the battery is being shorted through...

[staze]

Scrubbed... nothing, no change. And all the pins look good.

If you have the unit apart, could you tell me what R61 measures as (in circuit)? I ask because the reflow process for this meter and that resistor appears to have gone wonky, since it's at a diagonal. Probably not a real issue, but it has drawn my attention the whole time. It's supposed to be a 2.2M resistor (and that's what it's marked as (225), but in circuit, it measures only 7.2K, and it's the only SMT component that's crooked.

Thanks!

[retiredcaps]

If you have the unit apart, could you tell me what R61 measures as (in circuit)?

I see R61 on modemhead's picture here

http://mrmodemhead.com/blog/wp-content/gallery/f87_00/F87FR_009.JPG

but my board doesn't have R61.

My board is Rev H.

[staze]

lol! Crooked! Weird... it's like the pick and place for that one was misaligned.

My board is rev P in sharpie on the back, or rev K by the "silk screen" (looks more like solder mask covered trace).

I get about 1Meg between the battery terminals in the off position. Do you get OL/Open?

[retiredcaps]

lol! Crooked! Weird... it's like the pick and place for that one was misaligned.

I can open another 87 to take a look, but it is getting late so I don't want to make any mistakes.

I get about 1Meg between the battery terminals in the off position. Do you get OL/Open?

I get around 1.04Mohm.

This will be my last post (early AM).  I can take more tomorrow.

[staze]

well, then the 9.6uA when the unit is off makes sense. =P

At this point, I have a bad R1, bad R2, and a bad R29. Doesn't explain the voltage on the front end, but I actually haven't checked that since I removed those parts (will tomorrow).

What timezone you in (I see Canada, but not haven't looked where). =)

[PedroDaGr8]

Scrubbed... nothing, no change. And all the pins look good.

If you have the unit apart, could you tell me what R61 measures as (in circuit)? I ask because the reflow process for this meter and that resistor appears to have gone wonky, since it's at a diagonal. Probably not a real issue, but it has drawn my attention the whole time. It's supposed to be a 2.2M resistor (and that's what it's marked as (225), but in circuit, it measures only 7.2K, and it's the only SMT component that's crooked.

Thanks!

I just tore apart my working 87-I and the only resistor I could find marked 225 near u6. For some reason, I don't see r61 on the board diagrams in the service manual pdf I have. Anyways, when I measure across it in circuit (no battery, set to off) I get 8.554K.

Sent from my EVO using Tapatalk

[SoundTech-LG]

Don't know if this helps at all, but on my 79II , when I removed R1 (1K fusible) there is a trace connected to the via on the inner layer that leads to RT1 that broke off. Easy to happen as R1 comes out of the board. Might be good to check. Good luck.

[staze]

I'll check that.... but from what I could tell looking at the 87-I, it only seems like it's a front/back board. It's thick enough to be a multi-layer, but I don't see any evidence of traces on the inside of the board when holding it to a light... =/

[staze]

Checked with those components removed, and there's no voltage on the inputs (which, makes sense since they're basically completely disconnected from the rest of the meter at this point).

So, one question that comes up is...

Anyone know where I can get this: GS-3-100-9093-F-LF (3W, 909K 1% 100ppm resistor). Digikey is out of stock. http://www.digikey.com/product-detail/en/GS-3-100-9093-F-LF/989-1215-1-ND/

Can't say I've ever dealt with them having to order something. Anyone have experience? Or should I just call/email Fluke and see if they can send me replacements? Can't imagine that'll be cheap.

For now, might just try to add some resistors together to get ~909K and see what I get. Though I would think all that would lead to would be way out calibration rather than anything really strange. R29 is just jellybean.

Also need a replacement Fusible... looks like Mouser carries a 1K 2W 5% fusible which should work just fine.

[Vgkid]

It looks like I will need to follow this thread, my meter reads accurately on vollts/ resistance, not so well on mv range. The Input impedance is 10M on the DCV range, and 3M+/- .4 on, with an open circuit on the AC scale.

[SoundTech-LG]

The other thing you can do is buy a parts only Fluke on ebay, and harvest what you need.

[staze]

True, but those parts can often be bad... burned out by overvolting the meter.

Really, this meter is probably a parts unit itself... =/

I did just email Fluke asking for pricing for R1 and R2... guess we'll see.

[staze]

Ouch, got back to me. Lead time for the parts is nearly 6 weeks, and about $20. =/

Might have to see if Digikey can get them.

[lowimpedance]

 For the 909K , if space permits you could try the 1M ohm gs-3 in parallel with this 10M (High voltage and pulse withstanding 1/2W)
 http://www.digikey.com/product-detail/en/HVR3700001005FR500/PPCHF10MCT-ND/720093.
Both are in stock at Digikey. Should get you close enough to the 1%.

Have you checked R12 (most likely fine, but..).

Bodging in close enough values to test the overall function and fault find further is the way to go until your confident the meter is fully operational then get the Fluke parts to finish off.
 But if something major has gone  (IC!) then well nothing much lost, and you have a parts unit and motivation to get another one 

[staze]

For the 909K , if space permits you could try the 1M ohm gs-3 in parallel with this 10M (High voltage and pulse withstanding 1/2W)
 http://www.digikey.com/product-detail/en/HVR3700001005FR500/PPCHF10MCT-ND/720093.
Both are in stock at Digikey. Should get you close enough to the 1%.

Have you checked R12 (most likely fine, but..).

Bodging in close enough values to test the overall function and fault find further is the way to go until your confident the meter is fully operational then get the Fluke parts to finish off.
 But if something major has gone  (IC!) then well nothing much lost, and you have a parts unit and motivation to get another one 

Yeah, think I'm just going to wack in some 1/4W resistors for now to get 900K or so (since that'll be within 1%). Then a 1K 1/4W for R1... and a 100K 1/4W for R29...

I haven't done the math (since, I'm not sure how to start), but it would be curious to see if the combination of R29 reading 170K, plus R2 reading 850K, and R1 being open, combined would lead to how far out the DC volts is. I'm kind of guessing it would explain it.

R12 I haven't checked. It's on the back, so I hadn't even tried. =P

[staze]

bodged some 1/4w 1% resistors in to replace R1, R2, and R29. No change at all. Voltage on the input jacks in the DCV setting seems to be coming from the R1 side of things (lower voltage on the input jack side of R1). I need sleep, but I think I'll hook the scope up tomorrow and see if it's just a constant voltage, or if there's something obvious about where it might be from.

One of the transistors seem like an obvious, or something still wrong with the rotary switch, but I've cleaned that thing so many times at this point...

Thanks!

[retiredcaps]

Can you retest the input impedance on the 87 again?  Please report the measurement for DC mV, 4VDC, 40VDC, 400VDC, and 1000V (go into manual mode and cycle through the ranges).

Can you also get the 87 to measure a 12V DC power supply?  Please report the measurement for each range.  For example, if it is working properly, it should be

4VDC = 0L
40VDC = 12.00
400VDC = 12.0
1000VDC = 12

[staze]

Well get these tonight.

A quick measurement of input impedance with no battery connected was 13.6M on the DCV range, and I think 4.3M on the DCmV range.

Will confirm tonight, and get the other measurements, with a battery installed. That said, since there's still voltage on the input jacks in DCV, I'm sure the input impedance measurements are going to be WAY off.

Hopefully I can get the LCD to cooperate as well... it started out perfectly, and now it's pretty bad. =/

[retiredcaps]

That said, since there's still voltage on the input jacks in DCV, I'm sure the input impedance measurements are going to be WAY off.

I think you will be correct, but I just wanted to see if there is a change depending on the range.

The zebra strips probably just need an IPA cleaning.

[staze]

That said, since there's still voltage on the input jacks in DCV, I'm sure the input impedance measurements are going to be WAY off.

I think you will be correct, but I just wanted to see if there is a change depending on the range.

The zebra strips probably just need an IPA cleaning.

Yup. And yeah, will get those tonight. Also see if I can figure out where that voltage is coming from...

[retiredcaps]

So... where would voltage be leaking in from? My thinking would be in section 2-4 of the manual... the voltage clamping network of Q1, Q2, Q6, CR7, CR8, and R58. Maybe one of those transistors shorted from an overvolt?

I can't read a schematic to save my life, but I wonder if removing R59 (510ohms) would allow the meter to function while eliminating the above as possible source of leakage?  R59 is connected to S1 rear contact 2 (section 2-7).

edit: added more detail.

[staze]

So... where would voltage be leaking in from? My thinking would be in section 2-4 of the manual... the voltage clamping network of Q1, Q2, Q6, CR7, CR8, and R58. Maybe one of those transistors shorted from an overvolt?

I can't read a schematic to save my life, but I wonder if removing R59 (510ohms) would allow the meter to function while eliminating the above as possible source of leakage?  R59 is connected to S1 rear contact 2 (section 2-7).

edit: added more detail.

I'll give that a shot... should also be easy enough to see if the voltage is higher or lower on each side of that resistor. Yay 5.5 and 6.5 digit Multimeters. =)

[retiredcaps]

R58 is connected to VSS which might also make another interesting measurement point?

[staze]

okay, had my hopes up that I found the issue when I saw some solder-like material between the guard and input traces on U4, but seem to have been nothing. =/

Anyway, the info you asked for.

Input Impedance:
DC mV: 9.9845M
4VDC: -1.45M-1.74M (all over the place)
40VDC: 6.6M (all over the place as well)
400VDC: 9.990M (rock solid) (87 even sees 1.0V coming from my 189's resistance measurement)
1000V: 9.980M (ditto).

So the issue would seem to be the 4VDC and 40VDC. Hmmm...

Measurements of 12V DC PSU... set it to 1mA so I didn't risk blowing out my wimpy 1/4W input resistors I hacked in there). PSU always read 0mA. Actual measured draw on each range is listed in parenthesis below.

4VDC = OL (1.275uA)
40VDC = -25.55V (all over the place) (1.223uA)
400VDC = 12.0 (1.195uA)
1000VDC = 12 (1.196uA)

Calculating from that, the input impedance for each range, 12VDC input, with the listed currents, would be 9.411M, 9.811M, 10.042M, and 10.033M, respectively.

The 40VDC range is interesting since with no input, the 87 reads -33.45VDC. One would think with 12V input, it should read -21.45VDC. Huh.

R58 and R59, however, both check out as far as resistance. There's a 1.58V drop across R58, and a 0V drop across R59 (which, I'm not sure I understand, but like you, I have a difficult time with schematics... especially since Fluke didn't actually label the parts on their schematics) in DCV mode. Reading through the service manual section on Voltage measurements... I'm either thinking U4 is bad, or they've way oversimplified things... Because somehow, Q1, Q2, Q6, Q13, and Q14 enter into the equation.

One thing that keeps me wondering... did you ever manage to measure C20 (capacitance, maybe ECR/Resistance as well if you can)? I do see a bit of voltage on it in DCV mode...

Alright, time for bed.

[ModemHead]

I'm sorry I haven't been able to participate in this thread due to circumstances from which I have a temporary reprieve today.  So I'll toss out a few quick thoughts.

The current draw when turned off is a red herring.  The 1M pull-up resistor R54 is directly across the battery terminals when the unit is off, which accounts for the draw.  In addition the Vdd/Vss pins for U6 is also across the battery permanently, but the resistor accounts for most all the current.

Looks like an interesting case of serious over-volting as you know.  I'm just a hobbyist so I can't say I've seen everything, but over-volting seems to mostly affect the fusible resistor (R1), the MOVs, the thermistor, and (unfortunately) the main chip in cases I've seen personally or read about, assuming you clean up the carbonized mess sometimes left.  Of course there's a first time for everything.  Wouldn't hurt to check all the clamping transistors. (the ones with the base and collector shorted)

In addition to PCB contamination, failures in the main chip will lead to current being sourced or sunk where it shouldn't be, wacky resistance readings, and voltage appearing at the input jacks.  Do not hesitate to isolate the lines into the main chip by finding a convenient place to cut a PCB trace.  It will save you tons of time scratching your head, and as long as you're careful, it's easy to re-connect with a solder bridge or bit of 30ga wire.  I have an 83 here that reads 100 ohms from its ohms sense input (P93) to ground, with the lead isolated from everything else and the power turned off.  It must be toasted!

Pay real close attention to the APV0, APV1, and APV2 lines.  You might break one at a time and see if the voltage on the meter input jacks disappears.  Maybe start with APV2, since it seems that the 40V range is the screwiest?

[retiredcaps]

okay, had my hopes up that I found the issue when I saw some solder-like material between the guard and input traces on U4, but seem to have been nothing. =/

Yes, sometimes it is not obvious at first, second or even third glance where contamination lies.  I put tough to solve problems away for a couple of days and then look at it again with "fresh" eyes.

So the issue would seem to be the 4VDC and 40VDC. Hmmm...

By measuring the input impedance of each range, it gives us more clues as to what is wrong.

One thing that keeps me wondering... did you ever manage to measure C20 (capacitance, maybe ECR/Resistance as well if you can)?

I did measure C20 back in post #25.  Unless I misunderstood?

[retiredcaps]

I have an 83 here that reads 100 ohms from its ohms sense input (P93) to ground, with the lead isolated from everything else and the power turned off. It must be toasted!

Is it sheer coincidence that staze's Fluke 87 and modemheads's recent 83 both have bad main ICs from the same ebay seller?  Maybe the seller has some quick method of testing the IC to know it is bad?

I'm not saying the seller is dodgy as both were clearly listed as "parts/as is" and the problems with each meter clearly disclosed.

[staze]

Wouldn't hurt to check all the clamping transistors. (the ones with the base and collector shorted)

Will check other stuff later, but looking at the clamping transistors, Q1 and Q2 both test 0.61V one way, and 2.4V the other, which I believe is normal. They're also directly connected to each other. Q6, however, reads 0.61V one way, and Open (OL) the other. I think that's what it should be... not sure why Q1 and Q2 read how they do. =/ CR7 and CR8 both read 1.23V in one direction, and 1.83V in the other. And I previously checked R58, and it's fine. That's the whole of the clamping network.

One question is, Modemhead, do you have a full understanding of which pins on Z1 to test against? In a previous post I listed resistances, but the manual makes a complete mess out of simply describing how it should work (it's like a sentence was left out in editing).

Also, am I right in thinking there are guard traces that run along either side of the APCC and APV0 pair? there are two traces next to U4 that are not connected (they look like guard traces since they run down to Z1 and around).

Thanks!

[staze]

Yes, sometimes it is not obvious at first, second or even third glance where contamination lies.  I put tough to solve problems away for a couple of days and then look at it again with "fresh" eyes.

One thing that keeps me wondering... did you ever manage to measure C20 (capacitance, maybe ECR/Resistance as well if you can)?

I did measure C20 back in post #25.  Unless I misunderstood?
[/quote]

Yes, I think I might take a break here after tonight, unless the suggestions keep coming!

And yes, I totally missed that when you posted it. Sounds pretty similar.

Thanks!

[staze]

<edit>

Nevermind, I'm an idiot. Of course there's voltage across there with the fuses out, that removes the "short" to COM... so, ignore the below. Though it is weird there's ALWAYS a voltage across the shunt rather than switching it on/off with the rotary.

</edit>

The plot thickens? In DCV mode, with the fuses removed, there is a whole 1.4931V between the 10A jack, and COM. Between the mA jack, and COM, there's 0.7804V (basically, half). All the while, there's still the -0.1840V (AVG) across the Volts and COM jack. But, maybe that's normal, since it looks like those jacks are tied to VDD through a voltage divider... Is that for the shunt?

[ModemHead]

The diode arrays should read two diode drops in the forward direction.  The clamp transistors should read one diode drop in the forward direction.  In reverse, they should be open-circuit, so that any reading is the just the result of other parallel impedances reacting to the stimulus current. Somewhat unpredictable.

Fluke left off the pin numbers for Z1 on the schematic, so measuring it would be somewhat trial and error.  It's actually a very sturdy component, so it wouldn't be my first suspect anyway, as long as there's no physical damage.

Yes there are guard traces around APV0, this is the very high-impedance analog-to-digital converter input.  Not really sure what the APCC input is for, but if it's inside the guard traces, it must be a high-impedance input also.

[staze]

Always really hesitant to cut traces. Any chance I would see voltage on that pin when there is no input?

What is the op-amp in the unit used for if pretty much everything is in U4? I thought at first it was the AC buffer, but it seems that's in U4 as well. :/

Out of curiosity, what resistance do you get between black "pads" on the PTF when the switch is "shorting" them? From what I gather, the CPU just sees a "bit" when the switch is moved, and changes function from that.

[staze]

so, reason(s) I ask.

1. There's a "hole" in the top shield right above that op-amp. Doesn't look melted... but could be (op-amp doesn't get hot, it seems). Though, looking at modemhead's pictures, I see the same hole on his... must be a molding defect. =/

2. Voltages on the op-amp seem... off. Pin 4 and 8 both are right for the rails, but given the inputs on pins 2 and 3 (few hundred milliamps), pin 1 is slammed to the negative rail (-3.2V), and weirder, pins 5 and 6 are the same as pins 2 and 3, yet pin 7 fluctuates pretty wildly. Looking at the schematic, pin 1 being slammed at the negative rail makes some sense since the diode on it's output would mean it's not getting any feedback... so maybe I don't quite understand how this isn't the expected behavior of this thing...

[staze]

On another note... CR5 (up by the power regulator circuitry in the upper right corner), reads 0.61V in both directions... Which, doesn't seem right. But I'm not sure how, if it was bad, it would cause any of this... since it seems to just be involved in the power rail generation, which as I said before, all check out.

That said, I do have a dual tracking PSU coming, incase I want to bypass the battery input completely... =/

[retiredcaps]

Out of curiosity, what resistance do you get between black "pads" on the PTF when the switch is "shorting" them?

I'm not sure what you are asking here?  Perhaps a photo?

[retiredcaps]

On another note... CR5 (up by the power regulator circuitry in the upper right corner), reads 0.61V in both directions...

I get the exact same reading of 0.61V.

[staze]

weird.

And yeah, I'll post some pictures. I am going to break out the good camera and Macro lens and take some real pictures.

[staze]

Alright, didn't get pictures up. But I ordered some REAL IPA (99.9% anhydrous). Going to clean the whole thing again... since I'm really hoping it's not the CPU (Fluke said they stopped making a replacement in 2006).

I go see a voltage between the APV2 pin and the V? connection. =/

[PedroDaGr8]

Alright, didn't get pictures up. But I ordered some REAL IPA (99.9% anhydrous). Going to clean the whole thing again... since I'm really hoping it's not the CPU (Fluke said they stopped making a replacement in 2006).

I go see a voltage between the APV2 pin and the V? connection. =/

Thats overkill mate, there was no need for anhydrous. Unless you plan on storing the IPA under molecular sieves and using a needle containing N2 to pull out the alcohol it will rapidly rehydrate by absorbing moisture from the air (the molecular sieves scavenge any excess water that finds its way in). That's the "beauty" of azeotropes, they will rapidly form at will. I know this from experience being a nanomaterials chemist. I use anhydrous stuff regularly and its a pain to work with if you need to KEEP it anhydrous.

Otherwise, good luck mate.

[staze]

Alright, didn't get pictures up. But I ordered some REAL IPA (99.9% anhydrous). Going to clean the whole thing again... since I'm really hoping it's not the CPU (Fluke said they stopped making a replacement in 2006).

I go see a voltage between the APV2 pin and the V? connection. =/

There was no need for anhydrous. Unless you plan on storing the IPA under molecular sieves and using a needle containing N2 to pull out the alcohol it will rapidly rehydrate by absorbing moisture from the air (the molecular sieves scavenge any excess water that finds its way in). That's the "beauty" of azeotropes, they will rapidly form at will.

Otherwise, good luck mate.

True. I guess I meant more than the drug store 99% stuff has god-knows-what in it to keep people from drinking it, as well as getting the water out. I bought actual MGChem 99.9% anhydrous just because that, theoretically, doesn't have benzene, etc in it. =) Now, I doubt that'll make any difference, but it'll at least not make me question if my IPA is leaving behind some residue. Of course, I probably could have gone over to chem-stores on campus and gotten the same 99.9% IPA... but I don't like to abuse that (they gave me enough looks when I asked for straight Ethylene Glycol, and some nice Tygon 5/8th-inch tubing  (for a computer water cooling system)). =) Though, storing something under something just makes me recall dealing with solid K or Li stored in/under kerosene.

[JoeS76]

Hi , how are yous - enjoying the forum- can i ask- have you checked the bridge rectifier- i've found that sometimes with readings like yours- the bridge is blown and needs replacement which is a simple enough fix

[staze]

No, I haven't. I hadn't really thought that could be an issue since I wouldn't think it would be in the signal path in DCV...

I'll give it a check tonight though, thanks!

[lowimpedance]

I think you have reached the point where you need to cut some traces as Modemhead suggested to see if the dc is coming from U4.
It looks likely that U4 is toast.

[staze]

I think you have reached the point where you need to cut some traces as Modemhead suggested to see if the dc is coming from U4.
It looks likely that U4 is toast.

or at least, part of it is. since the higher ranges for DCV work, as well as DCmV, Ohms, etc.

Yeah, I'll just have to bite the bullet and cut them. Just for completeness sake I think I'll pull one of the MOVs too just to make sure they're not bad it some weird way.

Guess the real question is, if U4 is bad, do I keep the meter and use the ranges that work, or do I turn around and sell it on eBay as a parts unit?

[staze]

So, pulled the MOVs, no diff (as expected).

I'll see about cutting that pin tonight. Is there any chance it's the op-amp (AR1)? I just can't figure why the outputs were pegged like they were... but I guess cut the pin if nothing changes?

ugh.

[staze]

Pay real close attention to the APV0, APV1, and APV2 lines.  You might break one at a time and see if the voltage on the meter input jacks disappears.  Maybe start with APV2, since it seems that the 40V range is the screwiest?

Cutting APV2 (96) resulted in just killing that range (the meter goes into auto-range loop, or just OL on all but 4000V), and still voltage on the jacks. Cutting APV0 (99) oddly killed the voltage on the input jacks, but still shows -35V on the meter's reading. Cutting APV1 (98) resulted in no change.

Anywhere else I might want to look before I call it a goner? To satisfy my own curiosity, I cut the VDD and VSS rails for AP1, and that make no difference.

So... is U4 toast?

[staze]

So, thought I've had. Would I gain anything by re-cutting APV0 (pin 99), then checking for current/voltage on each side of the cut? Or am I just paddling against the reality that U4 is toast?

[lowimpedance]

Its a goner, keep for the parts on the off chance you get another, or just move it on as a parts only.
You win some and loose some, look forward to the next repair job  .