Fluke 77-IV Copper Corrosion

White residue on 77-IV PCBThe fourth generation of the classic Fluke 77 shares the same basic design and circuit board as the rest of the 170-series (175, 177, 179).  Apparently a common problem with these circuit boards is the appearance of a white powdery residue that collects in certain areas.  A couple have passed my workbench, and I’ve heard reports of more.

The source of this residue is unknown to me.  Chances are, it is probably environmental, perhaps humidity.  What is clear however, is that the residue can be accompanied by the corrosion of parts of the copper PCB traces.  [Added later: best theory is water-soluble flux residue has reacted with humidity to create a corrosive salt.]

This 77-IV was advertised as working, but missing its backlight and continuity beeper functions.  This description was accurate.  It has been well-used and is very grimy.  Other than the overall dirty appearance, there is no real evidence for what sort of environmental conditions this meter has been used in.  Salty air?  High humidity?  Who knows…

Inspection of the PCB reveals various locations on the board with concentrations of the mystery residue.  The area around the piezo beeper contacts shows a particularly heavy concentration of the stuff.  There are other deposits, including some around the backlight LEDs.  None of these areas are near the battery compartment, therefore battery leakage is not suspected.

The usual procedure for PCB contamination is a good scrubbing with a toothbrush and 91% isopropyl alcohol (IPA).  After this procedure, the PCB is allowed to dry thoroughly.  This gets rid of the white residue.  Examining the entire PCB closely with a magnifying lens reveals only one area of permanent damage.  One of the copper traces connected to the backlight LED circuit has the appearance of being completely eroded away, breaking the connection.  This damage is repaired by scraping some of the solder mask away, applying flux, and soldering a tiny piece of 30AWG tinned wire over the break.

After the repair, all functions including the backlight and continuity beeper work.  A thorough clean-up and lens polishing also improves the appearance.

Epilog: The next unit I found with this copper corrosion problem was a 177, with the exact same PCB revision as this 77-IV.  Unfortunately I did not photo-document the repair procedure, but it was quite similar to one described here.  One of the pads for the backlight LEDs was eroded away causing the backlight not to work.  The meter was non-functional with some digital-domain fault that caused it to always display “- – – -“.  The trouble turned out to be a corroded via that was open from the top layer to the bottom, due to the plated-through hole being eroded away.  A wire patch soldered in brought the 177 back to life.

About modemhead

Fixologist and multimeter junkie.
This entry was posted in Repair and tagged , , , . Bookmark the permalink.

41 Responses to Fluke 77-IV Copper Corrosion

  1. Phillip Schneider says:

    Thanks for this guide I used this technique to repair a laptop motherboard that had corrosion build up from being in a humid climate for years! Without this I would have never known it was so easily repairable.

  2. LDSisHere says:

    Thanks for sharing this information. I checked my Fluke 77 IV when I first read this article and found the beginnings of this corrosion in a couple of places. Later when I bought my used Fluke 175 the first thing I did was take it apart and inspect it for this problem but I did not find anything.

    I currently have a bad Dell PowerEdge 1850 power supply that has this type of corrosion all over a circuit board that I intend to see if cleaning this stuff off will restore its’ functionality.

    • modemhead says:

      I’m relatively convinced it’s due to water-soluble flux residue. Time and place of manufacture would have a lot to do with whether it ends up being a problem or not. The key is to find it and clean it up before it erodes the copper. On the Flukes it seems to affect the beeper and backlight circuits, which just happen to be the parts that are powered directly from the 9V battery. So tiny leakage currents, even with the unit off, may be a factor as well.

  3. Yahya says:

    Hi, this is the exact same problem i have with my 77-4.
    The backlight doesn’t work.
    I have got the 4 screws out and have the front and back apart.

    But how do i get the inside out so i can see the backside of the motherboard, so i can clean it up. I don’t wanna force it, but it doesn’t wanna come out. Are the more screws?
    I tried finding a teardown manual, without success.

    Thanx

  4. Yahya says:

    Thank You! Very simple :))

  5. Yahya says:

    Will a alcohohol cleaning cloth do also to clean it up, which is used to clean the processor from paste residu?

    • modemhead says:

      It will help, but I prefer using an acid brush with the bristles cut down and/or an old toothbrush. To find eroded copper you need some physical scrubbing action. Use plenty of alcohol and some paper towels to blot up the run-off.

    • Yahya says:

      I did it with an alcohol cloth, but to be honest, there was almost no resideu at all. Everything look almost shiny new. I Cleaned it up, put it together, but still no backlight. I did a continuity test on the leds, no beep but they start burning a bit when i do that :), but when put together, no luck unfortunately.
      Any ideas are welcome. I am a starter in these things and this is my first quality meter to be honest, but i have high interest and wanna learn.

      Thanx

      • modemhead says:

        There’s usually some tell-tale residue if copper erosion is a problem, but it may take a magnifying instrument to find it if there is any. Maybe this is a different problem. I’m running out of time for today, but here’s a photo of a 177 (same circuit board as 77-IV) with some very nasty corrosion. Check carefully the SMD transistor and resistor that I have circled in yellow. That includes the copper traces connected to them.

  6. Yahya says:

    I did it :))) after the 3rd time i opened it, after very close inspection, i saw that one of the legs of the led was off the motherboard. I soldered it back and voila :))
    The first 2 times i screwed back the screen though th numbers wherent showing :$
    Then i screwed it harder and then it started working again. Thanx for thinking with me 😀

  7. Stefano says:

    Hi Modemhead, I am writing from Italy. I have this multimeter 77IV (circuit board Fluke 17X) which does not work. In the display I can see only this sign “- – – -“, no numbers. I see right all the measuring selection (VDC, VAC, mV ecc..), I can read “LEAD” when I am in the A selection, and at the start I can read 77-IV.
    I read your message about this question and I will check better for corrosion. Do you remenber where was the hole eroded from the top layer to the bottom in the circuit board ?
    Thank you for the answer. Stefano

    • modemhead says:

      Stefano, this symptom occurs when the microprocessor loses communication with the analog measurement chip, for any number of reasons that would include a broken PCB connection, or even if the analog section loses its power supply. It is quite unlikely that yours would have exactly the same broken connection as mine.

      Nonetheless, here is a picture of a 17x board, the via in question is circled in red. I was never able to find complete documentation on this circuit or all of the parts, but data sheets for the two voltage regulators marked on the picture are easily found. This should help to verify that the +/- 2.5V power supply for the analog measurement chip is OK. The voltages are with respect to the COM input jack.

      • Stefano says:

        Thanks so much Modemhead for reply, until now I was in my lab to check the circuit board. The PCB is clean without sign of corrosion and in very good condition, so I think it is not a copper corrosion. The board is the same in your picture: Fluke 17X-3011 REV 015
        I can see “—-” in every measurement ranges, except in frequency mode (Hz, pushing yellow button when you are in VDC), in this range I can read 0.00. Is it a useful clue ?
        Now with your info I try to verify if the analog circuit is good supplied.
        Stefano

  8. Stefano says:

    I checked the power supply for the analog measurement chip and it is OK, +/- 2,5 VDC
    PCB connection OK.

  9. Stefano says:

    The multimeter, without LCD display, has 3 mA current consumption (9V)

    • modemhead says:

      The fact that you get “0.00” in frequency mode does sound like a clue, but unfortunately I don’t know what to make of it. Measuring freq is a matter of counting pulses instead of A-to-D conversion. I once tried and failed to find data sheets for those Linear Tech chips which might provide answers.

      3mA sounds too high. The 77-IV I have here draws 1.11mA in VDC mode with the LCD on.

  10. Stefano says:

    Yes, I have 3-4 mA, without LCD display, in VDC mode. I think these chips are custom for Fluke. Do you have any reference or measurement of some pin ? I cannot proceed…
    Thanks anyway.

  11. Stefano says:

    The 24351 is a LTC2435: 2.7V to 5.5V micropower 20-bit differential Delta Sigma analog to digital converters with integrated oscillator. The SL10327 is Fluke.

  12. Guillermo Rodriguez says:

    Best regards. my new used Fluke 177 this was damaged VDR input protection, which already I replaced and this work perfectly, only in capacitance I could not make it work properly in capacitance, within the lower range only and I can not read more than 20 microfarads. is like no change the range automatically, I change the range manually and not does not provide any read this still reading non stop. Any sugestion. Thank You

    • modemhead says:

      The only thing I can think of is the possibility of one of the series protection components failing with a high resistance. This would have minimal effect on voltage measurements, but could decrease the amount of test current for the ohms, continuity, and capacitance functions. Check the big green 1K fusible resistor and PTC thermistor near the input jacks. The thermistor should be in the 1K to 1.5K range, and the resistor should obviously be 1K.

  13. GeekPriest says:

    Hmm. A newly arrived 77iv has the exact same issues yours had: no backlight, no continuity buzzer.

    After disassembly, I see some of the crud you mention, although mine’s not in the same spots. After cleaning, still no backlight, although the diodes themselves do light up when checked with another meter. Guess I’ll have to break out a magnifying glass.

    Incidentally, the 77iv is billed as a model designed with automotive mechanics in mind. That might explain the extra grime on the outside, plus exposure to high humidity environments. (Many shops aren’t climate controlled much, if any, of the year.)

    • GeekPriest says:

      I take that back. I have white stuff corrosion exactly where you had to put your wire to fix the broken trace. Following everything around, and this is also the positive source for the piezo buzzer. Now to hunt up the wire and warm up the soldering iron.

      • modemhead says:

        For some reason the backlight LED and piezo circuits seem to be most susceptible to corrosion damage on this model series (17x). Coincidentally, these two circuits are powered by the full battery voltage, everything else is regulated down. Good luck with your repairs.

      • GeekPriest says:

        Yes. It’s an odd failure mode for a Fluke, it seems to me. The meter in question tends to be used in humid environments, and we have two meters failing in exactly the same spot. This is a current generation meter, so I’ve wondered whether I shouldn’t just avail myself of the lifetime warranty.

  14. GeekPriest says:

    Clarifying my previous point, “avail” of the lifetime warranty means “see if the lifetime warranty still applies.”

    I’ll probably be happier just fixing this one myself.

  15. Bob Grieb says:

    Hi,
    I am trying to repair a 177 for a friend, with similar corrosion issues to those mentioned here. I have found and fixed a bunch of vias that were no longer making the connection from top to bottom. I only have batt + at one LED and pin 14 of the 4069. I can’t find where else it is supposed to connect. The meter won’t turn on and draws no current at all from 9V supply. Any help would be appreciated. Too bad there is no schem on line for this series of Fluke meters. Bob

    • modemhead says:

      Vb+ goes to the lowest leg of a SOT-23 device just above the large 1-ohm 400mA current shunt resistor and to the left of the shielded resistor network. I presume it’s a MOSFET or something having to do with automatic power-off. Yeah, no schematic is like working blind, and why I much prefer to work on older models.

  16. Bob Grieb says:

    Thanks. That got voltage to the lower transistor, but still nothing around the chips,
    then I found an eaten trace under the upper of those two transistors that carries the switched voltage. Fixed that, and now have 3.3V and – – – – on the display. No voltages over on the analog chips, so I jumpered the switched batt+ after that lower transistor to the input to the 5V regulator. (Thanks for the part numbers) I don’t know if that jumper is supposed to be there, but now I get 5V across that one cap, and 2.5V at the output of the LTC reg with respect to batt neg. But I still get – – – – on the display, except at power up when it shows F177. I’m afraid this one has too much damage from corroded traces and vias. I have fixed a bunch of them already.

  17. Bob Grieb says:

    Hi again. In the photo here
    http://mrmodemhead.com/share/pics/F17x_Sections_Markup.jpg,
    I see +2.5 and -2.5V marked on a capacitor. I have 5V across that capacitor, but the top end is 0V relative to battery – Also relative to battery -, the output of the 2.5V regulator is 2.5V Should the 2.5V output be connected to the bottom end of the cap that is marked +/-2.5V? I see a trace running under the 24351 chip that has the +2.5V on one end and the +5V (cap bottom end) on the other end. Maybe that trace is eaten away under the chip? Maybe I screwed it up by connecting the switched battery+ to the input of the 5V regulator? At least with that connection I get some voltages over in the analog section…
    Thanks. Bob

    • modemhead says:

      Wow, I forgot that I had even made that graphic! Yes, the voltages are measured relative to the COM jack, which is usually considered to be analog ground on most DMM designs. Also, my recollection is that the negative battery lead is switched on these 17x meters, and positive goes through the previously mentioned FET. I’m away from home for a while right now and am unable to confirm that unfortunately.

      This is pure speculation, but I’m reasonably sure that a “—-” display happens when the processor loses communication with the A/D chips on the right side. Either they’re unpowered, dead, or the serial bus (I2C?) is interrupted.

      Similar to your problem, I have a wrecked 177 here with numerous corroded vias and traces, and I eventually had to give up on it and rob it for parts. I suspected problems underneath a chip that I could not see.

  18. Bob Grieb says:

    Think I see my problem. If I measure using the negative meter input instead of battery-, I get +/-2.5V on that cap, as in your photo. So I guess that’s all good. But I still get only – – – – on the display in all ranges. Also, no range indication, like AC, DC, etc.

  19. Bob Grieb says:

    Hi Again, I found the problem with the dashes on the display. The main chip feeds 3.3V to a resistor divider, and the selector switch selects one of the taps to pick a range and feeds it back to the chip. The trace that was connecting the voltage to the top of the divider was toast, so the chip was always getting 0V as the selection voltage, which is probably not a valid range selection. Anyway, now everything is working, except the mv range, which always reads 0. So I am making progress. BTW, is the white round selector switch part with the metal contacts on it in good shape from your “parts” 177? If so, and if you are interested in parting with it, please let me know. Thanks for your help with the meter.

    • modemhead says:

      Nice work! I admire your persistence. I was unaware that this model uses voltage-encoded function selection. The 8x series I and III use that method, not sure about the series V. And now we know that can also cause the “- – – -” display too.

      And unfortunately most of the front-end stuff and all the plastic bits from this parts unit (ie. carcass) has already been allocated.

  20. Bob Grieb says:

    Fixed the mv range. I was able to lift the white selector disk up without removing it,
    so that I could see the contacts and see where they were hitting in the mV range. Found one set of contacts that only seemed to be used in that range, and sure enough, the via to that contact area was corroded open. So now all of the ranges are working, but who knows how long it will last. Anyway, my job is done! I think a little water got into this unit, possibly through the probe connectors, and was allowed to sit for long enough to start things corroding. Lots more damage than the ones mentioned above. Thanks again for the help. BTW, I found a source for that selector disk. (Not Fluke!)

    • modemhead says:

      Once agsin, I admire your persistence. I know how tedious it is to do that kind of repair. Hopefully the meter will continue to work and give a few more years of service. And thanks for adding your experience to the pages here.

  21. Kitt1977 says:

    Hi

    what did you use to clean the case to nice ?

  22. Jarson says:

    I am from Venezuela,
      I have a Fluke 175 that will damage the regulator TPS77050 (I have to replace it)
    I removed the regulator but at the output pin of this regulator it has low impedance with respect to the negative of the battery (practically it is in short)

    And the component that is causing me this short is the Fluke SL10327

    I would like to know if anyone can measure the resistance between the output of this regulator and the negative of the battery to check if it is damaged.

    What does the Fluke SL10327 do in the circuit?

Leave a Reply

Your email address will not be published. Required fields are marked *