Not much time for blogging these days, but this one is overdue. A reader of this blog was kind enough to send me his retired IBM 8060A/AA multimeter. (Thank you HK.) This is really a Fluke 8060A in a light blue/gray case, with IBM branding. Presumably IBM bought a lot of these for their engineers, field technicians, and laboratories. It even came with its own special blue service manual, which another friend gifted me with a copy, so that I have a matching set for Mr. Modemhead’s Multimeter Museum.
This turned out to be what I would consider a textbook case of 8060A restoration with the usual problems brought on by electrolytic capacitor leakage, plus a damaged LCD. From initial appearances, it’s not worth the time to do any sort of trouble-shooting. Just dive in and take care of the obvious problems, then see what happens. For that reason I spent no time trying to power up the unit until the damages were assessed and the necessary parts replaced.
Leaks Don’t Always Come From The State Department
Electrolyte wicks up the legs of components and causes corrosion and problems. In this case it affected the elastomeric connector going to the microprocessor and its associated PCB pads. It destroyed one of the contacts in the 40-pin socket for the custom MAC (Measurement Acquisition Chip), which was “fun” to remove. No special de-soldering equipment was used, just a temp-controlled iron and a generic solder-sucker. And then some solder-wick to clean up. The PCB is very well-made and sturdy, it seems to withstand this rework quite well.
Another common failure, apparently caused by electrolyte, is the destruction of the 8-pin DIP ICL7660 charge pump. It provides the negative Vss supply for the meter, from the +5.2V supply. It was tested, found to be bad, and thus added to the list of replacements.
This is a list of the parts ordered from Digi-Key. The capacitors are all Nichicon UTT series. The capacitor selection was done on two basic criteria: physical size, and a 105°C temp rating for long life.
|Digi-Key Part No.
|CAP ALUM 100UF 20% 6.3V RADIAL
|CAP ALUM 10UF 20% 16V RADIAL
|CAP ALUM 22UF 20% 16V RADIAL
|CAP ALUM 47UF 20% 16V RADIAL
|IC REG SWITCHED CAP DBL INV 8DIP
|CONN SOCKET IC 40-PIN T/H
After the parts were replaced, the board was thoroughly cleaned by immersion in about 1/4″ (6-7 mm) of IPA (isopropyl alcohol). A disposable “acid brush” is used to get some scrubbing action in between components. It is important to remove flux residue and any other contamination from the PCB, or the meter may be inaccurate and/or fail to zero out with no input.
A word about the push-button switches: In my experience, I’ve had very few problems that I could associate with these very sturdily-made push-button switches. Others have reported problems with them though. After a period of non-use, it is a good idea to exercise them repeatedly and vigorously. As for cleaning, I didn’t deliberately spray anything inside, but I did blow them out with compressed air after dunking the board in IPA, because some liquid will usually splash up in there.
On The Road To Recovery
Powering up the meter after the new parts are installed is very promising, as the test point voltages are all within spec. It is permissible to power the unit without the display assembly and microprocessor in place. It is not a good idea to power it without the 40-pin MAC chip installed, because it performs part of the voltage regulation.
Liquid Crystal Destruction
The LCD on this one is in bad shape. The rear polarizer is definitely damaged, resulting in the burn pattern. For an extreme repair, it is possible to peel the back off and replace this with some generic polarizing film and a reflective surface. I’ve tried various things including tin foil, but what seems to work best is just some bright white paper, as in high-quality inkjet paper. But that won’t work in this case because the LCD also has the “bleeding” effect, which is a defect inside the glass sandwich, and not thus not repairable. I don’t know where to buy a new LCD, so I had to wait to find a parts donor unit on eBay, that wasn’t over-priced by the seller.
New Life For An Old Horse
The moment of truth arrives after the “new” LCD is installed and the pieces are reassembled. It turns out in this case the tactic of going straight for the usual suspects and spending minimal time trouble-shooting, was a winning approach. The unit came straight to life. It zeroed out after a few seconds within ±1 count, which I consider to be nominal. A quick check of its calibration revealed that it really didn’t need any adjustment at all, but I did tweak it to match my 1.9V reference.
Once again, thanks to HK for sending this unit in. Hopefully this story will inspire or help someone else refurbishing their vintage DMM.