Fluke 77: Strange DC Voltage Offset

Status
Not open for further replies.
Joined
Feb 6, 2010
Messages
4,836
Location
Central Texas
Got curious about the cause of this recently. It's a first generation 77. Bought it new in 1983. Worked fine up until last year. Perhaps some techs here have seen/heard of this? A 'net search didn't really turn up anything.

Powers on normally. Passes self-test. On Vdc, display reads -2.246 V then slowly begins to fall. With leads shorted, it reads 0.00Vdc. Unshorted, it reads -2.165 for a bit then slowly falls. I don't leave it shorted for long, just to see what happens

On 300mV, display reads "- OL"

On Ohm setting, reads -2.884M Ohm. I short the leads it reads -0.1 Ohms. Unshort back to earlier reading.

On diode check, display reads OL.

On Aac reads 0.02 amps with last digit ticking slowly. Looks normal for Aac.

On Adc, reads 0.00 amps. Looks normal.

Given the above, something's FUBAR on Vdc, 300mV dc/ac, Ohms and diode test settings. Ampere settings are unaffected.

This negative DC offset sounds like something is shorted. Capacitors block DC, but not a shorted one. Thus DC would get in somewhere it's not supposed to. On 300mV, this -2.3V is clearly over-range, so the display reads OL. On the Ohms settings, the -2.xx MOhms reading indicates a voltage being fed into the ckt, which isn't right.

May also be a shorted transistor. I opened the case. Inside is primarily film caps, a few tant. and one or maybe two AEC's under the display. The AEC's are suspect. Film caps don't usually go bad on their own. Tant. though....not sure if these go wobbly after decades like AEC's.

BJT's can go wobbly though. I'll have to dig up a service manual + schematic.

You bench guys ever seen this before? Dang odd. Just suddenly happened one day.

Another reason to heat up the iron.......
 
Ooen it up and look for anything “gooey” or dirty. Clean with flux cleaner.

Also look for cracked solder joints. A dead component is unlikely. A rusty selector knob is also possible.
 
Originally Posted By: danez_yoda
Ooen it up and look for anything “gooey” or dirty. Clean with flux cleaner.

Also look for cracked solder joints. A dead component is unlikely. A rusty selector knob is also possible.



Yeah, 30 year old flux residue can leak current across a PCB and throw things off! I scooped a
super expensive instrument that was sold off cheap, I found it drifted lots during bench test. Knowing the input
impedance would be well over 100 Meg ohms, flux cleaner did the job!

On cracked solder joints, you'll need a really good magnifier and light to see any kinda crack, sometimes
a crack looks like a cold joint but in a thin line either thru the solder fillet that's easily missed
or just around the component lead. Surprisingly, gold plated leads don't always take solder that well!

I've got 2 primo, expensive to replace Fluke 8842A bench meters that track the calibrator bang on! If they go south
I'll be in trouble! LOL
 
Fluke left no residual flux on the board. Very clean. Did a board hover with a mag. hood. No cracks or broken traces I can find.

Meas. ESR of caps I could easily get to. Nothing strikes me as the culprit...yet.

I did locate a very old service manual! Verified component locations, labels and values on schematic. Found a BJT with coll & base tied together, forming a diode. If this shorted, it may be the problem.
 
A capacitor doesn't have to be shorted .. just leaky. (I mean resistive leaky, not juicy leaky)
I'd clean the PCB with alcohol and let it dry.

Probably time for a new DVM - I recently got one with a large display for my old eyes.
cool.gif

The meter I used for decades was a Beckman TECH 310, I loved it
 
I vote for dirty contacts on the mode selection switch. Doubt it's caps. In 30+ years in the industry, I have seen very few caps short, they go open with age much more frequently.
 
Time for an update: I followed the trouble-tree in the service manual. Vdc reads correctly despite the no-connection offset. Ditto for the Ohms function. Checked ESR of all tantulums & one AEC. OK. The U1 IC looked straight on the pads with only a few legs maybe solder-starved. Not enough to warrant heating it up. No cracked/ringed solder joints.

I gave the board a good wash, a close examination and swept out any remaining debris with a stiff natural bristle brush. I'd read that Fluke had some issues on the earliest of models with solder-starved joints on the 60-pin U1 IC. One or two looked 'dry' but I left them. Fortunately pin/pad alignment was dead on.

I thoroughly cleaned out the banana barrels with ethanol, and around the posts. My 'board hover' showed a bit of green corrosion on the inside button of the NEG batt. terminal, which is riveted to the stand-up mount. I removed as much as possible (to no improvement), so I unsoldered it, took a wire wheel in a MotoTool to it, then soaked it in vinegar to eat the alkaline paste residue. Afterwards, I gave it a shot of DeOxit D5 spray and let it sit overnight.

I used a dental pick to rake out the joints between the + & - 9V batt. terminals and anywhere else that caught my eye. Over time, who-knows-what can build up on top cause tiny little conductive paths that weren't there when new. I also closely checked the probe connections and between the posts. Washed again, then blew dry with air, then had another look for any areas I missed.

Next day I resoldered it back in place, installed the 9V batt. and powered her up. Vac settled at 0.01. I noticed a + 1.5V offset on Vdc for a few sec. then it settled down to 0.00! Clicked over to 300mV, same thing. Ohms setting right at OL.

Success!

Looks like she was hankering for a good bath & a scrub and it did her a world of good! Interesting how Fluke uses a 9V batt. It's not 9V and gnd but rather +Vdd and -Vss. Internal to U1 there is more going on. Fluke used a dual slope A/D which requires two inputs as shown in the SM. Everything is converted to a voltage, filtered, buffered, integrated and counted before being outputted to the LCD display driver.

Speaking of a bath & scrub, the outer protective vinyl/rubber/plastic case needs a good one then a treatment of protectant. I mixed some powder Tide + hot water in a small cooler, dropped in both protective sleeves, agitated a bit, then shut the lid for an overnight soak.

Next day, the water was black. They didn't need much scrubbing due to the long soak. Both came out looking clean and new. Wiped them off then let both dry in the Sun for a few hrs. I'll wipe with some Mequiar's VLR.

Glad the old F77 is back in service. I've used it several times since with no more issues. Glad to cross this repair off this list.
 
great!
you wont likely find a better dvm unless you buy another fluke.
and i run lithium primaries in mine.
they dont leak corrosive stuff.
its worth $8 for an energizer lithium to protect a high end meter.
 
Thanks. Bought that meter in 1983 same year I graduated EE program. We've tested a lot over the decades too. I don't store batteries in any electronic equip due to leakage. But the DVMs get used so often it's a hassle to remove/install.
 
I still have my B&K from 1981. I wouldn't give it up for love nor money. It's the best diode and transistor tester I have ever used in a VOM
 
Status
Not open for further replies.
Back
Top