Fluke 8040A - Let's take it apart (and see if we can fix it)

[david77]

I went to a car boot sale today and found this little gem: A Fluke 8040A DMM.
For a few Euros it changed hands and was added to my collection.
It is a handheld/bench DMM 41/2 digits and was made in 1974.
Power is provided either via a plug pack/charger or via four built in NiCd C size cells.



When I got it to my lab I first took the NiCd pack out to see if there was any damage from
leakage. Sadly there was, although it obviously wasn't the fault of the currently installed cells.
They seem to be fine.
One of the battery contacts was corroded off. I replaced it with a contact from an old portable
tape recorder.

So next test: Let's apply power from the bench PSU. Sadly nothing happens, the display shows nothing.
I could hear some high pitched noise from inside, though. Like a switch mode supply. Ok, what's going on
here? Now I had to open it up.



Isn't that beautiful? More like a work of art then electronics, really. All components are through-hole, of
course. And all the IC's are standard CMOS 40xx types. Except one, wich is a custom Fluke type.
And sure enough, there really is a small switch mode supply inside.





Luckily I was able to find a manual online.
After a few measurements I found that all the supply voltages were present and correct. The SMPS is
an up converter generating +/- 11,5V and +/- 6,5V from 4,8V.
The manual had some suggestions what might be the problem when the display shows nothing, one
of wich was worrying: The custom LSI IC could be broken. A repair wouldn't be possible then. I don't
for one moment believe Fluke could still supply that one!

There were also several other IC's that could be responsible. One was a 4049 that is used as a 1MHz
clock generator together with a 1MHz crystal. I poked around there a bit with the scope probe and found
some very dodgy looking signals. The clock was not looking right. So I replaced the 4049 with a new one.
Power on....and hey! The display is back!
And it works brilliantly.



David

[slburris]

Nice repair!

Scott

[Kiriakos-GR]

I do not see any repair,  in pictures .... still I am happy for you ..

[david77]

Well, as I said. I fixed one of the battery contacts and replaced a 4049.

I'm happy to have repaired this meter particularly as I had a 8060A (Nixie tubes) a few years ago that I couldn't save.
Now I have the smaller brother.

David

[saturation]

Good sleuthing!  Congratulations!  Can you compare it against any DMM you currently have and lets see how off its calibration is?  Great job.

[Ernie Milko]

Nice work.

[david77]

Here's the comparison with two of my other meters.
The youngest one is the Metex in the middle (bought new about 10 years ago).

So far it doesn't look too bad. Mind you, none of my meters has ever seen a
calibration service, at least not since I had them!

[david77]

Update.

After a few days of joy my new antique meter's health deteriorated quite rapidly again.
It started to show only a tenth of the applied voltage (e.g. 0,1500 when it should've been 1,5000)
on all ranges but the 2000V range.
I followed the manuals instructions and started testing as described. After replacing a few J-FET's
and two OP-Amps as the manual suggested it still didn't work.
After endless fiddling and trying to understand the circuit I finally found the problem:
The range switching is done with small relais and the 2000V relais didn't switch off at all, so it influenced
the voltage divider.
The culprit was a partly dead 4019 AND/OR gate. Replaced that and it's OK again now.

I hope it stays that way now.

[MrPlacid]

Dave77, I have to agree your 8040's circuit board is artistically beautiful and very clean for something that old. I like the circuit boards inside these old but quality electronics. They're thick and has a quality look to them.

[JohnS_AZ]

8040's circuit board is artistically beautiful

That's probably because they were not autorouted. 

I have a couple of very old HP 1X8 RF switches (HPIB) that are pretty amazing inside. The circuit boards are actually gold plated. I will probably have no use for them but I keep them around as an example of what quality and craftsmanship used to mean.

[TheWelly888]

Have you replaced the electrolytic capacitors on your 8040? They could have dried out.

[david77]

No worries, have done that already and the tantalums as well

[House91320]

It's so awesome it  belongs in a museum.

[sonicj]

i attempted to power up my dad's old sabtronics 2000 dmm kit and i let out the magic smoke 

example

[saturation]

If its similar to the unit pictured, its very common for the power supply to blow.  Replacing it is maybe all that's needed to get it back up.

i attempted to power up my dad's old sabtronics 2000 dmm kit and i let out the magic smoke 

example

[david77]

@Sonicj: That Sabtronics meter looks funky. I would try and get the smoke back in, ahem repair it I mean
Can you get a closeup of the circuit board so we could see the markings on the IC's?

@House91320: Nah, stuff is built to be used, not to be looked at in a cabinet. So it now has a permanent place on my bench  .

[sonicj]

i'll try to get some photos up shortly....

a little more detail of the incident:
i read over the assembly & operation manual prior to connecting power. it ran off of 4 - 1.2v cells in its stock form but offered a 12v external power supply upgrade. the upgrade used a 7805 internally. i look on the back and see a coaxial plug that resembled the plug offered in the upgrade so i assumed he had installed the factory kit.  well... as you might have guessed, there was no 7805 inside  doh! lol!

one tantalum instantly exploded sending its hot molten innards throughout the inside of the case. (glad i had the lid on!) after replacing that 1 cap, the meter powers up but obviously has some issues. i haven't taken it any further since then. hopefully its not the A-D as i don't know how available those are. the rest of the parts look pretty common.

i'll try to get some pics up in a bit! cheers!
-sj

[sonicj]

pics...

[sonicj]

woops! the burgundy cap in second photo is the one i replaced. i put in a metal film, but its supposed to be a tantalum... is that bad? 

sorry if the photos are too large. i figured the forum would do a little more compression of the previews...
cheers!
-sj

[Mechatrommer]

i wonder how small the above circuits can be compacted by today's technology.

[sonicj]

a lot! that meter was VERY basic.


purdy sure its the ADC... unit powers up, the display driver appears to step through all the different settings, no readings at all though. 

manual lists the ADC as a Motorola 20-786. package is 24 Pin Dip. no luck finding any so far!
-sj

[JohnS_AZ]

The part number "20-786" is most likely a house number stamped on parts sold to the manufacturer.

If you don't have a schematic it would take a bit of reverse engineering, but that may very well be a Motorola MC14433 .

You might also compare your board to this .

[JohnS_AZ]

LAM! (laughing at myself)

Or you could just look HERE.

The PDF has the schematic for the Sabtronics 2000 as well as some text that says the A/D *IS* a MC14433. It also calls out the other house-numbered parts.

Interesting... that meter was actually sold as a kit.

[JohnS_AZ]

and if I haven't been helpful enough ...  

I think the about the only place on Earth that you can still buy an MC14433 is Futurlec .

Further I *think* it's pin compatible with the TC14433 which is available all over the place. But, I'd do a spec-by-spec comparison of the two datasheets to make sure they are the same.

[sonicj]

great work detective! 

i probably never would have come across that as i have a original copy of the schematic here, sans the handwritten part number!

i'll probably just order the original replacement from futurelec. im in no rush and they have a bunch of other goodies to stock up on.

so ya, this was a kit my dad put together 30 or so years ago. he described it as one of the first affordable DMM's.
-sj