Nixie Tube Display Problem with Fluke 8100A

[Paceguy]

I'm restoring a older Fluke 8100A DMM that has a nixie tube display. There is a problem displaying past 1.399, 13.99, 139.9 and 1399. It appears that displaying the number's 4,5,6,7 and 8 is not happening. It uses a DM8840N or SN7441 decoder driver. Is it possible that this chip is faulty and causing the problem?

I now that this is an old DMM and could easily be replaced by a more up to date one, but I have several older nixie DMM's in my collection and want to get this one up and running. I love those old nixie displays in test equipment.

[Martian Tech]

The Nixies are multiplexed on the 8100a - there is a single 8840 driving all four of them.  So if any of them can display all digits, the driver chip is not the problem.

But it's not clear from your description that the meter is not operating normally.  You only gave 3+ digits in your example, which would say that V2 is not displaying properly.  However, if you really mean 13999 (4+ digits) then there is no problem.  The 8100a is not a full 4 1/2 digit meter, but has (or is supposed to have) 11999 counts (4 digits with 20% overrange).  So you would not expect even to see a reading of 13999.

So to understand this better, please answer the following:
- which nixie tube (V1 or V2) appears to have a problem?
- does it digits over 3 if the leading 1 is not there? (e.g., values of 400/500/etc or 4000/5000/etc)

Photos or video would help.

[Paceguy]

Hi, Thanks for answering. none of the nixies will display digits 4 thru 8, when it's suppose to, it flashes 9's in their place. On the dc  range with nothing connected, the least significant digit flashes 9 constantly. It doesn't display 4 thru 8 on no matter what function or range. For example if I feed 15.00 volts DC on the 100 volt range it displays 19.19, same with 14 volts. As a 4 or higher is to be displayed 9 is displayed instead. I hope that helps.

[Martian Tech]

Interesting...  In that case it could be the driver (A5) or the counter (A4) or whatever drives the clock input to the counter (the A/D Converter)...

12V displays OK?  If so, the A/D converter and its remaindering circuit would seem to be working OK.
Do you have an oscilloscope? (may not be needed, but...)

I think I would try removing the counter chip (A4) and  testing it on a breadboard as a first step.  If it checks out OK, you might then try manually setting to inputs to the driver chip (A5) to +5V (through pullup resistor) or 0V to see if you can get it to display all digits...

[bdunham7]

I now that this is an old DMM and could easily be replaced by a more up to date one, but I have several older nixie DMM's in my collection and want to get this one up and running. I love those old nixie displays in test equipment.

Don't worry--we understand!   

I haven't dissected one of this model yet, but I had a quick look at the manual and I'd get ready to do some careful oscilloscope work as well as some reading.  I sort of doubt the issue is the 8440 BCD/Nixie decoder itself, but that is the first thing you want to check with the scope.  You'll need to look at the '4' and '8' input lines as well as the '9' output pin.  An 8-channel (or more!) scope would actually be helpful here, but I'm sure you can manage with less. The operation of the display here is closely integrated with the operation of the recirculating remainder AD converter, so at first glance it appears to me that the problem could be anywhere.

You mentioned that it 'flashes' 9-does that mean that on those digits that erroneously display the 9 are flashing on and off while the other digits do not?

[trobbins]

It may also be worthwhile removing one nixie at a time and see what is displayed by the rest of the nixies.  A bad/poor nixie can load down a display numeral('s) line(s).  I had to work my way through some poor numerals in some nixies in my Systron Donner DMM (7004 model) - ended up swapping one poor nixie in to the '1' digit LHS location as that nixie could display 1 ok but had other numerals that were bad - but that SD model's display used separate driver ICs for each nixie, rather than bussed numeral lines.

[Paceguy]

I have a few spare nixies, I'll try swapping them out as suggested to see if maybe it's caused by a defective one. I'll get back with the results.

[Paceguy]

It may also be worthwhile removing one nixie at a time and see what is displayed by the rest of the nixies.  A bad/poor nixie can load down a display numeral('s) line(s).  I had to work my way through some poor numerals in some nixies in my Systron Donner DMM (7004 model) - ended up swapping one poor nixie in to the '1' digit LHS location as that nixie could display 1 ok but had other numerals that were bad - but that SD model's display used separate driver ICs for each nixie, rather than bussed numeral lines.
[/

That's right, the 9 only flashes on the nixies which are suppose to be displaying 4 thru 8'squote]

[Martian Tech]

Assuming the nixies check out OK (and I'm guessing they will because it seems unlikely that all of them would have the same failure), the 8840 will blank the display for input values of 10-15 (or A-F if you speak hexadecimal), so the flashing 9 could be attempts to display those digits somehow.  But that would mean that the count for the flashing digit was not stable, and I'm not sure how that could happen...

[Paceguy]

The 8840 in mine has been replaced by the previous owner with a SN7441AN. It appears that this is a suitable replacement from what I see online.

[bdunham7]

How fast is the flashing?  Is it a flicker, or a steady purposeful blinking?  Any chance you can post a video?

The way this works, with the Nixies all being strobed in sequence and timed to the operation of the stages of the ADC without any latches, makes the whole thing very susceptible to timing issues.  I haven't looked at the schematic and worked it all out, but that's where you probably need to start.  But I would definitely verify the operation of the SN7441AN before doing anything else.

[Paceguy]

I tried to post a video, no go. It flashes about two-three flash per second

[Martian Tech]

7441 or 74141?  The 7441 will output something for the 10-15 input values while the 74141, like the 8840, blanks the output.  So if it's a 7441 the blanking is coming from the ring counter or a flaky +200V feed (or a bad 7441).  But it's odd that it only does it when it's displaying a 9.

Is the 7441 socketed?  You could pull it, ground pin 13 with a jumper and all the nixies should display 4 (similarly, pin 14 for 5, 11 for 6, 10 for 7 and 1 for 8).

IF the driver chip is a 7441, which has no blanking option, and IF we assume a single cause for all the weird behavior (unable to display 4-8, flashing 9) then it looks like the driver chip (or bad socket/solder joints on the driver chip) is the problem.

[Paceguy]

The 7441 is socketed, I'll try that too.

[bdunham7]

IF the driver chip is a 7441, which has no blanking option, and IF we assume a single cause for all the weird behavior (unable to display 4-8, flashing 9) then it looks like the driver chip (or bad socket/solder joints on the driver chip) is the problem.

What about illegal or mistimed BCD inputs?  I think he needs to scope it and make sure there isn't a 4 and 8, or 1,2 and 8, at the same time as the 9's pop up.

[trobbins]

... I'm guessing they will because it seems unlikely that all of them would have the same failure ...

Given the common drive line for each numeral, it is plausible that a combination of bad nixies relating to numerals 4-8 could be stopping all nixies from displaying those numerals (ie. it doesn't need to be just one bad nixie, or that all nixies have to have the same fault).

[Martian Tech]

... I'm guessing they will because it seems unlikely that all of them would have the same failure ...

Given the common drive line for each numeral, it is plausible that a combination of bad nixies relating to numerals 4-8 could be stopping all nixies from displaying those numerals (ie. it doesn't need to be just one bad nixie, or that all nixies have to have the same fault).

Oh, yes, you're right.  The cathodes for each digit are wired in parallel so one bad one could pull down the other three. 

[CaptDon]

The 7441 is a dog, get at least the 74141 or maybe the russian chip!!! Also, look closely at each nixie for metallic cathode sputtering!!! If the nixie looks like it a huge mirror like getter it is probably having multiple cathodes shorted to each other and why the original owner gave up on it. The cathodes lose metal from ion bombardment and eventually they develop shorts. This is even worse in multiplex modes with non-extra-life versions of nixies. Minor leakage paths make the driver chips go nuts!!! The 7441 is only rated at around 55 volts for unlit cathode segments. Almost all other driver chips are rated higher with the russian chip rated at 100 volts. The russian chip includes blanking for input codes over decimal 9.