Iskra 122 - Nixie Soviet Calculator - 1979

[Rafael]

Hello,

Can someone please give me some tips?

I have a very old calculator in my collection of devices that use nixie tubes, it has a strange behavior, if I press the number 5, it will turn the number 6 on and vice versa. This in any tube. I make any calculation correctly and if any result has 5 or 6, they will light up together.

Where should I start?

I was able to identify 3 drivers, should I replace them?

I researched and I understand that this calculator has little documentation. Despite being a beautiful piece of Soviet history.

Thank you

[schmitt trigger]

I would start by downloading the datasheet for the Nixie drivers (which could be a variant of the 74151 driver), connect a 4 channel oscilloscope to its BCD inputs, and make sure that the correct BCD code is actually reaching the inputs as you type each number.

[tszaboo]

It is cyrillic. In cyrillic, P is R, B is V and 5 is 6.
Its just the way they write.

[Rafael]

Yes...

And for capacitores mk(|) is uF.

I will start to replace capacitors, I want to replace for NOS capacitors...

Thanks!

[ivaylo]

It is cyrillic. In cyrillic, P is R, B is V and 5 is 6.
Its just the way they write.

Can't tell if you are joking, but the 5/6 part is wrong...

[glarsson]

It is cyrillic. In cyrillic, P is R, B is V and 5 is 6.

No. We all know a Soviet 5 is an upside down 2.

[voltz]

If you are seeing both 5 and 6 lit together (as in your image) this indicates leakage current between the 5 and 6 digits.

The leakage could be within a faulty tube effecting the whole common bus of 5 and 6, or it could be a faulty driver circuit driving the tubes.
To check this, with the power off, try measuring resistance between the 5 and 6 lines at any tube base. You would see a lower resistance between 5 and 6, compared to any other two digit pins. Hopefully this will then allow you to trace further and determine if this is a tube fault or a driver failure.
Are those tubes on holders? if so, pull them all out and test them on at a time.
Good luck.. nice calculator

[Ian.M]

That sounds likely.  However those tubes appear to be wire ended and soldered in.  It looks like a double sided PCB so disconnecting e.g the '5' anode of tubes one at a time by desoldering the '5' wire and making sure its not touching the side of the PTH is going to be rather tricky.   It might be a better idea to start by tracing it back to the driver and isolating the driver output, so you don't have to mess with the fragile tube wires unless you are certain there is a tube fault.

[tszaboo]

Can't tell if you are joking

Obviously joking.

[razberik]

Not faulty tube, but faulty driver. I have replaced few of them in instruments from various countries. Solartron A211 suffered the same problem - TI 74141.

[Rafael]

Hello,

The failure of 5 and 6 happens in any tube, the drivers board (74141) is attached.

There is one driver for each tube.

I wanted to try to find out the common ground of these digits, or if there is some bad contact/escape of the tube board?. Or if there is some common component in the output of these... No documentation is a treasure hunt.

For restoration masters, is it more interesting an old calculator restored with old or modern components? What is important is to function or what is important is to maintain caution with a part of the history of electronics?

Thank you friends.

[Towger]

Have you tested the capacitors?
The old capacitors in this type of equipment may be better than NOS.

[Rafael]

Have you tested the capacitors?
The old capacitors in this type of equipment may be better than NOS.

I have not tested everyone, I'm going to change all that I can, I'm waiting to arrive ... They are the biggest suspects, yes.

[PA0PBZ]

Have you tested the capacitors?
The old capacitors in this type of equipment may be better than NOS.

I have not tested everyone, I'm going to change all that I can, I'm waiting to arrive ... They are the biggest suspects, yes.

Capacitors would be the last thing I would suspect with an error like this.

[Cyberdragon]

Golden rule alert! Don't go changing stuff unless you are certain of the fault or the condition of a component! There is always room to make it worse. Those are mylar caps, and this machine was made in the late 70s to 80s according to Google, so they are probably fine.

What you have here looks like a display memory or data multiplexing fault. Given that each tube has a seperate driver, it is extremely unlikey they all failed exactly the same way (there is no connection between them). The driver inputs are probably multiplexed from a single output/display memory bank that reads out each digit sequentially. Somehow, the memory is cycling between 5 and 6 after each refresh cycle. There might be a logic fault that causes an increase in the digit value when 5 is read from memory and a decrease when 6 is read thus causing it to alternate on each refresh cycle.

Might be talking nonsense, but it sounds logical to me.

[z01z]

The failure of 5 and 6 happens in any tube, the drivers board (74141) is attached.

Those are not 74141 ICs. It should have 16 pins, not 14. The soviet TTL copies has a different numbering scheme, a 74141 would be K155ID1 or KM155ID1 (look it up in google for the Cyrillic characters and other possible numbers).
If you look at the Nixie PCB and check how the different pins are wired, you can deduct how many drivers are used. If all pins for the same number are connected for all the tubes, then one driver is used. If there are two groups, then two drivers, etc.

[voltz]

Very unlikely to be capacitors in this case.

The common point you are looking for is at the nixie tube pin connections. As mentioned above, trace the tracks and see if all the 0 to 9 pins are connected together on a single bus. Or it could be in groups..
As i suggested earlier, please to check the resistance between pins for digits 5 and 6. If all the tubes use a single common bus, a fault in 'any' tube will effect 'all' tubes. In other words, a short between digits 5 and 6 in say for example tube 1 (left) would show as 5 and 6 lit together in all the tubes... Take a close look at the nixie PCB layout and draw it out if you have to.
But, the most likely problem would be a faulty driver chip. So basically fault finding in the other direction away from the tubes.
A simple resistance check between 5 and 6 would confirm leakage either way and you can start working from there.

[Rafael]

There is a driver (74141 soviet version) for each tube, so it is correct to say that there is no defect in these drivers, because I can go from 0 to 9 in all tubes correctly. Failure of 5 and 6, no matter the tube.

I will follow the suggestion of the colleague and measure the resistance between the cathodes to see if something is generating contact between the 5 and the 6... In my not so ample experience with nixie clocks, I already saw a tube with two cathodes in short internally. That we can call lack of luck.

I'll take the calculator a little deeper and come back with more images... 

Thank you!

[Mickle T.]

Iskra-122 User Manual: http://www.mediafire.com/file/mmqbcvkdm46lod8/Iskra-122_User_Manual_%28RU%29.djvu
Chips datasheets:
http://www.mediafire.com/file/zeu27mb18sr0ht3/K144IR2.pdf
http://www.mediafire.com/file/dtthnq1en4odh48/K144IR1P%2C_K144IR2%2C_K144IR3.pdf
http://www.mediafire.com/file/wzm4l4rjly946ko/K1UI271.pdf

[Rafael]

Searching in Russian already returns something too, but not much.
http://myadel-gimnaz.by/iscra-122%282%29

[schmitt trigger]

My original suggestion still applies: check the BCD code reaching the driver IC pins.

[z01z]

I wouldn't dismiss the possibility that the 74141 is faulty (there can be leakage between the 5 and 6 outputs), without additional tests.
If there is cabling between the 74141 and the Nixies, the wiring could be changed so that for example the 1 and 5 wires are swapped.

Also, as suggested, the BCD inputs of the 74141 could be checked with a scope. Fill the display with all 5s or 6s and check the inputs, especially input B.

You are probably aware of that the Nixies operate with high voltages (130-150V is needed for strike) and need a few mAs (or even 10s of mAs when muxed). So be careful where you poke when the calculator is energized.