Elektronika EPOS-73A stuck with fixed display

[carlitosbala]

Hi,

I'm hoping I can get some ideas on what to check in this old (1978) soviet-era electronic calculator I got from a flea market, the Elektronika EPOS-73A.

Got it home and plugged it in. It worked just fine at first, I used it a few times for a few minutes each time, "kid with new toy" style.

Then I disassembled it to clean the plastic case and the keys. I couldn't separate the lower half of the case from the board because cables are tied to it, so during the cleanup one of the cables going from the power board to the main board detached, and a leg from one of the capacitors broke.

I re-attached the cable, replaced the capacitor, and now the calculator doesn't work anymore. The display shows "111" in the first few digits instead of "0", with other digits not turning on or doing so dimly. Button input doesn't change what's shown.

It has a reset button, when I press it it sometimes shows nothing, other times shows "999" or "ººº".

I noticed when taking photos for this post that if I press (slightly bend) the board in different places I get some of the digits to be brighter, and that one of the K264YM2 (which according to google translate is a "three display amplifier") is getting hot.

Regardless of the thing with the cable and the cap, I think something worked just enough when I turned it on the first time, and now has decided it's old enough to retire

Next time I check on it I'll reflow the connections in the back in case it's a bad solder joint.

I would like some general advice on, for example, how to properly measure the power board, and if someone has worked with this kind of old device, maybe some more specific recommendations?

Photos:
Working

Power board

main chip

Display as it is now when plugged-in

[stj]

Re: Elektronika EPOS-73A stuck with fixed display

the heater voltage/current is too high - you will kill the display if you keep running it.
the wires shouldnt glow like that - check the psu caps.

as it's soviet/communist there should be schematics available.

[carlitosbala]

I had forgotten about this thread

Thank you for your reply, I ended up taking the calculator to an electronics repair shop for diagnostic to avoid causing more damage while testing. The guy at least confirmed the power supply board looks OK, but apart from that was unable to find a fault and just told me that IC is bad and I should try to get a replacement.

I'm tempted to try removing it from the board and powering it, in my mind the best case scenario is the calculator should work and just have one of the digits not showing. If the IC is shorting the circuit to ground then removing it should also remove the excess current. Especially if I replace it with a resistor to compensate for the chip's voltage drop. However, I'm also afraid that might be disastrous so I don't think I'll try it, at least not until I try finding that chip for some time.

[stj]

find the schematics for it - then address the heater voltage

[redwire]

The VFD looks just like an IV-18, I used one in an Icetube Clock.

[carlitosbala]

Just an update: I got schematics, for the calculator itself (there is an emulator project at SourceForge, https://sourceforge.net/projects/emulator-epos-73/), the display controller chip, and the processor. However, their usefulness is limited by the fact I don't understand Russian and Google Translate doesn't help that much because the PDFs are images and not actual text.

As for further testing, I played with the potentiometers in the power and main boards, and there is a position where the display shows a "0" and the wires don't glow as much. But the calculator is still unresponsive, nothing changes when I hit buttons. Unplugging and plugging it again not always gets it again in that "0" state.

Took the "faulty" display chip out but nothing changed either. I noticed that the chip doesn't get as hot depending on what is being displayed, so I guess it is not faulty after all and is just getting hot because is being overdriven and always on.

At one hand, I'm thinking if maybe some of the orange components (which I assume are transistors) might be faulty and not giving the processor the right initialization signals. But I need to first confirm they are transistors, and then read enough of the schematics to identify proper replacements in case some are actually faulty.

Finally, what if the problem is that this thing wants ~220V and is just not liking the current ~240V mains voltage? It would be problematic because it would mean those first few times when I plugged it and it worked I was in fact causing a failure, but in any case I dont have equipment to help me test that theory, unfortunately

[stj]

attach the schematic

[carlitosbala]

The most complete documentation is what is there in SourceForge, really. Or at least that is the best I can find. It's 37MB, I'd assume attaching it is not possible

[stj]

i'll take a look

[carlitosbala]

Sorry to revive an old thread, but it's mine and wasn't solved so I guess I can do it

I left this calculator as a pending project for a long time, until I got the opportunity to ask a Russian speaker to translate a few parts of the schematics doc, and Google Lens helped me with the rest.

The power supply board is supposed to give -2.5V, -48V, -27V DC, and 6V AC. All voltages are present (within a tolerance) except the -27V, which is giving the same value as the -48V line. I can't say for sure it's a problem with the transformer, but the board is simple enough that its the only possibility I see.

So, with that in mind, I'd like to assume nothing got fried by the over voltage, and instead see if there is an easy way to get those -48V down to -27V.

My first idea was to search online for a replacement transformer, but I wasn't successful. Obvious in hindsight: even if I can get the same model number, it's still a 50 year old piece which can just fail at any time, and that will cost me 30€ or more because it is sold as a "vintage" part.

Next idea was to buy a dc-to-dc converter board and check if I could connect it "backwards", i.e. have it reduce 48v to 27v. The maximum voltage rating of the converter on the board is 40V, but I assume running it for a few seconds at the time won't be an issue.
Without connecting the main board the converter seemed to work, until I realized that since the "ground" of the psu is the positive side of the converter board, there is no direct continuity between ground and positive output (which should be my 0V if connecting backwards), and I get the feeling things will go very wrong if I link that positive output with the board's ground.

So, now I'm trying to find a negative regulator I can use in the dc-to-dc converter board, replacing the regulator it came with. All parts I can find only go down to -40V, but I assume it won't be a huge problem to run them at -48V for a test.

I guess another possibility is to reduce the voltage with voltage divider, at least for a quick test. I'll search my parts bin in case I can find suitable resistors, but if that works the dc-to-dc board seems like a better long-term solution

I'll keep going at it because I'm that stubborn, and because I prefer to think the CPU is not already fried due to the over voltage. IF the lack of -27V is the root problem and the calculator works once I find a way to have it, I'd still need to think about how to replace the power supply board before the transformer fails and takes an irreplaceable IC with it... but I guess I can worry about that later