SCM Marchant 414 schematic diagrams anyone?

[Dekatron42]

Hi,

I just bought an old SCM Marchant 414 with a 14 digit ZM1200 Pandicon as the display element!

The Pandicon contains Neon and looks just fine however the previous owner had disconnected the power supply from the main board and the power supply has been repaired extensively.

All of the transistors and one UJT 2N2646 and one SCR TIC125M, there are five MJE340 and a bunch of small TO92 transistors that have also been replaced.

Before I switch on the power supply I wonder if anyone here has a schematic diagram of the calculator, especially the power supply? A quick Googling didn't turn up any circuit diagrams and just a handful of websites with information but none with the Pandicon, all had separate Nixies instead.

The voltages from the power supply are printed on the circuit board but since the power supply directly rectifies the mains voltage I'd really like to know how it works before just turning it on.

/Martin

[Paolo Cravero]

Hello Martin.

No, I don't have the diagram. I have seen that calculator for sale and I was tempted to get it ... almost just to try to fix the power supply! At least I think it is the same machine. Would you share your findings, if/when you get it back in working state?

I have a Casio CL-200 with individual Nixie tubes. it accepts digits endlessly but does not calculate. It has a good amount of discrete components with obscure part # and reverse engineering from the board is no-go (too long). No way to find a schematic diagram which, at the time, was probably highly secret. It might still be somewhere in Casio archives, though.

Good luck with your SCM 414.

Paolo

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/f1fd03f9-ed10-4263-a833-1e8cd5ed223b%40googlegroups.com.

[Dekatron42]

Yes of course I will share what happens.

I am right now reverse engineering the power supply but since I don't know what the original transistors where, they've all been replaced with modern components, and I am not even sure the replacements are the correct types I don't know where I will end up. I'll do my best with the schematic and if I can't understand the design I might build another power supply to power the calculator with to see if it works. The necessary voltages are +100V, +197V, +200V, +20V, -35V, -18V and -15V and the total power is only 15W.

/Martin

[Dekatron42]

I've already seen that a lot doesn't make sense, some transistors, the SCR and UJT have all been connected in the wrong way compared to their datasheets and how they are soldered on the circuit board plus one trace has been cut which is necessary to power it all up. So, now I have a lot to do to really get it right and up and running.

/Martin

[Dekatron42]

So here is a first try at the reverse engineered schematic, please understand that most of the transistors, the UJT and SCR has been replaced at one time or another on my power supply so I can't be sure that the schematic diagram is correct, especially when it comes to the small TO92 transistors.

The SCR and UJT was definitely placed incorrectly on my power supply and I also found a few dry joints and one unsoldered pin on Q8 and I also spotted one cut trace that definitely need to be there for it all to work as it otherwise wouldn't have any connection between the SCR anode and the the rest of the power supply!

The repairs are really poor as the solder has bubbled and looks really messy, have a look at the photos!

So now I'll have to run it through LTSpice or something to see if all transistors are of the correct type (NPN/PNP). The transistors Q6-Q10 are suspect in their type and all have been replaced at one time or other. Now, Q1-5 and Q10 seems to be correctly replaced.

If anyone have one of the calculators in this series and could check what transistors they have in their power supplies it would be very kind!

/Martin

[Dekatron42]

Forgot to include a photo of the component side of the power supply!

/Martin

[Dekatron42]

I got it up and running!

I just replaced Q6 which someone had replaced with a BC327-25, but looking at the circuit diagram I couldn't understand how a PNP transistor could work there so I replaced it with a BC547B as I didn't have any BC337 which was used in the other position.

I also repaired the trace where it had been cut close to the right trim potentiometer for the SCR/UJT control, it should connect the wires from the big electrolytic capacitor down to the three half watt resistors (two of which are 10 Ohm resistors).

The Nixie voltages (+200V, +197V and +100V) were a little bit high so I had to turn the left trimmer to its leftmost setting to get the lowest output voltage at the 230VAC Mains voltage we have here in Sweden - I guess these will have to be adjusted when the electronics is connected as there is no feedback in the regulation loop, so I might have to adjust the left trimmer again.

Adjusting the right trimmer to far to the left is dangerous as the output voltage rise sharply up to the full wave rectified mains voltage and above, so be careful if you do this!!

The positive and negative voltages for the ic's is also high without a load so I'll have to see if I can figure out what resistor to load these voltages with so I can see if they will deliver enough current.

I've included a new circuit diagram here with the corrected Q6 transistor and some other small corrections I found while checking it myself.

/Martin

[Dekatron42]

Well, the ZM1200 Pandicon is working as it should on all digits, but the calculator itself isn't, it is almost completely dead. There are only four signals from the main board to the Pandicon board that shows any life whatsoever, a pity now that the power supply works.

The power supply is not such a good design because when you switch it off the oscillator for the transformer goes into self oscillation and the UJT controlling the SCR also drives the SCR into full conduction which means that at the moment of switching off the output voltages rises a lot! There is no regulation feedback on the output voltages and maybe this poor design is the reason for the calculator not working, just guessing here but the voltages were several 10's of volts higher than written on the circuit board just at switch off and you can see flashes inside the Pandicon. Maybe a fully working calculator doesn't see this over-voltage spikes but I think it will even if it draws more power working.

There were a lot more bad soldering, dry joints and corrosion on the main board, it smells like a rat has peed on it and it looks like that in some places. I cleaned all of the corroded areas and checked all of the traces and components. Now, the 10% resistors are at least 10% high so those might affect the overall workings, might change some to see if that differs. I changed two small electrolytic capacitors looking like Tantalum bead capacitors but that didn't change anything. Transistors and diodes are ok, but I had to re-solder a few of them as I could easily see the dry joints and also measure the dry joint as I only got contact when touching the pin of the transistor and not the trace it was soldered to, as is the keyboard magnetic switches but the ic's all are non responsive, not a single pulse anywhere except for the four signals to the Pandicon board.

Well, it was the Pandicon I was after knowing that the seller had disconnected the power supply due to problems previously, but it would have been nice if the calculator had worked.

So, if anyone has any ideas on what to check to get it running I would be happy if you could write a line here or send me a message offline.

I hope someone can have use for the circuit diagram of the power supply.

/Martin

[Paolo Cravero]

Thank you Martin for the detailed report! I wonder how many of those machines are still around, just in case a fully working calculator can be put together from parts. But probably the display is the toughest component in there!

Maybe the damage was caused by a lightning strike, at least I experienced similar faults when I tried repairs after those events. Hard!

Congratulations for fixing the PSU.

Paolo

--

You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/6ef531e8-0ee0-4eaf-8833-011bb16a3078%40googlegroups.com.

[Dekatron42]

Thanks!

I don't think that many prototypes like this one is left in the wild, so it would be nice if anyone had a schematics diagram of the calculator part or datasheets on the ic's used.

Since the voltages rise well beyond 2-3 times the rated voltages I think the poor design of the power supply here is to blame, I could measure peaks of 580V on the 200V line for the Nixie supply at switch-off, that meant that the negative voltages for the ic's were similarly too high at almost -100V for the -35V line. They should have put some kind of over voltage protection in the switching power supply to have protected against that I think.

I'll piece it together and put it on a shelf for the future, maybe I'll find one more!

/Martin