Denon DEC-61A4 (E) - 1969 - Nixie Calculator - No Function

[DrAlis]

Dear fellow friends of old tech. Another beauty is undergoing inspection. I have low hopes of fixing it and that is ok, but maybe some insight comes from sharing.

The machine lights the error and negative sign when turned on but shows no numbers. It is unresponsive to all keypresses except for (sometimes only) the CD (clear display) key. If it responds it shows a left aligned 0 and stays otherwise unresponsive to any further keypresses. Neither the error light nor minus sign can be cleared. Also no number can be entered.

I have carefully cleaned, reseated all boards, checked for visual anomalies, checked & measured keyboard for stuck or broken keys or reed switches and enjoyed the beauty of the machine. I also looked at it with infrared camera. Boards 1-4 are thermally more active. Boards 5-7 not so much. Resistors in power section are the hottest as usual but nothing to crazy. None of my activities have improved but also not deteriorated the lack of function.

I scoped the voltages from the power section and get the following voltages: 200v (yellow), 190v (orange), 100v (green), 7.5v (brown) and -5.8v (purple). scope lines available in the pictures for those of you who know what to look for. no clue if the voltages are good as no documentation available to me. the mitsubishi ic's seem to work on 5v.

I want to share some photos and invite your ideas on what to check or do with it to improve function. Thank you.

 

[DrAlis]

Maybe if somebody has some documentation or knows what voltages the descret diode logic ususally operates. I guess 200/190/100v drives the nixie tubes. -5.8V would be for the mitsubishis i guess (but why not +5.8V or does it even matter) and the 7.5v finds me puzzled. also, does the ripple seem acceptable?

 

[BlaBla1985]

From what I've experienced working on Nixie tube calculators (I've got a few) is that the high voltage is for the nixie tube display. Or at least one of them is, my guess would be 200 V. Not really sure why there is also a 100 V, as I believe most nixie tubes need that 190 V or more to stay on... but I stand to be corrected on that.

At first I thought you had some actual tubes in there, which might have used the 100 V as a bias voltage or something, but they are sleeved resistors. Those may still use that 100 V to do something, as there seem to be a lot large resistors on those boards.

The 7,5 V is probably used further down the line to generate the +5 V with a zener diode. I would guess the - 5,8 V is a bias voltage for either the IC's or some transistor circuit.

It looks like a really nice machine, but without some documentation it's really hard to troubleshoot. I wonder if those Mitsubishi IC's are special chips or just relabled TTL logic IC's.

On the ripple, I would guess it's very much a matter of what the voltage is doing. I don't think the nixies would really care that much unless it dips so far that it can't sustain itself. But the internal logic could be a lot more sensitive to ripple.

A small edit though regarding that +5 V, this is asuming there is a +5 V in the system.

 

[DrAlis]

Thanks, yes pretty machine but no docs. The pcb cards have all marked (+) (-) + G on the left connector on the center side. I could measure there but that would require powering the machine with missing pcb cards. I am not sure this is safe.

aside from power are there frequent failure components to check?

 

[durgadas311]

Did you find the article at the Old Calculator Web Museum? https://www.oldcalculatormuseum.com/denon61a4.html

Looks like those vertical (perpendicular to the PCB) components are for diode-resistor gates. Perhaps the ICs are more complex things like flip-flops. Being very early ICs, you'll have a harder time finding the datasheets. It's also before standardization of the numbering. But they might not require 5v either, depending on what they are - they may be more like a collection of transistors, diodes, and resistors and as such aren't dependent as much on a specific voltage - or just aren't standardized at 5V.

 

[DrAlis]

Yeah, I read the Old Calculator Web Museum Article. Thank you though. Indeed a collector in Germany https://www.thomas-kirchhof.de/ seems to have a working specimen. I might email him to maybe get some reference measurements.

 

[DrAlis]

BTW, I do have a 37 page operating manual, if anyone needs it please PN me. It is to large for upload here.

 

[ClydeBuckner]

BTW, I do have a 37 page operating manual, if anyone needs it please PN me. It is to large for upload here.

PM Please send it to me, I'm looking for it!

 

[DrAlis]

You have a message. Wow, I guess you just signed up for the manual. It is nowhere to be found online, so I guess the scanning was worth it. I am glad I found another shoulder to share the load of preservation.

 

[DrAlis]

Yeah, I read the Old Calculator Web Museum Article. Thank you though. Indeed a collector in Germany https://www.thomas-kirchhof.de/ seems to have a working specimen. I might email him to maybe get some reference measurements.

Thomas offered to bring the two machines together and measure things to compare in order to get it fixed. That is quite nice but my electronics repair skills are not up to the level needed for that task. Anybody in Germany willing and able to join us for this crazy idea?

 

[ClydeBuckner]

geometry dash subzero​

You have a message. Wow, I guess you just signed up for the manual. It is nowhere to be found online, so I guess the scanning was worth it. I am glad I found another shoulder to share the load of preservation.

I am so grateful to you!

 

[g arnold]

Hello DrAlis,
I also have one of these doing exactly the same thing.
I have gone over the power supply a few times and all my voltages are correct with virtually no ripple.
The chips on the vertical pcb's are mostly AND and NAND gates using old school TTL logic meaning its 5 volt logic but they switch states between 2 and 3 volts, there abouts.
The diodes are all OA90 small signal Germanium so they measure about 0.2 to 0.3 volts when measuring with a multi meter. Im currently going through the pcb's and checking components.
As mine was working perfectly then the next time I turned it on it would not flash up. I was sure I was going to find a bad capacitor but all measured ok and ripple is minimal.
If I find anything over the next few months I will post it here. My gut is telling me power supply, more investigation required in case I missed something.
Cheers, Geoff.

 

[g arnold]

Dr Alis,
Would you be so kind as to ask Thomas if he could please time how long it takes for the display to come on after power is switched on with his working calculator as I'm thinking it may be a reset issue
Cheers, Geoff

 

[DrAlis]

Hi Geoff, sorry for the late reply. I go in phases with my vintage collection. I will ask Thomas and report back.

Did you make any progress on yours?

Also, this forum is great! I am so amazed how people pop up from all over the world with the same machine or a similar problem. If you post something on some devices, you instantly become the #1 search engine result on it. Also, it gives me cold shivers thinking how much knowledge is lost on these machines and the engineering ingenuity that went in there. Sad to see that leave the universe.

 

[DrAlis]

Dr Alis,
Would you be so kind as to ask Thomas if he could please time how long it takes for the display to come on after power is switched on with his working calculator as I'm thinking it may be a reset issue
Cheers, Geoff

I asked Thomas with the working specimen.

He operates it on 200V (rather than the 230V we technically have here). He hopes it is more gentle for the device. First tubes only lit up after 10s after turning on. After 30s all are lit. Device immediately accepts keypresses though. He concludes the high voltage part takes longer to come up than the low voltage part. Retesting the booting after 5min, the first tubes lit immediately and after 15s all are active.

Let's see if that helps you. Best!

 

[g arnold]

Hi Geoff, sorry for the late reply. I go in phases with my vintage collection. I will ask Thomas and report back.

Did you make any progress on yours?

Also, this forum is great! I am so amazed how people pop up from all over the world with the same machine or a similar problem. If you post something on some devices, you instantly become the #1 search engine result on it. Also, it gives me cold shivers thinking how much knowledge is lost on these machines and the engineering ingenuity that went in there. Sad to see that leave the universe.

G'Day Dr Alis,
I have just started to trouble shoot again. After about 30 minutes a tube in the middle just starts to flicker and the only thing that lights up is the neon on far left of display when the CD key is pressed and CLEAR will clear it. All the low voltages on the power supply are ok and all caps check perfectly but the hi voltages seem off so i think its a tube bias issue. Sounds like we are all having the same or similar issue and I still think its to do with the high voltage.
I am currently tracing out the entire power supply of this beast so when I finish I will post it here for all. Mean while I will look more into the H.T.
The High voltage is derived from a doubler circuit, the capacitors in mine are ok but I think Thomas's might be on the way out.
Indications on the display PCB are marked 110 volts and 190 volts; mine are way higher so it seems something is not pulling it down and its left to float, kinda makes sense.
So I will make sure the power supply is ok and go from there but as all the other voltages appear to be ok I think all the plug in PCB's will be ok.
Being a shift worker I have taken this to work so I can play with it back in the Hotel room so you wont hear from me during my 2 weeks off rotation but im determined to sus this out. Oh and yes your comment about lost knowledge is 100% correct. I see this in my trade, aircraft avionics; the younger guy have been brought up with computer aircraft and are lost with the basics. I guess its like a Widows 10 person versus some one that has learnt on DOS ( no longer need it so why learn it )
Oh indecently the low voltages should be +5, -5 or there abouts, +8 and -23
Cheers, Geoff.

 

[g arnold]

Its been a while since I saw my unit running, are all the tube supposed to light up all zero's and the change with key strokes?
The higher than normal high voltages that I'm seeing are due to the fact that the tube arnt being turned on so they are not loading the supply lines.
Unfortunately all of the power supply outputs seem to be correct so it something deeper. I'll finish tracing out the P.supply while I think of the next step.

 

[DrAlis]

i have the original manual (in german) if you pm me your email i can send you a copy. It has zero suppression and resets upon power on. So on power on you have a single "0" on the right hand side of the display.

 

[g arnold]

My friend said he has the manual for this machine so if he cant find it I will contact you.
Thank you for your offer.

 

[g arnold]

Hi all,
Finally finished tracing out the power supply section, see attachment. In my case all the voltages are as expected now that I can see how it works.
The 5 volt and the -23 volt rails are the only 2 that are regulated. There is what I believe to be an 8 volt and a -6 volt rails that are adjusted by a 10 ohm resistor or rheostat if you like so the output on these are dependent upon the load from the cards.
The Hi voltage rails, 110, 190, 220 volt rails are also dependent upon load. As my Nixi tubes are not lighting up they are not loading down these rail so my voltages are higher than stated here. 110, 190, 220, are marked on the PCB with all the green cylindrical capacitors.
As there did not appear to be any problem with the power supply I started to trace back from the tubes to see how they turn on and it appears to be a matrix system, meaning a combination of pulses on the tube cathodes and anodes, so I'm guessing you would see a rolling flicker with a high speed camera if all "8"s were selected.
I have to look deeper into this but basically to turn on a tube an NAND gate on the display PCB pulses a capacitor low which biases the tube on, but this is not happening, so I put that aside as it gets its signal from the vertical cards.
Each card has been tested as far as I can at this stage. I made a very crude test jig to supply 5 volts to a card then probed each chip to see if the NAND gate logic I.C's were working.
As these are DTL they are slightly different to TTL. EG the M5496 looks the same as a 7400 but the M5946 has an internal 5K pull up resistor on its out put and the M5340 looks like a 7401 but the M5340 I dont think is an open collector like the 7401 but it also seems to have output pull up of about 5K.
I went through every chip and those that I was suspect on I pulled, then and tested in an other test jig. All checked out as best as I can see.
Notes on checking in circuit: inputs greater than 1.6 volt = High; inputs less than 1.5 = low. Output greater than 2 volts = High and output less that 0.8 volts = Low. Approximate readings.
Some M5340's were confusing to measure in circuit as both inputs measured High and so did the output ( remember NAND gates ) then going back one input was now low and the gate measured ok; the internal impeadence of the multi meter (Fluke) was just enough to load and change it state so be careful if you try this.
Found some info on the M5946 and M5340 here: https://www.zagroselec.ir/stfiles/getappdocument/1/true/099f85e7-9521-4869-9c6c-8e41cecd612f.pdf

All diodes were checked for forward bias and shorts with out removing, some are Germainium measuring 0.3 volts and some are the usual silicon 0.6 volts.
There were two metal can chips on PCB No. 6 that I could not check or find any data on, SM7006, mystery chip, timing oscillators ???
Did not check every resistor but those I did were well within tollerence, all capacitors and transistors all good.
I am confident that the problem is not one of these 7 cards and the power supply is good but not sure about the SM7006 round can chips.
Back to the display PCB's, I will try to trace out to see how this thing works.
Hope you can understand my drawings, I have checked and recheck but usual disclaimers apply. One interesting thing is how they get the negative rails, look carefully at the caps that are + to ground. Not the way I would have thought to do it but they must have had a good reason.
Cheers for now, Geoff.