Anita Mk9 restoration

[Jean-Pierre G]

Hi all,

I am new to this forum and new as well with "tubes electronics".

However I do like electronics and microcontrollers and Android programming... Well a kind of geek !

To know a little more about me, visit my site : http://freedom2000.free.fr or have a look to my youtube channel (search for f2knpw -it's me-)

I join this forum because I have recently found an old lady : a Sumlock calculator branded "Anita mk9 model2 from 1964"

you can see it here : https://www.youtube.com/watch?v=1fOkUnu6PWc&t=18s

Since this discovery I have successfully powered it and all the display works

And now...I would like it to make additions, substractions, multiplications and even divisions... but the dekatron is dead.

The model is a GS10D

So please if you have a (working) GS10D dekatron and would like to sell it to save the heart of this old lady, it would be really really nice :-)

Please consider also that I am a quite serious guy and that I am followed and helped by Roland, the famous Dutch guy who already restored and Anita MK8 : http://technischmuseum.nl/devices/Anita%20MK8/Anita%20MK8.html

Thanks in advance

JP

[threeneurons]

Great find !

Working high speed Ericsson dekatrons are next to impossible to find. The GS10D has the same pinout as the slower GS10C, and 6476/6476A. Also the GS10C is designed to operate at lower currents, but it may work, if you can reduce the clock rate. For short term operation, running the GS10C over its design limits does no permanent damage. Something to consider, until a working GS10D shows up.

An excessive portion of high speed dekatrons are found to be duds. Usually no gas, to ionize. This rarely happens with standard speed (neon gas / orange) dekatrons. The gas in the high speed tubes, I believe is either hydrogen or helium. Blue/violet glow is not unique to argon, as some assume. If the gas is helium, it can diffuse right thru glass. If its hydrogen, it can bond with the tube's metal structures. In either case, over time, tubes with these gases will go bad, just by sitting on the shelf. And these tubes have not been made in decades. 

[Jon]

I'd broadly agree with Mike, though a touch more optimistic. Although there is definitely a problem with fast dekatrons dying in storage, the frequency of dead tubes varies significantly from one model to another, and GS10D are far from the worst offenders. A quick look at the records of the GS10D that have come my way over the years indicate 8 dead out of 24 total. I am reasonably picky about buying these tubes so perhaps the sample is skewed a little - maybe assume 50/50 in the general case.

JP - you have mail.

Jon.

[Roland]

Hi JP, Mike and Jon,

I have several GS10D tubes which just work great in my Anita's. At the

moment I've got a MK8, MK9 and MK12 restored to 100% working condition.

I've now got four working GS10D tubes out of 16 tubes total. Three of them are in the

machines, one is waiting for my other machines to be restored... (three MK8 and another MK12)

But to be honest, I'm afraid that within 10 years a few more tubes will be faulty...

That is why I warned JP, just don't trust used or new in box tubes. Most sellers can't test them.

Nigel once pointed me to this forum once, and I've asked the same question as JP.

I also just wanted to buy a (few) good GS10D. But since I have running machines already, a good tube

needs a good home. And that is the MK9 from JP :-) I really hope someone here can help him!

That GS10D is key in these machines. It is the 3.4KHz clock divider, it is all about timing in those Anita's.

The clock setting mark-skip ratio is very important. If it's not set to the right mark skip ratio

your machine will start to make errors. I've tuned down the clock frequency of my MK8 to see how

low I can get. I had to do this with an external clock source to remain the mark time and just

make the skip time bigger... If I remember correct I could get to about 2,8 KHz. But the machine

becomes really unstable then...

So I bought an GS10C to try. But this tube won't spin at all. It glows at just one pin.

No movement at all... The current trough the tube is too high for that GS10C I think.

Funny to note, the current trough the GS10D is also outside the specs of the tube...

Jon pointed me to a GS10E which I bought. Still have to test that one...

Regards, Roland

[Jean-Pierre G]

I do have the strange feeling that I am looking for the St Graal !

However I would like to publically say that I do really appreciate all the efforts you do to help me.

Regards,

JP

[Dekatron42]

Hi,

Can someone please send me the complete circuit diagrams for these Anitas, or point me to somewhere where I can download them? I’d like to have a look at the circuit surrounding the GS10D.

/Martin

[Jon]

Hi Roland,

Good to see you here. Can you say anything about the guide pulses in the Anita (resting bias voltage, amplitude, duration etc)? The GS10C and GS10D differ in these parameters too, not just the maximum current. Perhaps that is also the thinking behind Martin's question too :)

Jon. 

[Dekatron42]

Yes, I was also thinking about the differences in the driving conditions for these dekatrons, and the possibility of adding, or changing, the driving conditions.

/Martin

[Roland]

Here you can see the dekatron circuit and the oscillator.

Dekatron cathode voltages are clamped between -20V and +12V.

Oscillator cuitcuit and voltages are shown in the schematics.

Regards, Roland

[Dekatron42]

@Roland: Do you know what the frequency is at pin 6 (Anode) of the triode V1B?

/Martin

[Roland]

The Anita runs at 3.6...3.8KHz depending how accurate you tune this oscillator on V1B.

This is a square wave (more or less). The GS10D can run at 20KHz sine, but square is quite lower(10KHz).

A GS10C should be able to run at 4KHz square, that is why I bought such a tube, just to test...

The GS10C lights up but doesn't move at all... So I thought, let's check the current trough the original tube.

The GS10D runs on 1,2 or 1,4mA (not sure any more) but that current was even above spec for a GS10D.

This ETL datasheet book comes from Sumlock, so they should have known about it :-)

To be honest, I think Sumlock needed the cathode current to be high enough to use in the machine.

The P0 to P8 pulses are used to scan the keyboard, P0 and P9 are used about everywhere in the machine...

That means a quite high cathode current load on the tube. So optimizing the oscillator would be great

if that makes a GS10C run, but lowering the anode current might be tricky...

[Dekatron42]

Thanks for the frequency information!

You need to increase the value of C5E and C6E to increase the pulse length and increase the bias voltage on R9E & R10E to match values in driving circuits for the GS10C. From the voltage levels shown in your second photo I guess that the pulse voltage is enough (datasheet shows lower pulse voltage for a GS10C), but otherwise R1E & R2E can be adjusted accordingly, hopefully without having to adjust anything else. Page CT-3 here https://www.die-wuestens.de/rd/DEKA.pdf shows the values used for a "selectors" in the right column of the table that ETL recommended. The reset voltage and other cathode voltages shown in the circuit diagram looks to to be ok to me. However the cathode current might be a problem, the simplest way I can see to fix this would need an adapter board where you can use 10 trigger tubes driven by each cathode of the Dekatron to reach enough current. J. B. Dance shows how this can be implemented on page 102 / Fig 4.28 of "Electronic Counting Circuits" for diving a Nixie but that circuit would need some adjustments of course. Or you can use subminiature triodes but that would need more heater power which might not be available from the power supply. Making an adapter board would require the base from a dead GS10C or GC10D or other similar dekatron plus an extra socket for the same dekatron, that might be to much to stuff into the case of the Anita unless there is a possibility of putting the trigger tubes and base into the socket for the GC10D and have wires run to the dekatron and socket which you put somewhere else.

You can experiment with this by just running two wires from the two sides of R2E, from the +495V Anode voltage and GND to a breadboard outside the Anita and then just add the necessary components there and hook up a GS10C there. If you get this to run properly then you will have to experiment with how to hook up the cathodes to deliver enough current. The GS10C with these modifications in place might survive for some time, it all depends on how much current is drawn.

I hope this will help you to get a GS10C to work in the Anita.

/Martin

[Roland]

Hi Martin,

Thanks for the information. At this moment I don't have the time to experiment with it.

Getting the GS10C to run is one thing, but to let the machine work with it is the next challenge.

Would you think the GS10C can run on a high current like 1.2mA too? Sumlock pushed

the GS10D really to the max. If the machine could run on a GS10C Sumlock would have

done that because they were much cheaper. Otherwise a simple emitter-follower transistor

on the P0 and P9 cathode might do the trick to get the right current...

To make an alternative for the dekatron, in a probably not working machine, might be quite problematic.

So for now I think JP just should get a working GS10D which is not impossible in my opinion...

Regards, Roland

[Jean-Pierre G]

Hi all,

I still follow this thread with attention !

For now, as Roland says, I should go on looking for a GS10D.

This will be my Grail Quest !

JP

--
You received this message because you are subscribed to a topic in the Google Groups "neonixie-l" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/neonixie-l/1I-rkLzor24/unsubscribe.
To unsubscribe from this group and all its topics, send an email to neonixie-l+...@googlegroups.com.
To post to this group, send email to neoni...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/ea0858f9-a9dc-4ad3-8705-0a216f598cb1%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[threeneurons]

On Wednesday, June 20, 2018 at 1:32:06 AM UTC-7, Roland wrote:

Would you think the GS10C can run on a high current like 1.2mA too? 

For short durations, an hour or less, every once in a while, I think 1.2mA will do little harm. I've run orange neon dekatrons, designed for 350uA, at 1mA, in the past, and they run fine. Though, I haven't run speed tests, to see if that's affected. Obviously, if you do this 24/7, the cathodes will prematurely erode. The excess sputtering may even cause shorts. Maybe an extended test, for me (or anyone interested), to run on an OG-4 !?