threeneuron's Dekatron clock

[Phill Scan]

Hi all,

Just to introduce myself I am a retired electronics controls tradesman (industrial process control) with an A grade Australian electrician's ticket tacked on (also industrial). 

I have two trades .. but am not an engineer... not too flash with theory but can fix most things.

I have been programming since 1977 on and off.

Know a little about a lot. Used to repair TV's and VCR's when the betamax VHS war was in full swing.

Familiar with just about everything used in process control including pneumatics and some hydraulics, even fluidics .. now there is an interesting field, and you think just mechanical and electronic devices can have logic elements in them !

Jack of all trades master of none really.

I like C++ the rest can go to hell :-) too old to change now ... and what little assembler I knew is long gone.

SO.....

Thanks to threeneurons blog I have now got the dekatron bug.

I ordered his Do-hicky kit today .. so I could test them out and look around at the timings and voltages etc.

I already have a bunch of Russian dekatrons.

A101's seem easy enough. And I think I will build a clock with them.

I also have 5 each of the OG8 and OG3. (the purple one's that people say are flaky) 

I have never seen a circuit with a high enough voltage to drive the OG8 though, so that will be a challenge.

Threeneuron's Do-Hickie seems to be able to do the OG3 though .. so I guess I will soon find out if I have bought dud tubes.

Anyhoo..... these things (OG8) look like they are a real bugger to light up from the look of the data sheet (Russian)

My guess is that it requires 500V typical and 480V min to light up these suckers. ?

And guessing again from the datasheet it looks like it takes 1.3mA Max and 1.1mA typical to run them. ?

The HV supply from threeneurons with the voltage multiplier looks like it might run it ?

The Main question I have though and perhaps the man himself might want to answer is what PPM signal is coming out of this circuit.

three neuron's dekatron clock circuit.

Pin PD.6 seems to be doing all the heavy lifting for the charge pump and then reading back in for voltage reference on AI+

So... what timing would I need to use to get the +220V for that circuit from an Arduino ?

I like the idea of running the clock with an Arduino because I have some of them and like the resources they have for them.

Any Arduino code for running Dekatron's would be very gratefully received.

My guess is that I would sample on the analog input less frequently per clock cycle so as not to get into a race condition or the like.

Cheers,

Phill 

[Dekatron42]

The OG-8 dekatron is not hard to drive, just use the circuit diagram in the datasheet and the other parameters found in the datasheet and they will step just fine. However some of the OG-8 dekatrons that are for sale are partially outgassed so they either won't step properly or sometimes at higher voltages. I haven't bought any OG-8 dekatrons recently but last time I did so, maybe some 4 or 5 years ago they all worked fine even though other people were commenting on them being hard to get to work. I was sent one such troublesome tube from a friend and it was definitely outgassed as it needed higher voltages to strike and it didn't step as easily as mine did, it also had a very different glow. I have read of the same problem, and also experienced, with other high speed Russian dekatrons and in my case it was all due to them being outgassed. The sellers I bought from then immediately refunded me for the outgassed ones, and I bought some more to make sure that I had properly working ones. These problems mostly occurred with the A-106 dekatrons.

/Martin

[Phill Scan]

Hi Martin,

Yes I have heard that a lot of the argon Russian dekatrons have leaks.

I am trying to get a tester asap so I can test the ones I have already bought.

Probably thought I could just make a circuit quicker myself but time is a rare commodity here.

Thanks for your thoughts.

What frequency do you think the link above is putting out to drive the HV ?

[Keith Moore]

Yes. I blame almost all of my relatively recent addictions to the mighty Mr.ThreeNeurons. 

I love my purple dekatrons and use them in my spinners and clocks at-will. I just treat them with kid gloves once I get them. They are cheaper but far susceptible to failure. I love my purple dekas, though. Don't be afraid. Just be aware.

 

- Keith   

[threeneurons]

On Tuesday, February 2, 2016 at 2:06:33 AM UTC-8, Phill Scan wrote:

Hi all,

Just to introduce myself I am a retired electronics controls tradesman (industrial process control) with an A grade Australian electrician's ticket...

I also have 5 each of the OG8 and OG3. (the purple one's that people say are flaky) 

I have never seen a circuit with a high enough voltage to drive the OG8 though, so that will be a challenge.

Threeneuron's Do-Hickie seems to be able to do the OG3 though .. so I guess I will soon find out if I have bought dud tubes.

...

Cheers,

Phill 

Welcome Phil,

Dekatrons are enchanting ... if you can say that about a tube.

The OG-8 is a real pain in [backside] ! I had two. One died, and the other I can't locate. Hope I didn't ship it off to a customer, as I was reselling OG-4s for a while, that I bought, in bulk, from Russia.

The OG-8 will not run in the DoHickie. It may light up, but the glow will just sit there. The OG-8 is a single guide dekatron, and the DoHickie only runs 2 and 3 guide dekatrons. The OG-3 will work just fine, as long as it still has gas.

I, personally, am not a fan of the "high speed" dekatrons. Those are the ones that usually glow blue. There are also high speed tubes that glow a funky yellowish pink. Those are the 6909, 6910, & Z505S. The blue (violet) ones are: OG-3, OG-8, GC10D, & GS10D.

All "high speeds" have a very high dud rate. I'm not sure if the gas in the "blue" ones is argon. I'm leaning towards hydrogen or helium. If its one of those two, then that would explain the high dud rate. Hydrogen will eventually react with the metal bits and form a hydride. Helium can actually diffuse thru solid glass. In either case, in time ... no gas, no glow.

[John Rehwinkel]

> All "high speeds" have a very high dud rate. I'm not sure if the gas in the "blue" ones is argon. I'm leaning towards hydrogen or helium. If its one of those two, then that would explain the high dud rate. Hydrogen will eventually react with the metal bits and form a hydride. Helium can actually diffuse thru solid glass. In either case, in time ... no gas, no glow.

However, it's helium at very low pressure. I wonder if the helium ones can be brought back to life the same way helium-neon lasers that have lost their helium are: put 'em in a trash bag full of helium at atmospheric pressure. Since the helium is at much higher pressure, it'll diffuse back in much faster than it diffused out. Years-old dead lasers can be brought back in a matter of days. The tricky part is to get the right helium pressure back. With the lasers, it's easy, you must measure the light output, wait for it to peak and just begin to decline (too much helium), then they'll be good for another few years. With dekatrons, I'd guess some funky sequence of drive voltages, and watch the operation to see when it begins to step at optimum drive, then step with less and less good waveforms.

- John

[Jon]

Indeed welcome Phill ! You are starting to slide down a very slippery slope indeed with the dekatron fascination. They are quite compelling tubes to play with, and the addiction gets worse if you start to collect the different flavours, look into the history etc. You're doomed mate. :)

On Friday, February 5, 2016 at 6:45:04 PM UTC, threeneurons wrote:

All "high speeds" have a very high dud rate. I'm not sure if the gas in the "blue" ones is argon. I'm leaning towards hydrogen or helium. If its one of those two, then that would explain the high dud rate. Hydrogen will eventually react with the metal bits and form a hydride. Helium can actually diffuse thru solid glass. In either case, in time ... no gas, no glow.

The hallmark of the fast dekatrons is hydrogen, first as a minority component of a gas mix in the 'mid-speed' tubes (say 10kHz upwards), and increasing in proportion generally as the max speed increases. The fastest tube types are just hydrogen. The bit of physics that limits the speed of glow-transfer is how fast you can deionise / de-energise the gas around the electrode that the glow just departed, and hydrogen acts as a quenching agent to speed this up. I've done quite some research into the dud problem, and I think I've pretty much convinced myself that it's bad chemistry rather than a physical leakage - will write up the arguments around this at some point.

Having looked at a lot of different dekatrons (my research collection is north of 600 now - eek!), the dud problem is quite variable across manufacturers and specific tube models. The Elesta tubes are generally pretty good - they specialised in hydrogen fills and it seems knew how to get it right. At the other end of the scale the tube that wins the prize for being dead most often - the Norwegian Blue of the dekatron world - is the Mullard Z505S. You might see it marked under other brands of the Philips empire, but they're all made by Mullard at the Mitcham factory. And they're almost all dead now sadly.

Cheers,

Jon.

[Phill Scan]

Thank you all for the replies and welcomes, I forgot how to find this group again after my computer had a Chernobyl event. But found it again at last !

Given that this tube is such a bugger to get going and it is not much use to me for a clock. I will be selling the one's I have on ebay.

I will test them all before I list them.

The data sheet that was given to me  by vannadiy, my go to ebay seller who I have bought hundreds of idems from has kindly translated (sort of) the data sheet which can be found here.

http://www.155la3.ru/datafiles/og8.pdf

His translation for the OG-8 is as follows.

OG-8
Electrode leads: 1 - cathode is zero; 2 - not connected; 3-anode; 4 - subcathode; 5 - the fifth cathode; 6 - the ninth cathode; 7th
- the third cathode; 8 - cathodes (general conclusion)

Not sure what it all means really, no idea what he means by a subcathode for example, so will just poke each pin individually to test.

Threeneurons Do-Hickie does indeed work the OG-3's perfectly.

Thanks again for the help and information,

Cheers,

Phill

[Jon]

Subcathode is the literal translation of the Russian electrode designation 'подкатод' which is how they refer to guide or transfer electrodes. Since the OG8 is a directional cathode tube, it only has a single transfer electrode rather than the pair of guide electrodes that you'd find on (eg) GC10B.

Jon.

[Phill Scanlon]

Thanks Jon,

Do you know which one is the K0 or Index ?

[Phill Scanlon]

Thanks Jon,

I think I have made a good translation of the original.

Could you please check ?

I think the cathode on pin one is the index ?

On Thursday, November 9, 2017 at 6:35:16 AM UTC+8, Jon wrote:

[Jon]

Pin 8 is better called the common cathode - it's connected to all the main cathodes except K0 (pin 1), K3 (pin 7), K5 (pin 5) and K9 (pin 6).

I guess K0 is what you might instinctively call the index cathode, but because the array is circular it's a bit of an arbitrary designation. For any of K0, K3, K5 and K9 if you start with the glow on that cathode, conduct 10 stepping cycles you'll wind up with the glow back on that cathode. So if your application just involves using the 'index cathode' as a reference point from which to count steps or complete revolutions, it doesn't really matter which of these four you use for that purpose. The most logical one to choose may depend on physical considerations such as the orientation of the tubein its mounting and therefore which of those cathodes is closest to 'up' or whatever rest position you might want to use.

This type of tube with some but not all main cathodes brought out onto the base is often referred to as a 'computing counter' as distinct from a 'counter' (one index cathode and nine commoned cathodes) and 'selector' (all cathodes individually available on the base). The ETL datasheet for GC10/4B has a useful table that shows how you can use this type of tube to count any number of pulses from 1 to 9.

Cheers,

Jon.

 

[Phill Scanlon]

Thanks for the extra information Jon,

From the translated datasheet I uploaded I have now named pin 1 index cathode.

Am I correct in calling it a uni-directional tube ?

If this is the case then why are there 3 "extra" cathodes ?

I am just trying to get the data sheet right. Then try to build a spinner for it.

Cheers,

Phill

[Phill Scanlon]

So to be clear,

I can use any of the K3 K5 or K9 cathodes and it will work ?

[Jon]

There is no electrical difference between K0, K3, K5 or K9, just a topological one with respect to the other electrodes in the array. Choose any one of them as your nominal index cathode, treat all the others in the same way as the common cathode K and you'll get a tube which spins giving one output pulse from the index cathode every revolution.

OG8 is indeed a unidirectional tube - the shape of the cathodes (main cathode and transfer electrodes) dictates the directionality. That's how you can get secure glow transfer with only one transfer electrode between each main cathode.

Jon.

[Phill Scanlon]

Thanks for clearing out some of the mud in my eyes Jon,

To summarise. The only pins which MUST be identified are the Anode, the Guide and/or Guides.

My understanding of the dekatron is that the index or K0 pin is NOT connected to any other pins.

The only reason I say this - as I have never taken one apart - is from this gif file I have attached.

Guessing that the gif is is over simplified.

Correct ?

[Jon]

On Friday, November 10, 2017 at 12:22:10 AM UTC, Phill Scanlon wrote:

To summarise. The only pins which MUST be identified are the Anode, the Guide and/or Guides.

 

Well, all dekatrons do have at least one main cathode brought out to the base individually in addition to any commoned main cathodes and it is necessary to distinguish this one if you actually want to be able to count using the tube or know where the glow is. However if all you want to do is spin a tube continuously, then for most basic dekatron types I guess you're right. Identify the anode and guide(s), apply the appropriate inputs to those, ground the other pins as cathodes and it'll work. However the situation does get more complicated for tubes like GC10D, GS10H where some of the guide electrodes are brought out individually to the base. Get the hang of things on the basic tubes first and then you should find it easy to deal with the extra bells and whistles of those tubes.

 

My understanding of the dekatron is that the index or K0 pin is NOT connected to any other pins.

The only reason I say this - as I have never taken one apart - is from this gif file I have attached.

Guessing that the gif is is over simplified.

Correct ?

The GIF is spot on for a two-guide counter dekatron like GC10B or OG4. And yes, the index or K0 pin is not connected to anything else in that whole array of electrodes circling the anode, just the individual electrode as shown. Some manufacturers call this the 'output cathode'. In a selector dekatron like GS10C or A101, every one of those purple electrodes that are connected together in the GIF as K1-K9 (RTN) is not connected to anything else and brought out to its own pin on the base. In a computing counter like GC10/4B, some of those electrodes have their own pin on the base, the remainder are internally connected to make a common cathode array like in the GIF.

Your OG8 is therefore different to the GIF in 2 ways. First K3, K5 and K9 don't connect internally to the common cathode line - they are just brought directly out to the base, exactly like K0 is shown here. Second, it has a single directional transfer electrode array instead of the two guide electrode arrays, so you need to imagine all the orange coloured pins and connections in the GIF as being absent.

 

Jon.

[Phill Scanlon]

You have done a great job of leading this blind man into the light.

I have finally got it !

Once Threeneuron's arduino sheild finally gets here I will practice coding with that. Then I will breadboard a circuit for the OG-8 with the correct pinout. The OG-3, which does work in his doohickie spinner, will not work with that shield.

Once again Jon you have been of great asistance and is very much appreciated.

Finished datasheet attached.

[Jon]

On Saturday, November 11, 2017 at 10:45:00 PM UTC, Phill Scanlon wrote:

Once again Jon you have been of great asistance and is very much appreciated.

No problem - happy to help. Enjoy playing with your dekatrons!

Jon.