Correct way to drive IN-28 "nixie dots"?

[Mark03]

I acquired some of these tubes on Ebay from the Ukraine, and plan on using them as the seconds flashers (colon) on a four-digit nixie clock.  I gather the recommended current is somewhere around 10 mA and the strike/hold voltages are a little higher than most nixies.  My supply is 280V so this shouldn't be a problem.  I found this page, which was of some limited help:

http://www.jb-electronics.de/html/elektronik/nixies/n_in28.htm

They have an anode, a cathode, and a grid, all brought out to separate leads.  I tried leaving the grid unconnected and just used the anode and cathode like a neon lamp, but my tube wouldn't reliably strike.  In the process of probing it and measuring voltages, etc, I must have touched something and it suddenly lit up.  I could then turn it off with the drive transistor but could not turn it back on.  So either the strike voltage is higher than I thought, or I'm supposed to be doing something with that screen.  Am I supposed to bias it somewhere in between?  What's it for anyway?

Hoping someone here has used these tubes successfully and can give me a few pointers...

Thanks,

Mark

[Mark03]

Some more research, some more questions...  Is this tube essentially a cold-cathode thyratron?  And could two (or more) be connected together in a ring-counter configuration?

[Mark03]

To answer my own question:  Yes, AFAICT the tubes behave just like what are called thyratrons.  Looking at the data sheet (translated with google's assistance), the anode-cathode strike voltage is 340V minimum, but the grid-cathode trigger voltage is 240V maximum.  (For my tubes it appears close to 175V.)  It only takes 50 uA max into the grid to trigger the main anode-cathode discharge, with an anode-cathode maintaining voltage of 155-175V.  So the simplest way to run these is to connect a ~ 1M resistor from the anode to the grid, then drive like any other neon bulb.

I had some free time today, and two IN-28s, so I decided to try building a ring counter---in this case, just A, B, A, B, ...   It is based on the ideas developed here: http://www.dos4ever.com/ring/ring.html --- a true gold mine of information!

Here is a quick-n-dirty schematic:  http://www.keteu.org/~haunma/neon_osc.pdf   The basic principle is the same as on the page above, but I have added biasing for each grid.  Those resistors will have to be adjusted depending on one's particular tubes.  I needed separate bias networks because mine were quite different in trigger level.  But the sum should be ~ 1-2M.  You want the bias so that once one tube is lit, the drop across the anode resistor keeps the other one from also triggering.  The capacitors may be larger than necessary, but I was worried these tubes, being larger than little neon bulbs, might need a few milliseconds for deionization, to keep the same tube from firing over and over.  In any case, after adjusting the grid biases, it seems very reliable.  Just pulse the open-drain (or open-collector) switch on and off briefly, and the tubes alternate.  I tried it over a supply range from 280-310V.

Hope this helps someone.  The tubes were pretty cheap on Ebay.

Mark

[Mark03]

Sorry, wrong link.  It should be http://www.keteu.org/~haunma/in28_osc.pdf.