EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: Alex on March 04, 2011, 01:44:47 am

Title: Looking for a jellybean transistor
Post by: Alex on March 04, 2011, 01:44:47 am
Hi, I am designing a switching circuit around a neon gas discharge tube.

This is a very common topology, common emitter NPN transistor in saturation.

Power supply is 145 V, striking voltage for the tube is 140 V worst case, and its maintaining voltage is 100 V. Current is limited by a 280 k resistor to 0.16 mA.

I am looking to select a transistor for this configuration. The only two criteria is excellent availability (local stores) and reasonable price (per piece). It needs to be in a THT package, TO-92 most likely.

I was wondering if you guys could mention a few transistors that keep on cropping up and are readily stocked in local stores or are otherwise easily accessible.

The only important parameter is the collector-emitter breakdown potential (Vceo). Normally you would select a device with significantly more than 145V, maybe 250V or 300V such as the MPSA42 which is typical in neon-based circuits. However a lower Vceo transistor can be used due to the nature of the circuit. The BC547/BC550 come to mind with Vceo of 45V.

Any other jellybean transistor that you are familiar with?

Many thanks.
Title: Re: Looking for a jellybean transistor
Post by: Zero999 on March 04, 2011, 05:33:41 pm
Did you do any experiments to confirm the BC547 will do?

Are you sure the leakage current won't cause the tube to strike?

Yes, on paper it looks like you should be able to get away with a VCEO of 45V but I think you should do some tests to make sure.

My favourite transistor is the BC337 because it has a high gain which is specified with a VCE = 1V
Title: Re: Looking for a jellybean transistor
Post by: mikeselectricstuff on March 04, 2011, 06:11:10 pm
MPSA42 is the standard jellybean high voltage switching transistor
http://uk.farnell.com/on-semiconductor/mpsa42g/bipolar-transistor-npn-300v-to/dp/1126473?whydiditmatch=rel_1&matchedProduct=mpsa42&matchedProduct=mpsa42&whydiditmatch=rel_1 (http://uk.farnell.com/on-semiconductor/mpsa42g/bipolar-transistor-npn-300v-to/dp/1126473?whydiditmatch=rel_1&matchedProduct=mpsa42&matchedProduct=mpsa42&whydiditmatch=rel_1)
It's likely be too much hassle relying on the neon voltage drop to use lower voltage transistors as you need to accurately regulate the HT supply.  With the right switching transistor the HT rail can vary +/-50V with no problem.

Title: Re: Looking for a jellybean transistor
Post by: Alex on March 04, 2011, 06:20:51 pm
Thanks Hero.

Thats the thing, I know 145V is the total,  I can't be sure as to how it is distributed across the transistor, resistor and tube.

In the Townsend region of a glow discharge I-V characteristic, just before the normal glow discharge, the current is in the order of 10nA to 1uA for your average switching diode. These diodes has a similar geometry and gas filling as the ones I am working with, so I expect the same currents to occur.

Looking at the BC337, Icbo at 20 Vcb is in the order of 100 nA, but, as expected, increased to uA at higher temperatures.

So I am worried that the maximum Vceo or/and Vcbo will be exceeded under the right conditions IF the tube doesnt strike from the leakage current as you pointed out.

The MPSA42 has a stated Icbo of 0.1uA at Vcb of 200V. This is much better than for say the BC337.

I know the MPSA42 works reliably, but have not tried 337 or 550. In fact, for the 550 Vcbo was found to be 3 times the datasheet value, 50V. How does that vary across devices with different doping levels? I dont know. The 550 has beed used successfuly, but using numerical 'nixie' tubes and not over the commercial temp range.

Seems to me that using a transistor with max Vceo/Vcbo less than 100V is pushing it.

I could set to measure currents in the nA range, not sure if it's worth it. Is it? lol

This is academic, but how about a very large value resistor across the transistor to limit the voltage across it?

*Just saw Mike's reply*

Thanks Mike. Regulating HT supply is not a biggy as it is usually generated with a boost converter using a switching regulator IC. Unless you are looking into powering from the mains direcly or 555-based designs with no feedback loops... This is powered by 12V DC then boosted with a regulator.

Anyway, I think your reply further reinforces the point that relying on voltage drops across other components in this case is in the realm of 'bad design'.

And a mini-rant on how some datasheets have NO graphs whatsoever   >:( ...