AC Characteristics of Glow Tubes?

[dsavitsk]

Has anyone done any experimentation with the AC characteristics of glow tubes? Or know of any datasheets that might indicate this? I am looking for how linear they are, frequency response, impedance, etc. Anything that might be useful.

[dnsey]

Years ago, I discovered by accident that glow tubes (small neons, anyway) are photosensitive, so you'll need to specify the ambient light intensity for the characteristics you want.

[EC8010]

The data sheet gives the slope resistance (an AC parameter); it won't change with frequency. However, the striking voltage may change slightly with ambient light, radiation, FTSE index.

[dsavitsk]

Originally posted by EC8010
The data sheet gives the slope resistance (an AC parameter); it won't change with frequency.

Sorry to be dense, but I can't find this anywhere -- I did find the shipping weight, but no slope resistance. Do you have a specific datasheet in mind?

[EC8010]

If you download the 85A2 data sheet from Frank's site, you will find that although it doesn't give the slope resistance explicitly (it does on my Mullard data sheet), it gives a graph of Va against Ia, allowing you to measure the slope at a given current to determine the slope (AC) resistance. It also gives a figure for the noise.

[dsavitsk]

That explains why bypassing a glow tube that is in the signal path with a small cap helps. I tried it without the cap, and it sounded okay, but was a little flabby in the LF response. I am not sure why the small cap helps with low frequency.

I am using this schematic



and I think I am not sure how to think about the impedance that the cap sees. Obviously, the cap can't be larger than 0.1uF. With an ultrapath cap, it is typically a couple of orders of magnitude larger, but I am not noticing any low frequency roll off even with a 47nF cap -- maybe I just can't hear. I still need to take some measurements, but I could use some help in determining this.

[EC8010]

The capacitor sees the slope resistance of the 0D3, which is pretty low (100 Ohms?). In that context, it's no wonder you don't hear a low frequency roll-off even with 47n - the 0D3 is very nearly a short circuit even at DC.

[dsavitsk]

I figured about 300, but, whether 100 or 300, this is why I am confused. I am assuming that this impedance, whether 100 or 300, is going to be multiplied by mu and added to the rp of the tube. And this I would think would account for the flabby bass as mu is pretty high. But it wouldn't account for a 100nF cap seeming to fix the problem. At any rate, I think it is time to do a few measurements -- a good excuse to buy a new sound card I suppose.

[kstagger]

my Audio Cyclopedia mentions using gas tubes in place of capacitors -

"...Instead of the usual coupling capacitor, the control grid of the second stage is fed from the plate of the first tube through a small neon tube. Voltage regulator tubes, as a rule, are not satisfactory because of their high operating potentials. The purpose of such coupled amplifiers are to reduce the phase shift generally associated with resistance-coupled amplifiers at the extremely low frequencies. Gas coupling is only practical at fairly high signal levels because of the internal high noise level of a gas tube. The frequency response and stability may be improved by the use of negative feedback by eliminating the usual cathode-bypass capacitors." - Tremaine / 1975


I know - not very useful for what you are looking for!

[dsavitsk]

Maybe this will help think about this -- this circuit should work the same. In fact, maybe it will work better and I should just do this instead. I have not figured out what the second tube should be yet. It could just be the same tube, but I don't see why it would need to be ... on the other hand, have I just drawn up a primitive push-pull amp with single ended iron?