Cooling of Vacuum Tubes

[PRR]

> What advantages would cooling of tubes have? Would it cause less noise?

If you could cool the cathode enough to reduce noise a measurable amount, it would just stop working at all. The cathode MUST be hot to make those clingy little electrons jump off where the plate and grid can affect them. And emission is like 3rd-power of cathode temperature, while noise changes very slowly.

If you are over-heating the plate, cooling will increase life. However small glass tubes don't have enough thermal contact between the plate and any part you can put a cooler onto. Some huge (kilowatt) transmitter tubes do have provisions for cooling the plate directly. But don't over-do it: the transmitter tubes are designed to run hot and won't getter well if you manage to cool them too much.

If you work the tubes very hot and turn them off and on many times a day, the seals may crack. Cooling around the pins is usually something to check if your tubes go gassy quickly.

> drilling some holes in the chassis near the tubesockets and use some fans

Do as much of that as you like. It will keep the resistors and other parts on the socket from cooking and drifting over time.

[EC8010]

Interesting post, PRR, and I agree with most of your points.

However, (you knew this was coming), noise is only signficant with small-signal valves, and it doesn't solely come from the shot noise of electron emission. There are also some distinctly dodgy insulators with considerable voltage across them that are connected to a high-impedance input (the control grid). The micas.

The anode cools by radiation, but some of this is absorbed by the envelope, raising its temperature, and simultaneously reducing the temperature differential between the anode and envelope, which reduces anode cooling. The micas are in close thermal contact with the anode, and insulators become leakier as temperature rises. Alternatively, it could be that hot micas begin to release water vapour, causing grid ion current.

I accidentally proved the noise theory by fitting snug-fitting screening cans made of 22mm copper pipe with a slot in the side to B9A valves. After half an hour, there was a lot of LF noise, and the valves were hot. I removed the cans, and the noise subsided. The cans, being relatively shiny, had prevented radiation.

[fdegrove]

Hi,

Interesting post, PRR, and I agree with most of your points.

So do I... but there's more to tube coolers than that; lowered microphony is one thing.

If we'd want to extend tube life regardless of top performance we could just as well lower cathode emission by reducing heater voltage/current.

Cathode emission is not going to be affected by a cooling element on the glass envelope, so I seriously doubt that in this application it's going to affect this in any negative way.

I accidentally proved the noise theory by fitting snug-fitting screening cans made of 22mm copper pipe with a slot in the side to B9A valves. After half an hour, there was a lot of LF noise, and the valves were hot. I removed the cans, and the noise subsided. The cans, being relatively shiny, had prevented radiation.

Yep....I found the best screening cans to be matte in colour, i.e. none reflective.
There's a lot more that can be said about this, we're entering muddy waters here.

Cheers,

[Gunders]

Originally posted by ShiFtY
As for reduced noise, have you heard how loud computer fans are? They suck, I think any decent music system should be fanless.

Yes, for Hifi equipment the noise from the fans will be annoying. But what about guitaramps?
Well the noise from the fans would probably still be annoying when you're playing at low levels, but in many(or maybe most) situations guitarists would like to play rather loud.

[Gunders]

Well, I don't know if guitarist care much about noise(tube noise) in their amps when they are playing loud, but cooling could probably extend the life of the tubes.

I have searched a bit for some low-noise fans, and I found a couple which have a noise of 12 dBA, thats substantially lower than the noise produced by computer fans.

[jevta]

First, you have to know reason for cooler. Amps with baby tube EL 84,EL34 , EL 88, don't need cooler.Big tube like 6c33b need cooler for socket and area near .I made on my OTL 6C33B two cooler 12v 12 W with 25 db noise.For put down this noise very efecty put down voltage on minimum 6 V.

[fdegrove]

Hi,

First, you have to know reason for cooler.

Sure...nonetheless a reading of Bill Perkins papers is worthwhile...

There's more to a tubecooler than just temperature.

Ciao,

[jevta]

Really? I dont care this fact. If that had rapid influence only on sound please tell me adress where I can learn about that.

[fdegrove]

Hi,

If that had rapid influence only on sound please tell me adress where I can learn about that.

You'll find all the appropriate info in post #10 by Planet10.

Another good reason for using a wel designed tubecooler is that it reduces the effect of airborne and structure borne vibrations.

Cheers,

[planet10]

Originally posted by fdegrove
Another good reason for using a wel designed tubecooler is that it reduces the effect of airborne and structure borne vibrations.

Especially if used in conjunction with an iso-socket. Parts Connexion talked Bill into redesigning his isolation socket so that it plugs into the existing socket which means they can be installed without soldering (althou at the expense of another contact)

dave