Forced-air tubes

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Quince

Well-known member
Joined
May 31, 2006
Messages
68
Location
Vancouver, BC, Canada
Interesting way to avoid the blowers one would normally use with tubes like 3CX300A1:
http://www.vongaylordaudio.com/product_html/productviewuni.htm
I didn't figure out what the liquid is, probably some sort of flourinated fluid. Which I'm sure contributes to the $50K price tag. Still, it makes me wonder if this may not be the type of solution to try with my 4X150A, as I've yet to manage to quiet the centrifugal blowers. Any ideas what the cheapest appropriate fluid would be?
 
If you need to use water-cooled transmitter tubes in your audio amplifier, it could be time to look for more efficient speakers.

Anyway, I don't think they use anything exotic in transmitter service... I think we may have used regular glycol solution to cool the klystrons in our UHF transmitters. But I wasn't directly involved with that so I can't say for sure.

This document recommends deionized water.

http://www.cpii.com/mpp/company_info/PDF/AEB_17.pdf
 
I need transmitter tubes since I'm direct-driving high voltage (plasma). Though glass-envelope TV horizontal sweep tubes can do that voltage, they have very little current.
When I saw that photo, I seriously started thinking if this may not be a better solution than the centrifugal blowers, which make a hell of a lot of noise.
 
Hi there,

even never used for transmitting tubes, most of the transmitter I know about used steam cooling or distilled water as a heat transfer agent, plain old transformer oil ( used for power transfer x-former), or even Diesel (used as an insulator and heat transfer in spark erosion machines) will work more stable and generate less problems dealing with plasma amplifier than water.
 
8560AS.jpg

8560AS
They make a conduction-cooled version of a 4CX250.

Just get a big, honking heat sink.

more: http://www.diytube.com/phpBB2/viewtopic.php?t=626
 
> centrifugal blowers, which make a hell of a lot of noise.

> used steam cooling or distilled water....,

There are several ways to cool over-hot tubes. Each way requires a different tube.

Good old radiation through glass.

Run water through the plate. Plumb it like a car. Since the plate is high voltage and safety/sanity demands the pump and radiator/drain be safe to touch, you put in a couple lengths of glass or ceramic pipe, and use highly pure water (which is a remarkably bad conductor).

A variation brings in liquid water and flashes it to steam. The inlet pipe can be a little smaller, the outlet has to be much larger. I'm not sure that steam allows a higher power density than pressurized water. Have to hit the old boiler-books to figure that out. The main advantage may be that you could run a steam engine on the exhaust, which may be more entertaining than most music.

Fins and fans. Lots of air. Since these rigs are usually out in a shack, they design so that rated dissipation can only be reached if you blow until it whistles. Sometimes you can de-rate and blow less hard.

Conduction cooled. Like transistors. You are responsible for carrying the heat away. Huge passive heat-sink, sink+fan, water-block as used in hot-rod PCs.... your choice.

> plain old transformer oil .... even Diesel

Never heard of using oil. It could work. Oil won't generally take as much heat as water, so you need to check that before oil-cooling a tube made for water-cooling. Oil that has been exposed to air is not a great insulator. At kilovolts it breaks down, makes gas, leaves carbon-tracks to encourage more arcing. In big transformers, you fill with oil, short the HV winding, put a low voltage on the LV winding, and let it cook a few days to boil the water out. That's why PCBs were such great stuff: they didn't have as much trouble with water.

> what the cheapest appropriate fluid would be?

For an air-cooled tube: air. Liquid Nitrogen would crack the seals, most liquids would give local over-cooling and stress. Just blow on it. If you need to blow quietly, use a bigger tube and de-rate, or don't air-cool. It's like the VW Beetle: great in many ways. Cheap, lightweight, don't have to drain it in the winter, won't leak on the carpet. But if you have a lot of power in a small space, the fan will be loud.

For a water-cooled tube: nothing beats water. Carries the MOST heat. Not free: it's gotta be very pure, the plumbing has to be very good. If you get things really quiet, you need to add alarms: water stops, tube dies fast.

Conduction cooled, IF you can get the appropriate insulator, has promise. A practical passive heat-sink may be obscenely large, but a water block and car heater core could work with plain water (add a dose of Prestone for anti-corrosion). Or even a dead-loss system, trickling water from domestic plumbing to a drain.
 
Of course, I could run blowers outside and use a duct.
I need to cool a pair of those: http://www.df6na.de/surplus/tubes/4x150a.jpg
They're very small so shouldn't be hard. Probably can manage with 1 L flourinert total, and though flourinert is very expensive new, I often see old equipment on eBay that contains some.
 
What about a Peltier effect device mounted to a heatsink? Silent and can get down to freezing point. Surely that's enough cooling.

If the tubes are cylinders (i.e. not bulbous) you could get a block of aluminium, drill out holes suited to the OD of the tubes and mount the peltier device on the side. You'd probably have to use a heatsink on the hot side of the peltier.

For better heat conduction, you could submerse them, but mount the tubes upside down so no liquid contacts the pins:


tubecooling.gif


Just a thought.
 
What a bunch of wimps you all are! <g> Goto:

http://www.cpii.com/eimac/

and look up the specs for a 4CV250,000B.

Length 28.02 in; 17.17 cm
Diameter 15.06 in; 38.26 cm
Weight 180 lb; 82 kg

Filament/Cathode Thoriated Tungsten

Voltage 12.0 Volts
Current 660 Amps

A pair of them running in Class AB1 can produce 660 KiloWatts with 20,000 V at 40 Amps on the plate supply.

Now THAT'S a tube!! <G!>

Bri
 
That's nothing compared to an 8974/X-2159.
http://www.cpii.com/eimac/catalog/169218.htm
Output power: 2,158,000 W
Drive power: 7000 W
Filament current: 600 A
And it's a mere 80 kg, compared to the 82 kg of the tube you pointed out that does a fraction of the output power :p
 
[quote author="Larrchild"]
8560AS.jpg

8560AS
They make a conduction-cooled version of a 4CX250.

Just get a big, honking heat sink.

more: http://www.diytube.com/phpBB2/viewtopic.php?t=626[/quote]

Not only do you need a big heatsink, but you also need Beryllium Oxide insulating blocks in between the tube and the heatsink. BeO dust is a known carcenagin.

-Bruce
 
Actually, it dosn't need to be a carcinogen. BeO is highly poisonous and the dust will kill you in short-order.

I have a pair of these on 144 mhz pumping out 5-600 watts.
and I'm EXTREMELY careful, when handling the insulator-block.
 
[quote author="Quince"]That's nothing compared to an 8974/X-2159.
http://www.cpii.com/eimac/catalog/169218.htm
Output power: 2,158,000 W
Drive power: 7000 W
Filament current: 600 A
And it's a mere 80 kg, compared to the 82 kg of the tube you pointed out that does a fraction of the output power :p[/quote]

OK, your TUBE is bigger than my TUBE...

LOL!

Bri
 
> BeO is highly poisonous and the dust will kill you

I have a quad of the JBL 2440 horn drivers with the short-lived Berillium diaphraghms. "Short" in the sense that JBL stopped making them almost as soon as they started. The Be makes a stiffer diaphraghm than the Aluminum everybody used since 1923. With computer-aided designed phase plug, and CD horn throats, it was another half-octave of bandwidth, eliminating super-tweeters for most work. And then they realized that a half-octave is not worth dying for. (Berillium's toxicity was sorta-known, but some specific government project disaster about that time really put the lid on Berillium fabrication.) JBL fiddled with Aluminum again, then Titanium.

Surrounded by 30 pounds of Alnico, I'm not afraid of handling the drivers. I did open one to look at the Berillium, but held my breath and sealed it right back up again. Looks like Aluminum with a faint bronze-tone anodize.

> What about a Peltier effect device...?

If Quince is running to the full 150W rating, it will be hard to find a Pelt which can move that much heat at any temperature differential. It's a bit more than the Pelts we use for CPUs. A Pelt is good for pulling something below ambient, but we don't need that. We just have to keep the seals from cracking, the grids from warping. It can be mighty hot, we just have to limit the rise. Huge amounts of ambient will do that.

> run blowers outside and use a duct.

The fan is a noise source, true, and within limits you can remote the fan.

But IIRC, the required velocity over the fins is a sizable fraction of the speed of sound. And those fins are sharp-edged, and surely run turbulent flow. Seems to me they are bound to make noise.

How nasty is your electrical signal? A 40V 5A signal for a common speaker, you can put the amp out on the porch and bring wires in through a towel-stuffed window. But it sounds like you are working far above 300V, which is a very different level of danger. OTOH, high voltage low current means negligible copper loss, not like wrestling #10 wire for 8 ohm speakers to keep losses low.
 
[quote author="Larrchild"]Actually, it dosn't need to be a carcinogen. BeO is highly poisonous and the dust will kill you in short-order.

I have a pair of these on 144 mhz pumping out 5-600 watts.
and I'm EXTREMELY careful, when handling the insulator-block.[/quote]

I had forgotten about that. It's been a while, although I still remember the warnings:

Do not lap, grind, or chemically clean.....
 
Back
Top