• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Heatsinks for tubes?

Does anybody know if it is possible to source heatsinks for tubes?


  • heatsink.jpg
    49.6 KB · Views: 1,681
Heatsinks on tubes may actually make them run hotter, since they intercept the normally radiated radiant heat and have poor thermal contact with the glass. Probably were used to keep the surrounding components from getting hot by converting all the radiant heat to vertical convected hot air instead. A fan could help, but that would likely work even better without the heatsink in the way.
  • Like
Reactions: 1 user
Morgan Jones wrote about valve heat sinks in Building Valve Amplifiers (The blue book)

I dont have the book to hand at the moment but I think I remember him quoting some source saying that just keeping the envelope temperature down can extend life.

I can see that radiated heat from the anode would be partly absorbed by the glass because it does not completely pass IR. Also the black anodising on the heatsink will absorb the rest of the IR and as long as it has enough surface area for conduction/convection to the air will prevent heating of other devices around it.

I think I saw some the same as in the book at Ask Jan First. The Pearl ones look good if a bit pricey.

What tubes are you cooling Wavebourn?


It looks better above SOA, on higher than specified voltage, with higher load resistance than usual.


Pearl Tube Coolers do in fact work, especially the newer ones (without the fibre sleeve).
Though you could also experiment with thin section aluminum extrusion which you can buy be the foot, cut it to desired length then hand roll them to dia. over an appropriately sized dowel rod (od a dead tube of the specific type). Use eithe Viton or Silicone "O" rings to hold to the tube wall.

Just to be curious: what magical thing are you designing with these wonderful tubes??? Please tell us it is a Wavebourn OTL!!!!!
Last edited:
Just to be curious: what magical thing are you designing with these wonderful tubes??? Please tell us it is a Wavebourn OTL!!!!!

No, it will be a home theater. :D
5 triode PP channels above 300 Hz, with small transformers for outputs. And a pair of class A+C SS channels for below 300 Hz. LC crossovers using modem transformers.
I dont have any 6s19 to look at and I cant find any decent pictures of the base.

Is the anode pin nice and chunky, or is it connected to the anode with a thin strip?

If it is directly connected to the anode you could add extra cooling at the pin and also use a heat sink on the envelope to try and keep the glass cool.

Only trouble is any failure due to long term overheating is likely to occur at the glass to metal seals of the pins. Take a look at the QQV06-40/GU19 for an example of heavily constructed seals.

What, DIY chainmaille sleeve wasn't good enough?
Euro4:1 pattern conforms and hugs the glass just
due to its own hanging weight... Make the top row
of a few links less, and it won't slip down.

16ga is real easy, especially if you use aluminum.
I think I showed 17ga stainless, maybe too hard?
Uncertain if stainless is a good sink.


Strange, was 18ga aluminum??? What did I do with
that spool? I can't remember...

The rings were scored and broken from the coil
using tungsten tile nippers, modified to close down
a little more than usual, but still not touching.

If this had springs pulling down the skirt like some
tube retainer clip might have, this maille would hug
the glass even more firmly. Behaves like finger cuff
Last edited:


2009-09-24 1:53 pm
Sounds like an incredible way to use the 6S19P. A home theatre system seems like an interesting way to use them. What is the max output you get from them? How do they sound?

My worry is that the tubes may not tolerate the increased dissipation even with the heat sinks. Tried that with mil-spec 6N8S and found them becoming noisy after around 100 hours even though I had a dedicated fan blowing on it with glass temp not exceeding 80 deg Celsius. Let us know the results of your experiments.
I have always been puzzled by valve heatsinks. Unless silicon grease is used, they will make only patchy contact with the glass so there will be cool spots and the rest hot - a recipe for problems? Glass is a poor conductor of heat, so even the contact points won't cool it very much. Most of the heat gained by the heatsink will be radiant heat, which was already outside the valve anyway!

The solution to valve cooling is either ceramic valves or air flow. Air does contact the whole glass fairly uniformly so it can take the heat away. Maybe Wavebourn should be looking for a very quiet fan?
I would have to agree with DF96- I cannot see the electrode structure of a valve as anything other than a blackbody radiator with very little direct conduction. Is it possible, however, in this case, that because of the very high heater current,1-1.1A for a relatively small envelope that there is significant conduction to the glass via the heater pins? I must say that I liked MJ's solution of mounting 6528 double triode on a fan guard of a low noise fan with the fan blowing from underneath.


2006-07-06 10:27 am
Maybe the glass will be cooler ( better convection ) , but the temperature of the plate will rise because some of the radiated heat will be reflected back from the heatsink , wich is very close and not a perfect black body to absorbe all radiation ...
Last edited: