It is a tricky subject.
The heat transferred through the glass is not just a factor of glass outside surface temperature, as heat can transfer via transmitted radiation which has no relationship with external surface temp.
Similarly, the temperature of the anode is not just related to the glass outside surface temperature. The mix of influences is very difficult to rationalise, as the reflectance and transmission of an interface are not zero, and vary with IR wavelength, and IR spectrum changes with body temperature, and bodies with thickness that allow transmission make it even trickier.
Imho, although the content of unrefereed papers/articles is fair and reasonable, it is part snake oil to try and push a psuedo scientific experiment and discussion as justification for a commercial product without making the experiment and results a proper scientific paper, and attempting to get it done by or peer reviewed by appropriately experienced professionals, especially where the measurement is not done with a mature commercial instrument and a widely accepted measurement technique.
The heat transferred through the glass is not just a factor of glass outside surface temperature, as heat can transfer via transmitted radiation which has no relationship with external surface temp.
Similarly, the temperature of the anode is not just related to the glass outside surface temperature. The mix of influences is very difficult to rationalise, as the reflectance and transmission of an interface are not zero, and vary with IR wavelength, and IR spectrum changes with body temperature, and bodies with thickness that allow transmission make it even trickier.
Imho, although the content of unrefereed papers/articles is fair and reasonable, it is part snake oil to try and push a psuedo scientific experiment and discussion as justification for a commercial product without making the experiment and results a proper scientific paper, and attempting to get it done by or peer reviewed by appropriately experienced professionals, especially where the measurement is not done with a mature commercial instrument and a widely accepted measurement technique.
Because you should consider both effects working against each other: cooling of the glass by a heatsink in tight mechanical contact with it, and heating it up by hotter due to reflection anode.
A possible 'thought experiment' for increased anode temp, and reduced glass temp, could be to assume the glass has relatively high transmission of IR radiation and relatively low reflectance, and to have the valve in a large box with a relatively high temperature blackbody surface for the anode's IR radiation to 'see'. Assuming that heat is only transferred from the anode by radiation, due to the internal valve vacuum, then the anode temp would mainly be dictated by the external large box surface temp. The glass temp could be low or high and would almost entirely depend on the local air temp.
If people did as you say then half the audio industry would vanish. Silver plated speaker wire and hand rolled capacitors would not be sold.
After all, this is entertainment. If snake oil entertains people, I guess that's OK.
Exactly. Like I said, put a thermocouple on the anode, take some data and then we'll talk.
Talk about what, if I may ask ? ?
And while I'm asking about things: what causes the anode to get hot and why is cooling it necessarily a good thing, given of course it's not going to melt or outgas with heat ? ?
And while I'm asking about things: what causes the anode to get hot and why is cooling it necessarily a good thing, given of course it's not going to melt or outgas with heat ? ?
I was searching for heatsinks in order to pack tightly many tubes in a limited volume. If it is not a case, heatsinks are useless.
Talk about the merits of any specific "coolers".
Cooling is not necessary most of the time, in fact I've never had to use it. But I readily concede that there may be cases where cooling may be desirable to keep a tube from overheating, which may destroy it or reduce its life. As in Wavebourn's case above.
You'll need to be able to move some air over the coolers and thru the enclosure, obviously. That done there are no issues, the coolers cool tubes . . .
You're sure you've, " . . . never had to use it" ? ? You must be running nothing but 12AX7s and/or variants . . . 🙂
Last edited:
I'm sure. I'm running a set of old Sylvania 6L6GCs right at the datasheet dissipation limit (30W plate) right now. They get hot. They have for many years. No issue.
In the end it is just kinetic energy. Fast moving electrons smash into the plate and are stopped by it. The energy of the movement gets turned into heat. (law of conservation of energy)
The same thing happen when water falls down a waterfall. It is accelerated by the gravity field and then all of that gained energy is removed when the water makes that 90 degree turn at the bottom. The electrons do pretty much the same thing as they "fall" through the electric field.
The thing is you can't cool the plate with a heat sink because it is in vacuum. It can only cool by radiation. That is one reason why the tube is made of glass, so the radiation can pass through.
The reason you would like it to be cooler is so that you can crash even more electrons into it and get more power. The BEST way to do that is simply to make the plate physically larger. Or use more plates in the amp.
Last edited:
I think you would do better by simply using forced air. Heat sinks will just take up space. Unless you plan on liquid cooled heat exchangers, those can be very compact.
Assuming air cooling, I think bare tubes with a blower fan would give you the minimum number of cubic inches. reason is that heat exchange is most efficient when the difference in temperature is greatest.
When 14 of hot power tubes stand close to each other on a single 17" wide chassis?
May be if to put them on the edge of a chassis, and cover from the rest of the amp by a reflecting zigzag... At least they will look interesting... 😉
Ya I was picturing a thin rad of sorts separating the tubes, absorbing radiant heat as well as isolating the tubes from each other.
If it's already built>not practical.
Something to consider on a future build.
If it's already built>not practical.
Something to consider on a future build.
whats the problem . put a PC fan in there and just buy enough tubes to retube the amp five times . they can handle overload course of the anode coating
As it was mentioned here several times, PC fans would be fun, but unfortunately vacuum is needed...