Enclosure considerations and thermal design for vaccum tube design

[etl17]

Hi there,

I am starting a new hybrid design that incorporates vacuum tubes, analog solid state and digital electronics! :-)
The ideal desired form factor is 2U or 3U 19' rack mountable unit, powered directly from mains (so, power supply should also reside within the enclosure).
Although I have worked with high-voltage tube designs in the past, all of them had been open chassis enclosure. So, I am a little nervous on how to approach the enclosure selection (or fabrication if needed) and thermal management in general.

Any recommendations/suggestions on where I may find useful info for vacuum tube enclosure / thermal management planning?

Thank you in advance!

[PwrElectronics]

I don't have any specific suggestions but some things to consider...

How many tubes and how many watts per tube?  The minimal would be the filament current x voltage from the tube data.  But, something like a power amplifier tube also has plate power to add to it.

Tubes not heat sinked so air cooling only.  Your rack mount has to account for enough ventilation somehow.  Convection maybe enough depending on my first point.  Old test gear that had a lot of tubes (think scopes) often had forced cooling.  I think of some 500 series classic Tek ones that were something like 5-600W power consumption.

The tubes themselves may be fine to a point but think about your other parts and how how will they get and how closely mounted to the tubes.

EDIT:  Found a 1961 textbook I have that has coverage of tube electronics design but there is nothing in there on thermal management.

[TimFox]

There is useful information on this in Morgan Jones Building Valve Amplifiers, 2nd Edition Elsevier/Newnes 2014, pp 13-27.

[jonpaul]

1960s RCA recieveing/transmitting/industrial tube manuals

All have mountung, and cooling reccomendations.

Some tubes are position sensitive vertical or horozontal.

We used 3 x 6L6 in 0-400V 250 mA lab B+ reg var PSU.

slow rotron fan cooling.

No issues


Jon

[etl17]

Thank you all for the helpful recommendations and tips! :-)

[TimFox]

Specifically, for horizontal mounting of tubes, some data sheets explicitly state which pins of the base should be in a vertical plane, to avoid sagging of internal structures.
Others are rated only for vertical operation.

[mawyatt]

All the old guitar amps like Fender, mounted the tubes upside down, these included the popular 6L6 and 12AX7s tubes.

Best,

[trobbins]

So, I am a little nervous on how to approach the enclosure selection (or fabrication if needed) and thermal management in general.

Any recommendations/suggestions on where I may find useful info for vacuum tube enclosure / thermal management planning?

I'd suggest you need to spend some time and effort on the basics.  As above you need to prepare a thermal budget, which may be beyond you if you don't yet have an actual design and prototype of operating circuitry?

There are also books that are worthwhile having - such as Kraus and Bar-Cohen's Thermal Analysis and Control of Electronic Equipment, and the 1962 RCA manual Electron Tube Design (this one is on-line).

PS. excerpt from Kraus/Bar-Cohen for indication: https://dalmura.com.au/static/Cooling.pdf

[jonpaul]

Power tubes like transmitting, 100..5 KW glass have

1/ Ventalated chassis/sockets
2/ Glass air chimnerys
3/ Soecified forced airflow.

Larger tubes like 5KW ..1MW (klystrons, beam power tetrodes) are ceramic and metal and heat is dissipated by conduction via   earthed   heatsink,

See Eimac and Varian.

Jon