@JoeAlders
Not a stupid question at all. This specification is for each triode individually but there is also a total dissipation limit (both triodes operating) that isn't mentioned in this particular datasheet. However, the total dissipation limit is almost always less than twice the per-triode value (3.5 watts in this case).
Here is another JJ datasheet that specifies a 5 watt total dissipation rating:
http://www.amplimos.it/images/ECC99.pdf
And here is a Reddit article that gives the same 5 watt dissipation rating:
Tube of the week: ECC99 : diytubes
So it would be wise to limit the per triode dissipation to 2.5 watts or less.
Not a stupid question at all. This specification is for each triode individually but there is also a total dissipation limit (both triodes operating) that isn't mentioned in this particular datasheet. However, the total dissipation limit is almost always less than twice the per-triode value (3.5 watts in this case).
Here is another JJ datasheet that specifies a 5 watt total dissipation rating:
http://www.amplimos.it/images/ECC99.pdf
And here is a Reddit article that gives the same 5 watt dissipation rating:
Tube of the week: ECC99 : diytubes
So it would be wise to limit the per triode dissipation to 2.5 watts or less.
5W was max plate dissipation for each triode in the old datasheet. It actually does not specify when only 1 triode is in use so it can be assumed that it's same also when both are in use. However being cautious and doing some statics over the years they have changed it. So when both in use it can be 3.5W per triode, 7W in total. This was the practice in the golden era, pretty common with most double triodes really. This might be the reason why the current datasheet says 3.5W. At 3.5W dissipation per triode I would expect a lifespan similar to a power tube.
Last edited:
The other heat factor that limits total plate dissipation . . .
ECC99:
The filament is 800mA at 6.3V = 5.04 Watts.
5W in plate 1, 5W in plate 2, and 5.04W in the filament = 15.04 Watts total heat.
Would you operate a 15W light bulb that is in such a small glass envelope?
You might, but then . . .
How much air space would you put around that 15W light bulb?
And do not forget, the ECC99 has fine wire grids that do not like to be subjected to too much heat . . .
The grid wires might become mis-shaped.
6SN7 types:
6SN7-GTA, GTB:
5 Watts per plate
7.5 Watts total plates
6.3V @ 0.6A = 3.78 Watts
7.5W + 3.78W = 11.28 Watts total heat
The 6SN7-GT model was rated at 3.5W/plate, and 5W total plates dissipation.
5W + 3.78W = 8.78 Watts total heat
Compare the two:
Consider how big the glass envelope of a 6SN7 is compared to an ECC99.
3.5W + 3.5W + 5.04W = 12.04 Watts total heat sounds better to me, versus 15.04W for the ECC99 glass envelope (and better for its grid wires).
ECC99:
The filament is 800mA at 6.3V = 5.04 Watts.
5W in plate 1, 5W in plate 2, and 5.04W in the filament = 15.04 Watts total heat.
Would you operate a 15W light bulb that is in such a small glass envelope?
You might, but then . . .
How much air space would you put around that 15W light bulb?
And do not forget, the ECC99 has fine wire grids that do not like to be subjected to too much heat . . .
The grid wires might become mis-shaped.
6SN7 types:
6SN7-GTA, GTB:
5 Watts per plate
7.5 Watts total plates
6.3V @ 0.6A = 3.78 Watts
7.5W + 3.78W = 11.28 Watts total heat
The 6SN7-GT model was rated at 3.5W/plate, and 5W total plates dissipation.
5W + 3.78W = 8.78 Watts total heat
Compare the two:
Consider how big the glass envelope of a 6SN7 is compared to an ECC99.
3.5W + 3.5W + 5.04W = 12.04 Watts total heat sounds better to me, versus 15.04W for the ECC99 glass envelope (and better for its grid wires).
Last edited:
Good points, I was going to put mine about 2 tube widths apart - 12BH7A at the front and the two ecc99s behind as a triangle. I may need to make that larger due to radiated heat rather than simply the convection cooling. I may make that closer to 4 widths instead.
Automotive lamps are much smaller and run much more power, in an enclosure next to a hot engine on a warm summer night.
They don't have grid wires though. 😀
And the smaller 5687 has 4.8W per triode / 7.5W both and both of those tubes have smaller plates. If the tube can't take the power, I suspect they used "normal" glass instead of quartz like in a halogen lamp.
There's a reason a bulb the size of an IN-9 indicator tube can be 500W without melting... The first post of the repair threat shows a 6P43P left to orange-plate for so many minutes... The glass started to implode.
The much thicker glass of the 6P45S will crack instead of melting if you orange-plate that - take a 6P45S, tie screen to plate. Ground cathode and grid. Apply 320V. Watch it attempt to dissipate over 300W until the envelope fails! (Actually I didn't see it fail, I heard it as a reduction of volume on the left channel. The amp had happily played music while it was killing it's tube).
George isn't the only one who's killed a share of tubes, right? LOL
Last edited:
