cathode bias resistors getting hot

[phillyhudson]

just wanted to check before it all ends in smoke....I have a set of fender twin transformers running 4 6L6's with a 100r 100W aluminium sag resistor and four 470r 10W cement resistors for cathode biasing on each of the valves- I've wired in then glued the cement resistors under the power board - plate voltage 364v , cathode resistors 27v which according to my calculations makes the 6l6's VERY cold running - everything sounds ok but the cement resistors are getting so hot , its starting to melt the glue after only 10 minutes or so - have I got something very wrong ?

[xtian]

27 volts across 470R is about 1.5 watts. So that's no problem for a 10w resistor. What glue did you use? I've never seen anyone (amateur nor production amp) use glue on a power resistor...)

[phillyhudson]

dunno what type it is - from a glue gun - they're soldered to underside of board , glue is just belt and braces - so its normal for them to get that hot ?

[Colossal]

Some glues melt at fairly low temperatures, but there is also the issue of the ambient temperature rising in the chassis space because of trapped heat. There should really be no need to glue cathode resistors down if they are well soldered. You want all sides of the resistor exposed to air (maximum surface area) so that it can shed heat.

[FUCHSAUDIO]

Either get a bigger resistor (a 10-w aluminum encased resistor mounts nicely onto the chassis), or mount what you have on a terminal strip and off the PC board.
That's why the current Supro amps are failing. The cathode resistors get so hot they desolder themselves, heat the parts near them, and destroy the board....plus they are based too hot.

[roberto]

Just note one thing: glueing the resistors you are thermally insulating them, so trasforming them in 1W resistors (or so).
That makes their temperature to raise until the thermal glue liquifies.

[phillyhudson]

ok thanks - will re-do the resistor placement and probs get some aluminium resistors to bolt on the chassis - I think I can probably get away with 5W ones for the space

[Stevem]

You do have the bias voltage disconnected from the 220K resistors in the PI section right?

Off the top of my head without doing the math, I think you need much higher wattage resistors since Fender themselves on there partial cathode bias amps use 7 watt resistors!

[roberto]

No Stevem.
27 V across a 470 Ohm resistor is around 57 mA.
57 mA across a 470 Ohm resistor are around 1.55 W.