Carlsbro K115 Whispering Buzz

[Ted]

I've got an 80s Carlsbro K115 here. I don't have a schematic, but it appears to be a pretty uncontroversial 2 x 2N3055 100W amp design.

It makes a sort of whispering buzz underneath the main note when you play through it. Its noticeable at low volume, but harder to make out as you get louder.

I've tried an external speaker and it's still there.
I've tried plugging straight into the FX return (i.e. straight into the power amp) and its still there. Using sig gen and scope, it looks clean enough from low volume up to max.

Any suggestions?
I really don't want to start replacing power amp components one at a time...

[Steve Conner]

Maybe it's biased too cold, crossover distortion and all that.

[Ted]

Maybe it's biased too cold, crossover distortion and all that.

I'd expect that to be visible on the scope, especially at low volumes.

[Steve Conner]

Not necessarily, when it comes to high-order distortion, your ear can be more sensitive than a scope. So check the voltages across the emitter ballast resistors and see what you get.

If it's not that, maybe it's instability?

[Alex R]

Another possibility - speaker coil rub is more noticeable at low volumes. I know you tried an extension speaker, but was the combo speaker nearby? I've noticed an unconnected faulty speaker will pick up vibration and emit coil rub noise, confusing things very effectively.

[Ted]

Not necessarily, when it comes to high-order distortion, your ear can be more sensitive than a scope. So check the voltages across the emitter ballast resistors and see what you get.

If it's not that, maybe it's instability?

Hmm. No current flowing at idle in 0.24R resistors...
But the thing has no bias adjustment. What would you expect to see?

The schematic for a Carlsbro Colt 120 appears to have an identical power amp. This could help.

[Ted]

Schematic enclosed. Only difference appears to be a 33R fixed resistor in place of the bias trim pot.

[Ted]

Maybe it's biased too cold, crossover distortion and all that.

Good call. WInding up the bias so that both 3055s are dissipating a couple of watts at idle seems to fix it.

[olivier zolive]

Hi,
Thanks a lot for shematic, and the idea.
I've just fixed a carlsbro colt 100 keybord whose output transistors (nothing written on), were dead.
Schematic is the same.
I've put two 2N3055 who seem to work perferctly. I just don't know how to adjust the bias trim pot and if they won't heat too much.

[Ted]

Like Steve says, you can calculate the current through each transistor by measuring the the voltage across the emitter resistor.

You can then calculate the transistor dissipation by measuring the power supply voltages and multiplying by the transistor currents

[Sjoerdz]

Hi Ted, I have exactly the same problem with my k115.
How did you solve the problem?

[Ted]

I adjusted the bias - measured and calculated transistor dissipation as described above.

[Sjoerdz]

I adjusted the bias - measured and calculated transistor dissipation as described above.

Ok thanks