Kustom 150 run hot?

[drewl]

Got a Kustom 150 that had Q5 blown and it's emitter resistor R22 burnt open.
Looks like the output 3055's Q18 & Q19 were replaced sometime in it's lifetime, and they tested good. Q5 and Q6 were also replaced with T0220 TIP31 and TIP40-something.

So I replaced Q5 & 6 with proper rated T039's and swapped out Q18 & 19 with MJ15022's.

Amp works fine, but all of the above transistors are a bit warm, voltages and current through the output devices all look good, power supplies clean.

Just curious if this is normal for the output drivers to get a bit warm.

I was thinking of replacing Q12 & 16 (2n3055's) also with 15022's.

[g1]

They are showing .1V around the outputs on the schematic. Looks like that would be .1V across each of the emitter resistors R26 & R28 or 200mA current. That seems like fairly warm bias to me.

[Enzo]

You sure that 0.1v is not for the base of the transistor?

What is a bit warm? Are they WARM, or are they just more than room temperature?

See the three bias diodes in series, CR1 CR2 CR3? SHort across CR2 or CR1. Does that cool it down? Does it make objectionable crossover distortion. CR3 is a different type, I have not seen one of these in a while, is that diode mounted on the heat sink or against the body of a transistor?

Q5 Q6 are drivers, and those Kustom part numbers translate to 40409 and 40410. They were indeed TO39, but they also had the integral heat sink. If you crosss them to something like NTE, you get the voltage and current rating, but not the heat dissipation capability. Did you use 40409 or did your TO39 sub have heat sinks mounted? They will get too hot if you don't

[drewl]

CR3 is mounted on the output transistor heatsink, and I installed a couple of those cool looking old finned heatsinks on the T039's.
The amp has the correct + & - .1v at those .5ohm resistors. Like I said, all the voltages are as they should be, everything tracks good, no output or crossover distortion.

How hot is hot.....hmmm, I have a temp meter at work, I'll see. None of the transistors in question are too hot to touch, they don't burn you but with the large output transistor heatsink block I thought it would dissipate more heat, that whole section of the chassis gets pretty warm.

I'll check the bias diodes with the shorting trick, see if there's any change.

Thanks.

[drewl]

You know something about those diodes, don't you.....

[g1]

You sure that 0.1v is not for the base of the transistor?

No, I was a little iffy on that, which is why I said the voltage was shown "around" the outputs.
However, his measurements show the .1V across the ballast resistors, and wouldn't voltage between the bases normally be in the .5V vicinity?
(still should be a good idea to cool the bias anyway, barring objectionable x-over distortion).

[The Dude]

Just curious.....When you say it runs hot, are you talking about at idle with no signal, driving a load at max power, or something in between?

[52 Bill]

If it runs warm at idle try changing the two diodes on the board, I think I remember a service note about this. Be careful when moving the heatsink mounted diode as the leads will break off very easily.

[Enzo]

That is why I pointed that diode out, stick with the other two if you mess with it.

Drewl, the only thing I know about those diodes is what they do in the circuit. You will see a similar bias string in many amps. The whole poimt of bias is to maintain a certain voltage difference between the ends. SO shorting one of the diodes will reduce that space, and thus also reduce the idle current of the outputs it biases. If it runs cooler, great, then liten. It will increase crossover distortion, at least potentially. If it is idling at over current, then a reduction my not get it into crossover land. But you also may not hear a difference, and if you dont then leave it running cooler.

[drewl]

The reason I mentioned the diodes is when I swapped out the two main T03 drivers, I didn't test it until I got home from work and then there is some strange 5MHz oscillation on the power rails, those transistors and even the chassis!
It's only about .5v peak to peak, or ~176mv, and just clipping a meter to those diodes makes the oscillation go away.
I wasn't sure if it was something in my house causing it, or if you knew something about those diodes going bad/leaking in some odd way.

I'll try it here at work again and see if it's still there.

[Enzo]

Oscillation on the chassis? Your scope is grounded to the chasssis, isn't it?

[drewl]

Yes, of course it's grounded.
I even clipped another ground lead to see if something lost it's ground.

[Enzo]

I wasn't trying to be flippant. I was trying to figure out how we could ground a scope probe clip to the chassis and still see a signal on the chassis on the scope.

[52 Bill]

You may need to add a small value cap to the driver transistors to stop the oscillation. The modern transistors have better high frequency specs than the ancient originals.

[Enzo]

Good point, I don;t see a zobel network on the schematic. One could be easily added.