Biasing a Sunn Coliseum Bass amp

[spy]

I have a Sunn Coliseum Bass on my workbench and did a full recapping job, also I replaced all the power resistors plus the bridge. I'd like to check the bias and bias the amp again but I don't have any info on it.

The schematic is at the beginning of this page,
The SUNN Shack
and here is the output stage,
http://www.richbriere.com/images/The_Su3.gif

Can anyone help to bias it properly?

[Jazz P Bass]

Measure across wire #7 & #5.
I would go for something realistic, like 10 mv's.
Without factory documentation it is impossible to guess the setpoint.
Still, I would go with a low current draw.
Or you could set it to produce crossover on the loaded output & tweak it until the CO just disappears.
Or you could do the "Enzo Thing".
Set the bias cold as a stone.
While monitoring the mains current, adjust the bias pot until the mains current starts to climb.
At this point back it off a hair.

[spy]

That's across R1 and R2, right? Do you mean #7 and #6, across R1?

What I don't like is that the amp has mixed brands of output transistors, one Motorola, two old RS, three newer RS, is this a problem? Would it be better to replace them with new ones, same brand so they would be matched? I took some measurements across each .33R emitter resistor and on five of them the bias is around 2-3mV but on the sixth is around 8mV.

[Jazz P Bass]

Across R! & R2 will give you the total draw.
You knew that.
I know it has been said here before 'do not mix & match'.
Personally, if the transistors work, I would just test it.
If it works, it works.
I would not worry about the one that wants 8 mv,s.
What I look for at test is whether or not the bias settles down at 'no signal' after it has been driven hard.
It's your call.
There is no guarantee that new ones will be 'matched'.

[Enzo]

Or you can set it cold and slowly turn it up while running a sine wave through the amp and scoping the output. Observe the crossover notch on the waveform. Turn the bias up to the point the notch JUST disappears. The I always back off from that just a hair. But my "thing" always works for me.


Whenever a manufacturer says something like 10mv across some resistor, it is a sop to techs. It means they adjusted it right then measured the resulting current. They then tell you that current and expect that if your amp is also set to the current they found, then your amp will be more or less properly biased too. I find it interesting that people really want some current figure for this adjustment, but when faced with the Peavey tube amp system of seting it by voltage, they balk. PV knows that at those voltages, tubes will be biased as they want them, in much the same way.

The way to do it is with a scope and watching the crossover distortion. My field tech approach works because I understand how the amp works and why the notch forms at all. SO my way is also a sop for techs, but from a different angle.


Looks like all your outputs are the same types. The amp is quasi-complementary. Three on each polarity, total six. The reason you want them of the same type is that the three work in parallel. And if they are different, they can either hog current or not carry any. A transistor turns on at a certain threshold. If one in the row turns on at 100mv lower voltage, then when signal approaches, it turns on first, and the voltage clamps off at that point and none of the others get turned on, it is then said to "hog" the current. And so all the output power comes through just the one part. Likewise another never gets turned on, so it is as if it were not there at all. Part of the reason those ballast resistors aree there is to promote current sharing.

Also, temperature affects current flow in transistors, so as they warm up they change. It is important that transistors thermally track together. And equivalent RCA and Motorola transistors way work fine individually, but noit work together at all. I am not concerned over gain, I worry more about thermal issuues and junction voltages.

NO they don;t have to be matched, but transistors are made by different processes, and some may work better with others than some alternates.

I don't care if the three on positive are different from the three on negative. But on eaither side, I'd like all three to be the same type, whatever type you chose.

[spy]

Ok, I did all this work with the recapping, etc, I think it worth it to replace the power transistors, too. Hope, I won't find them a fortune locally.

The "Enzo Thing" sounds as the best way on biasing a solid state amp, I'll give it a try and I'll let you know! Thanks!

[spy]

I replaced the transistors with 6x TSL 2N3773 and everything went well. Any idea about the TSL on them? What brand is this? Tesla?

About the bias, I tried the Enzo Thing but I couldn't see a crossover distortion, only a little when pushed really hard. I left the bias around 10mV across points 6-7 and no problems with full output, the amp sounds really great!

[gbono]

On semi, STMicro, Central Semi, Microsemi, Boca and Micro Commercial all produce some :version" of the 2N3773 - TSL? What are the other markings on the part?

[spy]

Nothing else on them! Just TSL 2N3773.

[gbono]

Nothing else on them! Just TSL 2N3773.

where did you purchase these transistors?

[spy]

I purchase them at a local electronic shop, why? Are they the worst ever made?

[spy]

All this year the amp was active and kicking but suddenly a few days ago the amp died. No signal with a load and full signal without! Any thoughts?

[J M Fahey]

You mean you do have voltage output (some 30V RMS) at the speaker terminals unloaded but they disappear (or go down, how down?) when you plug a speaker there?

Looks like an "open" somewhere.

1) without load , measure voltage from each speaker terminal to ground (yes, you might have an open ground).

2) if ground fine, go backwards towards the Power Amp PCB output rail, power transistors legs, same with drivers, try to find signal again.

[Jazz P Bass]

Bad speaker?

[spy]

I'm connecting a dummy load and tracing the output with an oscilloscope parallel to it.

Yes, without the dummy load (8 ohms) I have full power and with the load no power, just a line with small round edges at the frequency I have set from the generator.