SWR California Blonde troubleshooting tweeter/power amp repair

[Enzo]

I don't think in terms of more negative or positive, though that would be the case. I think of the bias as maintaining a voltage space between those opposing driver bases. That is what prevents crossover. Instead of the bases sitting at zero volts, each sits a bit above and below zero. We want the DC levels to be just where the transistors start to turn on. If a transistor base starts at zero, then the signal has to rise up to about a half a volt befiore that transistor turns on. So a little half volt bite comes out of the signal. But bias the base up to half a volt, and now the signal can start amplifying at the first instant, no little crossover bite. That covers the driver, but add in the output transistor, that adds another half volt. So two half volts up and two half volts down, that means Q4 would needd set a 2 volt difference across itself. (I am not looking at the schematic, so if there are additional transistors, we adjust for them.)

Left on its own, Q4 would center around zero. Whatever voltage we put on it, the output follows. it just maintains that 2 volt space wherever we push it. That is how we send signal through it. Q3 tugs it up and down, through your R11. Signal moves Q4 up and down, and the outputs follow.

To do that we need current flowing through Q3, so there is a circuit strip from V+ down through Q10, R5, R9, Q4, R11, Q3, R32, to V-. Q10 is more or less a switch, the mute circuit turns it on to energize the front half of the power amp. SO normally Q3 turns on enough to pull Q4 to center. Now signal from Q1 will turn Q3 on more and less to increase and decrease current through Q3, which has the effect of moving Q4 up and down in voltage.

So shorting Q4 just means all that happens without leaving the 2 volt space. And that means as your signal crosses zero it has go go that extra volt each way before the outputs conduct current to the speaker.

[DrGonz78]

Enzo thanks so much for explaining all these questions and it sure is helping me understand the bias circuit so much.

Well I shorted the Q4 E to C and on the scope there is still no cross over notch present. I twiddled so many different ways of looking at the scope and the amplifier in conjunction so I am very confident that there just is no cross over notch to view. Just seems very strange indeed that shorting the driver bases did not reveal cross over distortion. Perhaps it needs a load to set the bias in this manner? I still need to turn the bias pot a tad to see if that changes voltage drop across emitter resistors R19 & R20.

[g1]

Try it with a load. That is what they specify in that bias info. (load and low level signal)

[DrGonz78]

Yup ran it through an 8 ohm load and viola! I measured about 1.8mv across R20 and new right away that it was probably going to do something different. Ran it through the scope and presto notch appears, finally! So question is now which load should I be using to set the bias? I have 8 and 4 ohms large resistors for loads. I could parallel those together to make a 2.6 ohm if that might be necessary in this case. Here is another example of workingmans amp that also says 2 ohm load. At least this is starting make sense a lot more. Thanks so much for the help up to this point.

[Enzo]

Do the work yourself, run up a medum signal on 8 ohms and then 4 ohms, and see how much difference the load makes to crossover notch.

[Jazz P Bass]

What is the purpose of 'bias setting' ?

It's purpose is to take a direct coupled amp (in this case, an amp that is designed to be in Class B output mode (Class B = crossover distortion) & and turn on the output's just enough to take the output into Class AB.

If Fender wants a load attached so be it.
You can use 5 watt resistors for the test.
They specify a 2Vac RMS signal at the output.
Even at 2 ohms, that's only 2 watts.

Do you have a Kill-A-Watt meter on your test bench?
You should. (or some way off monitoring line current draw)
Overall amplifier current draw is a nice thing to know.

Do this. Try the "Enzo Bias Method".
Set up the DUT and monitor the idle current draw with the bias control set at minimum current draw (Class B), (with or without a load).
No input signal (Note: Some techs prefer to short the input jack.).
Turn the bias control until you see the idle current start to increase (Class AB)
Back it off a tad.
Done.

[DrGonz78]

So I ran it on 8 ohms and then 4 to see the difference. The difference seemed to be the limiter would get triggered badly under 4 ohm load and made problems for setting the bias that way. That may be why they send test signal through effects return for setting the bias like in the Workingman amps. Either way I was able to combine the Enzo Method while using a scope to set the bias at the same time running a test signal. I monitored the line current through my ammeter on my variac while setting the bias and looking at the scope while the notch slowly went away. At the end I set it cold again and just went with the Enzo Method while the amp sat at idle with no signal injected. Got the amp all back together and it sounds great! Thanks to everyone here for all the help. I no doubt will return to this thread to re-read it many times and really let it all sink in.

[Enzo]

Yes, usually bypass the preamp if possible. And there is no reason to max the amp, if you trigger a limiter, then back off your levels.


I think in most cases you will find the bias setting where Xover just disappears and the point my cheating method comes to are pretty close to one another. If nothing else, my way gets you pretty close to your goal, and you can get fancier for finer tuning.