Quote from: saturated on October 01, 2026, 02:58:01 PMI did have a who's on first moment though
I had to ask myself if increasing the resistance of the variable resistor eliminated the crossover notch or if decreasing the resistance eliminated it.

Your voltage measurements between the bases (post #4) showed that 1V was not sufficient to remove the notch, but 1.2V was sufficient.
Increasing the value of the trim pot increases the voltage drop across it.  So it was increasing the value of the trimmer that eliminated the notch.

Ok, well points back to the speaker relay and you mentioned it clicks just before you get sound.
This is highly likely either the Relay itself or the circuitry that runs it.

Others here will know more but often you can just bridge the relay terminals and bypass the relay. Outside of rare situations the only real reason for the speaker delay circuitry is to mute the turn on thump through the speaker.

Sadly it looks like you would have to remove the whole PCB to access the terminals
so don't try anything until others here with better experience can verify what I'm saying.
Phil.

1. Yes, you have 8V going into a 10:1 voltage division, split between the transistors. So each one gets 8/11 ~ 0.36V.  There is a dead band.

2. Removing the "bridge" resistors means that you have something very similar to "base bias" (except that there is no collector resistor) rather than voltage divider bias. You've eliminated the certainty of the voltage drop created by the middle resistors. The 10K resistors are now base resistors, limiting how much current flows to/from the bases, and that current, together with the temperature and BJT parameters, determines VBE.  Current does flow, and so the transistors are turned on, and there is no visible crossover distortion. With no collector or emitter resistors, you risk instability: thermal runaway. The only thing limiting collector current is VBE in response to the base current, which is temperature and parameter dependent.

Btw one particularly annoying discovery

No crossover distortion when the bridge resistors are removed entirely

However it now looks like the poor little transistors are over biased and crying out for me to turn down the DC voltage

You cannot view this attachment.

Wow I was able to turn the DC down from 8V to 2V without a change in output

And get the Vbe down to about 0.65 on each

I still question why there is no voltage gain but this thread has already gotten out of hand divergence 

Anyhow I replaced one of the 1k resistors with the 10k variable and was able to adjust it and find the sweet spot

You cannot view this attachment.

I did have a who's on first moment though
I had to ask myself if increasing the resistance of the variable resistor eliminated the crossover notch or if decreasing the resistance eliminated it.

Then I had to try and reason how the DC voltages at A and B would respond 

Life is full of tough questions

Time for another side project 

I went to redo this circuit with a variable resistor and forgot which of three terminals was variable and which was fixed 10k

So I stick it in and attached two leads and couldn't get any resistance

You cannot view this attachment.

So I pulled the trim pot out and saw sure enough 10k on outside pins 

Rechecked and messed around finally thought hmm what if it's one of the wires 

Sure enough

You cannot view this attachment.

 

i tried that but it didnt work. however the other day i played my amp for about an hour and no issue. havent played it in a few days but.. weird problem

Hmm ok yeah I went to tear it down and disconnect and unplug everything I had an idea why not turn up the DC supply voltage so sure enough when i hit about 12.7 V

You cannot view this attachment.

And now remeasuring Vbe

You cannot view this attachment.

So seems kinda simple oh well 

But I'll go ahead and proceed onward I anticipate they use the variable resistor to obtain needed voltages with the original 8 V DC but the part about the diodes has caught my interest    should be cool

Hmm this might be over pretty quick

Maybe they are just trying to demonstrate that if you don't have enough forward voltage then stuff isn't gonna work right

You cannot view this attachment.

Seems like I remember threads where you were telling the noobies how to adjust the resistors until the notch disappeared

But seems like an elaborate circuit to demonstrate something that simple using two transistors and all.

I guess we'll see what's in store 

The other thing that is messed up is there's no amplification/voltage gain.  I went back and made the volts per division equal for both channels and the output is actually SMALLER 

You cannot view this attachment.

I know I really need an amp in need of repair.  Trust me ive been looking.

Anyhow I was flipping through a book looking for something cool and there is an experiment for creating and then abating crossover distortion.

Here is the plagiarised schematic

You cannot view this attachment.

Ensuing rats 🐀🐁 nest

You cannot view this attachment.

I managed to throw this together pretty quickly and to my astonishment it worked.  They called for 2.2 uf electrolytics the closest I had was 3.3 uf

I didn't even measure the input I just increased it from zero and BAM there it was on the output

You cannot view this attachment.

My instructions are to learn about the existing phenomena then take steps to cleanup the output by
-replacing one of the 1k resistors with a variable resistor
-replace both 1k resistors with diodes

But before I do that I need to try and grasp what's going on here   

I cheated and read a little seems like biasing is the main deal here I'm gonna measure the Vbe on each transistor at present and see what's up