How to follow a signal through amplifier schematic

[WaldPinkler]

My car amplifier has one distorted channel. My purpose isn't really to fix it but more to learn what is happening and where pinpoint the fault is through understanding rather than just replacing parts.

The nice part is that there are 2 channels and I can follow the good signal and compare with the bad signal.

Some of the music can be heard through the bad channel but it's rough. A pure sine wave gets skewed and weakened at that speaker output.

The schematic is very readable and it's clear how the stages are divided. The part I'm struggling with is where should I be testing?

I thought I could scope the outputs of each op amp stage but I'm struggling with where I should put the negative lead? I can find the sine wave randomly by probing over smd caps but again there's no understanding of what I'm doing.

The physical board has some burn damage on the differential absorption stage of the bad channel. I replaced the smd caps there already. If that's an issue, I should be able to find the clean signal before then right?

I assume that the power rail side is also good because of one working channel.

The schematic has nice divisions below from the input, gain, lpf, etc.

[andy3055]

Signal is indicated with the red arrows. Negative or ground is the heavy black ground. So, your scope probe ground goes to that.

[xavier60]

My car amplifier has one distorted channel. My purpose isn't really to fix it but more to learn what is happening and where pinpoint the fault is through understanding rather than just replacing parts.

The nice part is that there are 2 channels and I can follow the good signal and compare with the bad signal.

Some of the music can be heard through the bad channel but it's rough. A pure sine wave gets skewed and weakened at that speaker output.

The schematic is very readable and it's clear how the stages are divided. The part I'm struggling with is where should I be testing?

I thought I could scope the outputs of each op amp stage but I'm struggling with where I should put the negative lead? I can find the sine wave randomly by probing over smd caps but again there's no understanding of what I'm doing.

The physical board has some burn damage on the differential absorption stage of the bad channel. I replaced the smd caps there already. If that's an issue, I should be able to find the clean signal before then right?

I assume that the power rail side is also good because of one working channel.

The schematic has nice divisions below from the input, gain, lpf, etc.

There is no one signal path. Most of the signal flow is in the form of balanced currents. You are likely to see larger signal voltages in the faulty channel caused by the negative feedback trying to correct the output error.
 A good place to start is checking the current in the Voltage Amplifier Stages by measuring the voltage drops across the 100Ω Emitter transistors of the VAS transistors, eg, Q136 and Q137 while no signal is being applied.

[Doctorandus_P]

(Analog) Audio amplifiers are mostly build upon the same same principles.

At the input there are usually two transistors "opposite to each other" (I think this is called a long tailed pair). and these two transistors compare the input signal with a feedback signal (from somewhere near the output).

After those two transistors the input signal get split into something that drives the top power transistors, and something that drive the bottom power transistors. The power transistors at the end usually only deliver current gain.

In between the top and bottom power transistors there is a sort of "power zener" built from a transistor and a voltage divider (or similar).

Often there are a few constant current sources, but in the simpler amplifiers these sometimes get replaced with simple resistors.

I suggest you start by reading some literature about audio amplifiers, and combine that with studying some schematics of simpler amplifiers. Once you understand how those work, you will start to recognize the blocks in this amplifier.

A simple thing you can do is to measure both diodes in each transistor with a simple diode tester. If there are big deviations form 600mV, then it's very suspicious, ans as an extra check you can verify it whit the working channel.

[Doctorandus_P]

I looked a bit around for a simple schematic, and I found this one:


TR1 and TR2 compare the input signal with the feedback signal.
TR3 and TR6 are both current sinks (with 2x 1N4148 as voltage reference)
The signal goes trhough TR4.
TR5 is the "Power Zener" (Big capacitor of 68uF) (a.k.a Vbe multiplier).
TR5 keeps the voltage between the base of TR7 and the base of TR8 sort of constant.

The current sink TR6 tries to pull the "power zener" downward, while the signal transistor TR4 modulates an "upward" current. Result is that TR5, and with it the power transistors at the output go up and down with the current though TR4.

The current through TR6 is (nearly) stable, which means that small current changes through TR4 can lead to big voltage swings in the end stage.

A 27k resistor between the output and the TR2 provides the feedback to limit the amplification and keep it all stable. The 47pf and 2k2 resistor marked as "no longer needed" reduce the amplification at higher frequencies (above audio) and are just for stability purposes.




 

[WaldPinkler]

Thanks for the advice. I've played around with it and done plenty of reading. I'll study the materials given.

There was a short from the -28V power rail to the -15V rail and ran through the right channel. The side of the tracks are deformed so it's easy to follow.

I checked diodes, resistor values and for shorted caps. Looks good.

Q237 gets 1.2V with diode check from base to emitter. Good channel has 0.7V. I guess that one is dead but it's also affecting the top signal too.

100R (R241, R245) have no voltage drop. Good side 435mV.

I attached some scope readings. It's been interesting playing with this. A great way to learn.

[WaldPinkler]

Broke open an old computer power supply and found an npn transistor. Changed Q237. It changed the signal but still looked a weird shape.

I changed C233 47pF purely because I had the hot air gun out and the colour looked a bit off.

The signal is perfect!

This is my first repair success. It was for learning but there's no need to put the amp in landfill. I'm pretty happy.