1

I made a stereo amplifier on a PCB which is two of the following circuit.

One channel amplifies the audio decently, but the other channel has NO sound regardless of how loud I make the input.

My sound source for testing is the earphone jack of my computer. I have tested each channel independently and have confirmed that one channel does not work.

I have set my multimeter to diode mode to check for obvious short circuits and could not find any. I also could not find any disconnections.

One source online stated to look for an open PN junction on transistors but when I measured some transistor connection points, I get numbers that constantly change (probably due to the bootstrap capacitors).

Is there a way I can figure out what part to replace using my multimeter before I start taking the whole board apart and replacing everything (possibly including parts I didn't need to replace in the first place)?

I attached the circuit for reference and the sim confirms that the circuit can amplify sounds. Yes 6V is low-ish but I'm testing with batteries so I don't explode parts.

circuit

JYelton
  • 32,302
  • 33
  • 134
  • 249
mike_s
  • 11
  • 1
  • 2
    Since you have one channel that works, compare voltages (AC and DC) between the two channels at several points. If you find a point where the voltages differ significantly, that should lead you to the fault. – Peter Bennett Jul 25 '23 at 01:02
  • 1
    Mike_s, that is a kind of funky "bootstrap" technique (\$C_2\$) there to create a semi-constant current source via \$R_1\$. I'd have done it differently, with one less capacitor to achieve it. But no taking away from what I see there, either. I'd love to see your analysis of the NFB via \$R_9\$ and the resulting expected voltage gain. Did you do any such? Or is this a copy from somewhere else? – periblepsis Jul 25 '23 at 05:51
  • I imported it from some author a long time ago. I think his amateur radio call sign was va3sp0 or ve3sp0. He had a large tutorial on transistor amplifiers. I cannot find his website lately. – mike_s Jul 25 '23 at 19:27

2 Answers2

1

Set the signal voltage to zero. (In the simulation as well).

The first is a dc voltage test of all the PN junctions and node voltages. You can use the simulation to get a reference.

The voltages across the PN junctions should be between 0.62 and 0.68. These junctions usually fail open so it goes really high. Replace the semconductor. They can fail short so a zero volt drop is a warning sign. Check the voltage across nearby series resistors to see if there is current through the PN junction. A zero voltage means the junction is defective.

If it is low or just a little high then the problem may be somewhere else (or maybe not).

Test the node voltages to ground using the simulation as reference. Any discrepancy can lead you to suspect a nearby part.

Test the voltages across the resistors. If it appears out of range, it or a part nearby is defective. If there is a voltage across a resistor than there is a current through it. So you can map the currents as well.

This will give you a map of voltages and currents. Use your electronic knowledge (KVL, KCL, Ohm's law) to "deduce" where the problem is.

Set the signal generator to a small amplitude that the ac voltmeter will measure (in the simulator as well. Repeat all the measurements as before.

As mentioned by Peter Bennet, you can also use your working channel as a reference.

Don't blindly take all the measurements then try to sort them out. Work through the circuit in a methodical fashion trying to deduce the failure at each measurement. It may not be a part, but a wiring or soldering issue. Don't use diode mode when the circuit is under power. Use the dc voltmeter and the ac voltmeter. Often the failure will reveal itself in a few measurements.

Trust KVL, KCL, Ohm's law and your meter. Well... check your meter with a known voltage.

RussellH
  • 12,496
  • 2
  • 9
  • 34
  • After measurements with my voltmeter I noticed that the working side had showed an almost 5V difference when measuring the 2N3904 collector pin in reference to ground on the working side. on the faulty side, that 2N3904 collector pin measured around 0.5V. I'm suspecting I melted the inside of a transistor – mike_s Jul 25 '23 at 19:28
  • Not necessarily. Measure the base-emitter voltage drop. If 0.7 or higher, then the 3904 is ok. Perhaps R1 or R2 is open, mis-wired, or one of the diodes is in backwards or failed open. One measurement is usually insufficient to determine the fault. Check your theory by measuring other components or nodes – RussellH Jul 25 '23 at 20:53
0

ok first I thought it was the 2N3904 that need replacing but after closer visual inspection, I installed the transistors in the wrong places. The orientation was correct but the part numbers were off (I put in a TIP42 where a TIP41 is supposed to go and vice-versa).

After making those changes everything started working.

mike_s
  • 11
  • 1