Suddenly an amplfier shows terrible noise on the digital input. The internal pre amplifier is supplied by the source below, which has about 1V ripple on its +/- 21 volts output. Could that be the cause? No idea what the output looked before the noise occured. But I assume supply voltage to a DAC section of an amplifier should look more flat? What part in the cuircuit could be the cause? The rectifier? Explanations on the basic principle of the cuircuit would also be helpful.

The measurements are at "-21V" (AC, 10x probe) and at the collector of Q910 (DC, 10x probe). I only measured the negative voltage output, the ripple on the positive one looks the same.

Thanks!

 

1V ripple on its +/- 21 volts output. Could that be the cause?

No, 1V is not the problem. Because the circuit you have shown is a pre-regulator I guess. There must be a voltage regulator after this circuit.
So check the voltage regulator in the preamp section and voltage reg for a digital section as well.

 

Thanks. Indeed there are two voltage regulators for the preamp at + and - 18V, the 18V then also feeds a 5V digital voltage regulator. There is still ripple after the 18V, but much less. If I short the output from the DAC to the pre-amp, the noise is still there, so it's not a digital problem. (shorting the output of the pre-amp removes the noise) As there are two symmetric circuits left / right in the pre-amp - which are not connected except for the supply voltage - and they both show the same issue, I thought it can't be a problem within the pre-amp?

 

My bet would be the large electrolytic capacitors (2200/35), you could replace all the electrolytics – they are notorious for failing. People sell kits consisting of all electrolytics for an amp.

Alternatively you could probe the voltages at each electrolytic to pinpoint the faulty component(s).

 

Thanks Hymie.

One addition: The noise starts very strong when I power on the amp. After a while (2..3 min) the more or less "white noise" disappears, first on the right channel, then on the left. The humming (100 Hz) remains. Funnily in between all on a sudden the fault was completely gone for a while (about 4 weeks), only to reappear again now...

What I don't understand is the behavior that occurs on both channels, which are obviously independent, except for supply voltage.

 

To get a 2V ripple voltage on a 2200µF capacitor would be draining at around 0.4A – that gives a total power draw of 17W for the preamp supply, which would appear quite high, implicating the capacitor(s) as being faulty.

Often electrolytics exhibit a bulging of the casing when they fail – you may be able to observe this with careful inspection. If you have access to a capacitance meter, often an in-circuit measurement will give an indication of a much reduced capacitance value, indicating a failing capacitor.

Within the circuit, Zener diode D907/capacitor C943 combined with D908/C944 set the +21Vdc output – if there is significant ripple on either capacitors C943/C944 this will be transferred to the output.

 

Its the caps

 

You mention intermittent problem and this leads me to think about cold solder junction. Check/resolder componets in the power supply that may have lost good contact. Especially bigger components like caps etc. Even if you do not see the problem, just solder them again. Will not hurt.

 

You mention intermittent problem and this leads me to think about cold solder junction. Check/resolder components in the power supply that may have lost good contact. Especially bigger components like caps etc. Even if you do not see the problem, just solder them again. Will not hurt.

^+1
I have a (now ~22 yrs. old, was abt 10 yrs. old at the time of the problem) Klipsch subwoofer (10", 150W) that developed a hum. I initially thought caps. Then I did a close inspection and noticed a couple of small cracks (figured cold solder joint) in the leads of the caps. Soldered those and problem went away.

A few years after that, the subwoofer developed a higher pitched noise (almost like a "scritching" sound) that occurred only when audio was pumped thru it. This noise sounded much higher than the top end of the woofer freq. At first I thought it was a voice coil issue. Then, remembering the previous prpoblem, I decided I would go thru and resolder most (maybe all) of the connections. Solved the problem again.

It's now more than 10 yrs. later and the sub still sounds (feels?) as nice as it always has. Moral of the story: (re)solder first, replace caps second. After that...more serious troubleshooting

Cheers....