Oscillation

[raveenvijendren]

Good Day,


I would like to know how to I get rid of oscillation from the output of an amplifier? This oscillation which I have is around 20MHz at about 5mV peak to peak. Please tell me all the steps necessary to reduce or eliminate the oscillation. Thank you.

[MikeB]

Schematics ?
Have you already added outputzobel ? Basestoppers to drivers and output bjts ? 20mhz is quite high and smells like local oscillation in the outputstage...

Mike

[darkfenriz]

Check power supply first. It may be oscillating by itself.

[raveenvijendren]

Good Day,

My power supply has a "pi" filter network, along with 22,000 uF filtering per rail and ceramic disc capacitors from each rail to ground. The power suply can still oscillate with all this around? I will check the power supply, but other than that, where else a fault in the amplifier can cause oscillation? Thank you for your time.

[Stocker]

If you pick exactly the wrong combination of components, even the most complicated power supply filters can turn into a big resonant tank.

[amplifierguru]

Have you tried a 5 MHz scope?




Seriously it may be the CRO bringing it with the viewing window.

Cheers,
Greg

[raveenvijendren]

Good Day

The weird thing is, the oscillation problem is only in the left channel amplifier. The right channel is perfectly fine and both amplifiers share a common +/- 44V power supply (60 W RMS per channel). The two things i did that i think may be the cause:

1) To avoid corrosion of the hand drawn copper PCB, I sprayed the entire bottom of both amplifier PCBs with laquer, several times. This is the common wood type that is available at hardware stores. Could the oscillation be cause by the parasitic capacitance of the laquer?

2) Between the bases of the output Bipolar Junction Transistors, there is a 1 uF capacitor that is paralleled with a 150 ohm resistor. The capacitor is supposed to speed up the "turn off" of the output pair. I forgot to buy the capacitor when I bought the components for the amplifier, so I used a second hand one I salvaged from somewhere. The second hand capacitor had too short legs to be directly soldered to the PCB, so I ran about 1 cm of single uninsulated strand of wire to connect the capacitor to the PCB. Could the oscillation be caused by the inductance of the single strand wire together with the capacitance of the capacitor?

Please tell me if any of the two conditions above could contribute to the oscillation. Thank you.

[Bensen]

HI raveenvijendren,

I don't think the laquer is the problem because you did this on both PCB's, and only one has the oscillation. Using laquer is a good trick to keep your PCB's very nice throught the years, I also do this with every PCB on the copper side.

I don't know whether the capacitor of 1µ can be the problem. I would probably remove it or change it with another one, and do a test. Schematics would be very usefull here.

Good luck
Ben

[raveenvijendren]

Good Day,

Here are the schematics of my amplifier. It is an ordinary class AB amplifier capable of 60 W rms per pair of output BJTs at +/- 42V. Looking at the attached schematic, the capacitor which I was talking about was C9.

Also, I am using a 10 ohm ground lift resistor between the signal grounds of the amplifiers and star ground for both amplifiers. I did this to remedy a +40V DC offset which came on both channels. After adding that resistor, offset came to about 6.2mV for the right channel and 15mV for the left channel.

The left channel output transistor also blows for no reason. This could be due to oscillation and the weird thing is, only the NPN output blows but the PNP is fine. It is always the case.

If this does not help, I will take some pictures of the amplifier and paste it here for you to see it. Thank you.

[thanh]

Q12 Q13 ,protection tst
a 40n cap from B to C
a 47p cap is parallel with R10
good luck !