puzzled

[sam18]

can anyone tell me why my amplifier i have built is doing this when it starts up? after it has done this it works fine

http://i1094.photobucket.com/albums/i452/Sam_Burridge/osiliscope.jpg

can upload schematics and psu if needed thanks.

 

[Osvaldo de Banfield]

This is a conditionally stable circuit. Perhaps a bad closed loop margin. In other words, try another way to feed back the amplifier, and try first to fix open loop issues in it.

 

[sam18]

and how would i go about doing that? sorry im still new to building amplifiers

 

[Osvaldo de Banfield]

Review RC time constant in the amp.

 

[sam18]

sorry i still dont understand

 

[DF96]

Yes, looks like conditional stability either in the main feedback loop or from parasitics in one stage. We need more information. Tell us everything you know about the amp.

 

[sam18]

Amplifier Specification:
Maximum Output: 240 watts rms into 8 Ohms, 380 watts rms into 4 Ohms
Audio Frequency Linearity: 20 Hz – 20 kHz (+0, -0.2 dB)
Closed Loop Gain: 32 dB
Hum and Noise: -90 dB (input short circuit)
Output Offset Voltage: Less than 13 mV (input short circuit)
Phase Linearity: Less than 13 0 (10 Hz – 20 kHz)
Harmonic Distortion: Less than 0.007% at rated power
IM Distortion: Less than .009% at maximum power

 

[DRC]

Is this a simulation or physical circuit ?
Is this just when power is first applied ?
A schematic would be good ...

It would be worth checking stability anyway but even an otherwise stable amplifier can sometimes have problems when the power rails come up during the first few hundred mS.

dc

 

[Nico Ras]

That looks like a simulation to me or is it a real life oscilloscope plot? From just looking at the part of the circuit that I can see it is a dual differential Hitachi based MOSFET amp. These normally would not cause the problems that you described. I know of several thousand out in the field.

If it is a real plot of the output I suspect from the first part that there are unbalanced current draw from the power supply or different time constants thus the two rails are not firing up in a balanced manner at start up but settles after a few cycles of power.

It would appear as if the problem have a time constant around 300 mS, so look for something in the power supply or decoupling caps that has a time constant around this value.

Also check the time the input differential amplifier bases are charged up, they may be charging at quite different rates which would be caused by a large NFB decoupling cap and a small input cap.

Some designers by-pass the NFB decoupling cap with a diode to stop DC voltages of more than 0.7V to develop across them.

That is my two cents worth. Good luck.

 

[sam18]

this is just a simulation, im hoping to build this circuit soon. the +45v dosnt seem that stable and seems to drop to about 40v then goes back to 45v

 

[Osvaldo de Banfield]

I strongly dislike the input stage. It is not DC coupled, and with single supply. When this stage starts, all Caps are discharged, and while doing it, it impulses a CD peak trough the final stage causing it to saturate, and causing the voltage sag. Review this first preamp, please. And the coupling to the discrete stage.

 

[Nico Ras]

I see another possibility, that the background circuit is not input DC decoupled so this could generate a large bang on the speakers at turn on and also a high output off-set.

 

[sam18]

so your suggesting a different first stage? have you got any schematics for a diffrent pre amp?

 

[Osvaldo de Banfield]

Yes: try removing it at all.

 

[Nico Ras]

Do not expect the claimed output at 40V (full load) supply rails into 8 ohms, it will more likely be 75 watt.

 

[sam18]

what would you suggest for a first stage? because i tried it before and it didn't work without a first stage. i only want a volume and bass control for the first stage, thats why i chose the other one.

 

[Nico Ras]

The input stage is a differential amplifier and to my knowledge there is nothing wrong with this kind of input topology. You can make it much more complex adding current sources and mirrors it in its current form it works fine.

 

[sam18]

right so how do i make the second stage work then?

 

[Nico Ras]

Place a decoupling capacitor of about 10 uF between the tone control op-amp and the input of the power amp. This should solve much of the problem.

 

[Osvaldo de Banfield]

This is the real trouble: try make work the last stage, then build the preamp. ONLY if the power stage is properly, then start job at the pre. And, try to use a split supply also for this first stage.