turn on noise

[junglejuice]

Hi guys i have a pheonix gold m50 amplifier that has a loud turn on thump. I have measured the output of the amp at
turn on and it gets an ac wave form of 43hz for about 120ms with a peak voltage of 6v which
tapers back to zero. The wave forms are the same for each channel but 180 degrees out of phase.
This spike occurs at the exact time that the voltage starts to rise on the high voltage rails.
My question is does anybody have any thoughts as to how i can fix this or better still does anybody
have the circuit for this amplifier.
Any help will be greatly appreciated.....

[junglejuice]

Is there no one on this forum that has any ideas on how to fix this problem????

[vbd]

The simplest thing to do, without any circuit diagram of the amp : add a delayed turn-on relay circuit... See for instance another aussie's project :
http://sound.westhost.com/project33.htm

[BrianL]

The slow turn-on should help, but not always. One could
also put a relay in series with the the output that delays
connection of the speaker until the amp settles.

But why not just let it thump? A number of amps I've
had over the years thump on turn-off (sometimes turn-on)
and it's never damaged anything, including when I was
triamping and tweeters were directly connected to the amp's
output.

So, why worry about it?

[AKSA]

Junglejuice,

Turn on/off thump is very difficult to eliminate.

It is USUALLY (there are always exceptions!) caused by the power supply to the diff pair, which controls voltage offset at the output.

Most diff pairs are fed from current sources (CCS), and these will come to life within a couple of volts of switch on. However, the rest of the amp takes more volts that this to properly bias up, and strange things happen.

I have found that replacement of the CCS with a simple resistor allows the diff pair to come to life at the same rate as the rest of the amp, and this helps considerably. Many designers don't like resistive feed, however, since it is just too simple! However, at switch off there is then an infrasonic pulse which takes the speaker cone to full excursion for a second - you can't have it both ways.

The good news is, as Brian says, these pulses contain very little energy, well below the rated dissipation of all real world voice coils, and no damage is ever done.

If you like eggs, you must break shells.......

Cheers,

Hugh

www.aksaonline.com

[traderbam]

"Many designers don't like resistive feed, however, since it is just too simple!"

I don't like resistive current biasing because it compromises the CMRR and the PSRR of the amp!