DC offset after power disconnected

[Pooley]

Hello fellow DIYers

I've just finished building one of a pair of active speakers using Nuuk's excellent guidance on Decibel Dungeon and Carlosfm's subberised power supply:

http://myweb.tiscali.co.uk/nuukspot/...ecircuits.html

http://myweb.tiscali.co.uk/nuukspot/...clonesnub.html

However the second amplifier has me confused. When measuring the DC offset it sits at a nice 21.9mV until the power is disconnected - then it increases to 23mV (fairly linearly) in 15 sec before dropping to -7.3 V and returning to zero slowly.

When monitoring the positive and negative rails independently they both fall off at the same rate after power off and at 15 sec are /pm 7.3 V respectively.

Any ideas/help would be greatly appreciative. I'm stumped the other amplifier doesn't show this behavior and as far as I can see they're identical. I don't really want to have to pull it to pieces and start again...

Thanks for reading

Tom

[lanchile]

well, that can be because of the caps discharging.and remember all "chips" do not have the same dc offset.check this picture from Peter Daniel.http://www.diyaudio.com/forums/attac...amp=1212163032

and when you said you turn it off it gets from 21.9mv dc offset to 23 mv dc offset all that is a little more than 1mv(that is nothing!).

But again ...I can be wrong.

[Mooly]

Are you getting 7 volts across the speaker or is that with no load ?

[Nuuk]

And is that with the source connected?

[AndrewT]

did you say that the smoothing caps discharge equally?

When monitoring the positive and negative rails independently they both fall off at the same rate after power off and at 15 sec are /pm 7.3 V respectively.

what does { /pm 7.3V } mean?

[Pooley]

Yeh I'm not at all worried about the 20 or so mV when the amp is running!

This is measured across a 10 ohm resistor. It must be related to the caps discharging because the change from -7 back to zero looks exponential, however there must be equal total capacitance on each rail because when watching the voltage on each drop after power down it is exactly the same.

This is with no source connected, does this make a difference? I'm away from home for a while so I can't check for a few days

[AndrewT]

7volts across 10ohms for a few seconds and decaying should not harm big bass drivers. But a small efficient bass/mid might not like that.

It sounds like this amp really needs an "instant off" relay on the speaker output. The same trigger signal could also pull in an input mute relay.

Finally the same circuitry could enable a "delayed on" for both relays.

It is quite likely that one smoothing cap has discharged to near zero volts while the other is still holding 7volts or so. This could be due to non equal capacitance values and/or non equal quiescent drain currents from the supply rails.
A high power bleeder resistor on each power rail will discharge the caps more quickly, but this should be relay activated as an "instant off" current draw.
Diodes between ground and each supply rail will help prevent reverse voltage building up on the first to discharge smoothing cap.

[Nuuk]

The snubberised supply has bleeder resistors.

It's quite normal to see a very brief fluctuation in the DC offset when you remove power although I have only noticed up to around 2 volts and that dropping to zero again in a second or two.

Do connect your source and see if that makes any difference.

[Peter Daniel]

Originally posted by Pooley
However the second amplifier has me confused. When measuring the DC offset it sits at a nice 21.9mV until the power is disconnected - then it increases to 23mV (fairly linearly) in 15 sec before dropping to -7.3 V and returning to zero slowly.

When monitoring the positive and negative rails independently they both fall off at the same rate after power off and at 15 sec are /pm 7.3 V respectively.

Any ideas/help would be greatly appreciative. I'm stumped the other amplifier doesn't show this behavior and as far as I can see they're identical. I don't really want to have to pull it to pieces and start again...

That is normal behaviour for a chip when no load is connected.

With 10R across the output you should not see any voltage fluctuations at power off. Are you shure the 10R load resistor was connected properly in both amps?

[Pooley]

Thanks for all the suggestions folks, turns out the problem was the reason I suspected all along - I was being an idiot.

I had rigged together a load of resistors in parallel to get to 10 ohms and despite this upping the power rating it was in sufficient and my little measuring jig had burned out Got a shiny new 10W beasite and the 7V jump turned into a 7mV jump.

Everything is working fine now, sorry for your troubles