dc on output of leach

[Wagener]

Hi, My leach amp worked perfectly for the past month now.

I have been running two 8-ohm 175Wnom subwoofers (in parallel giving a 4-ohm load) on the one channel only

Yesterday I had it running the whole day. During late evening I heard a buzzing sound from the subs. I immediately switched the amp off.

I took off the subs and measured the output of this channel. From time to time I get 20Vdc on it. It drops and then goes away for some time (+- 20 secs) then appears again.

I don't think I could have overloaded the leach because I constantly touched the heatsinks of the channel. It never became so hot that I could not hold my hand on it.

What could be the matter? Where shall I start looking for the problem?

[janneman]

Originally posted by Wagener
Hi, My leach amp worked perfectly for the past month now.

I have been running two 8-ohm 175Wnom subwoofers (in parallel giving a 4-ohm load) on the one channel only

Yesterday I had it running the whole day. During late evening I heard a buzzing sound from the subs. I immediately switched the amp off.

I took off the subs and measured the output of this channel. From time to time I get 20Vdc on it. It drops and then goes away for some time (+- 20 secs) then appears again.

I don't think I could have overloaded the leach because I constantly touched the heatsinks of the channel. It never became so hot that I could not hold my hand on it.

What could be the matter? Where shall I start looking for the problem?

How do you measure it? With a multimeter on AC or on DC, unloaded amp, music playing? Or shorted input?

Jan Didden

[Tony]

turn the volume to zero then measure the outputs, if problem go away, your problem is not the leach amp.

[Wagener]

I measure a dc voltage with a multimeter on the output when there is no signal connected to the input of the amp.

The dc voltage is not constant. It appears and dissappears.

[janneman]

It CAN be misleading. Make sure that you short the input. If not, there may be some spurious input noise or something that occasionally gives an output pulse. The multimeter registers that. How is the spacing of those 'DC' signals, in time?

Jan Didden

[Wagener]

apart from the dc on the output of the one channel I found the following: The other channel had both its fuses blown and there is something big wrong there. I measure a short between collector and emitter of all the output transistors in there (with the whole circuit connected up.

I do not find this short on the other channel which gives out dc.

So basically the one channel is completely fried. Can get it running without popping fuses / 100R resistors.

How do I know for sure it is the output transistors without disconnecting them? A short over CE should be a good indication?

I don't know how this could have happened, I only used the channel which is now giving out dc.

Also, the dc last for about 15secs at 20V and then decreases to about 18V, then dissapears completely for some 15secs and then appears again. No fixed pattern but definitely not noise!

[JohnW]

I wonder if the amp has gone unstable - could explain the fully blown second channel, which from accounts blown when not used / loaded. Is the Leach design MOSFET OPS?

Badly drifting DC and blown channels is a good indicator of an unstable amp.

I've seen amps that go unstable while varing the gain control - I guess the change of input impedance....

Unstable amps can be sometimes be hard to see on a Diyer's scope, say 500MHz @ 1uV....

John

[janneman]

Amps don't just decide to go unstable on a nice sunny day. The Leach is a very stable and mature design. I wonder what has been going on when these catastrphes happenend.

Wagener, are you SURE you just were listening happily and all of a sudden poof! Don't be embarassed, we all have our bad moments, but if there was something unusual, like you were just changing speakers or CD players when it happened, let us know as it can save both you and us a lot of time and frustration.

Anyway, I would first tackle one channel, like the one that has the spurious DC. Trying to do two different cahnnels with different problems at the same time will make sure you get nowhere at all.

Make sure you short the input. Don't assume that no input means zero input voltage. And load the output with a resistor, not necessarily 8 ohms, 10 or 20 ohms is also fine. Then measure again. Try to follow the error. For instance, if the output jumps up, what does the other side of the feedback resistor do?

Jan Didden

[JohnW]

Yes its true that an amp should not need Sun Block, but there is no guarantee that the amp was not already on the edge of instability (or unstable, but oscillating at a low level) – only to be pushed over the edge by different speaker / input cables / loads. From the sound of Wagener’s original post, it’s the first time the amp had been really used in anger…

A “Mature” circuit design can be stable, only to become unstable due to PCB / construction layout – circuit diagrams only tell half the story.

Wagener, if you have excess to a decent scope, then theres no harm in doing a quick check for HF / RF “noise” on the speaker outputs while the amps in its “DC” condition.

Unlikely but (20V is too high)…. you could also check that the Capacitor to ground on the negative Feedback input has not gone leaky..

Best of luck,

John

[pooge]

I'd like to 2nd checking out the feedback cap. I don't know which brand of cap you have, or when it was bought, but 3-4 years ago some bad caps were made in Japan with bad electrolyte that ended up leaking. Concentrate on the input stage for other source of imbalance.