• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

My amp has 600mV on the input jacks--how do I fix this?

jhoyt

jhoyt

Incomprehensible space llama
I'm working with an amp that was emitting a terrible turn-off noise. Come to find out, there is 600mV across the input jacks. Shorting them out sends a DC spike out to the speakers.

Unfortunately, I do NOT have a schematic for the amp, nor am I likely to find it. It is unobtanium AFAIK. However, the topology looks to be fairly straightforward.

Any ideas where this DC voltage could be coming from? How do I get rid of it?
 
Register to hide this ad
Are the power supply rails symmetrical? Same (-) voltage as (+)? If something is out of whack there, you may find bias at the signal grounds...

What are the offsets at the speaker outs? Does that change when you short the inputs?

John
 
DC on the input line could be entirely normal, if you're measuring with nothing connected there other than your voltmeter. All coupling caps have some amount of leakage, so if there's no input bleeder resistor, that leakage could easily overcome the voltmeter's (very high) input resistance.
 
Tricky without a schematic or knowing the layout of the amp's input stage. Is it possible to draw it out by hand?

One thing I'd try is putting a 47k resistor from each input to ground, and seeing what effect this has.
It could well be leaky DC blocking caps causing the voltage. If the resistors stop the voltage appearing at the inputs then I'd be looking to recap the preamp stage.
 
jdurbin1 said:
Are the power supply rails symmetrical? Same (-) voltage as (+)? If something is out of whack there, you may find bias at the signal grounds...

What are the offsets at the speaker outs? Does that change when you short the inputs?

John
Click to expand...

I won't be able to fire it up until later (to measure voltages), but both input stage rails are on 7815/7915 regulators. The output DC offset is controlled by a servo, but when the inputs are shorted the DC at output spikes up and off the meter.
 
BinaryMike said:
DC on the input line could be entirely normal, if you're measuring with nothing connected there other than your voltmeter. All coupling caps have some amount of leakage, so if there's no input bleeder resistor, that leakage could easily overcome the voltmeter's (very high) input resistance.
Click to expand...

The input has a 100 ohm resistor and 100 pF film cap (in series) shunting to ground. DC resistance (from input to ground) measures at just under 50K ohms.
 
BinaryMike said:
DC on the input line could be entirely normal, if you're measuring with nothing connected there other than your voltmeter. All coupling caps have some amount of leakage, so if there's no input bleeder resistor, that leakage could easily overcome the voltmeter's (very high) input resistance.
Click to expand...

Most likely, I say this as well.
 
Maybe this amp is DC coupled on the input... If so, my guess is that something's very wrong in the input stage, or possibly in its current source(s), but the servo has enough range to null DC offset on the output side despite the failure.
 
There is a .470uF film cap in the signal path, so it's not DC coupled. My next step is to re-cap the circuit board, and see what that does.
 
Well, I'm an idiot. I took another look at the circuit today, in order to replace the "DC-blocking" caps--and I was wrong; the caps couple to ground. The audio path is DC all the way. So the first thing the input sees is the (presumable) diff amp transistors.

Ergo, could all this DC be coming from "leaky" transistors? Allowing DC to pass when they're not supposed to?

Sorry. Sometimes I'm a dunce.


:arrow: :withstpd:
 
Not necessarily leaky, but yeah; it's likely to be some sort of defect in the input stage transistors or their power sources.
 
curious, maybe stupid, but do you think they made it that way to compensate for something counting on having something plugged into the jacks ? then that would short out the voltage.

and or, not engineered to care about it considering normal use would be with something always plugged in anyway.?
 
Good thought. But actually the inputs don't short with something plugged in. They do short, however, when switching or muting the inputs. That sends a little DC voltage spike through the amp which comes out as a great big spike into the speakers.
 
Hmm, transistors do the muting ? if so, i might look at that area. buffer something or another type of thing shorted.
 
Muting would take place in the preamp.

I suppose I should try to find some replacement small-signal transistors for the input stage (MPS8099) and switch them out.
 
You must log in or register to reply here.
Back
Top Bottom