Problem with preamp on audio amplifier

Hi everybody,

Skip to the bottom for the questions or read through for the fault description

Since I play electronic drums, I've been meaning to get hold of an active floor monitor (speaker with built in amp) for a while for gigs and stuff.

I found this faulty one brand new from a shop on ebay, so I decided to buy it as a little project.

There are a choice of two inputs into the amplifier, one is a microphone and the other is a line level input.

The issue with it is that the line-in input hisses when the level control is turned up.

The microphone input is perfect and there isn't a single whiff of ground noise when it's cranked to max.

If the input levels are set to zero and the master volume control is set to max there is no hissing either.

This leads me to believe that the preamp of the line-in input is at fault. However, I should point out that if I play an MP3 player or whatever through the line levels, it does produce the sound as it should, except there is hissing accompanying it. The hissing gets louder or quieter as the line-in level potentiometer is adjusted on the control panel.

So today I took the amplifier apart and it all looks fairly straight forward.

The back of the control panel of the cover (with the EQ controls and input sockets) has a PCB on the back with a single 4 pin connector going to the PCB from the rest of the circuitry of the amplifier (see picture of PCB removed).



Back of the PCB (which I photoshopped to show the various duties of the PCB tracks for reference purposes)...



The board appears to use 4580 IC, which is a duel operational amplifier and uses four of them in total on the PCB. The 16 pin IC is an LM13700 (Dual Operational Transconductance Amplifiers with Linearizing Diodes and Buffers)

The far left 4580 appears to be responsible for the Line level output.
Top left 4580 i think might be something to do with EQ
Top right 4580 I think deals with EQ and combining the two inputs (mic and line in)

Bottom right 4580 duel opamp seems to be amplifying the mic input on one side and the line level on the other side before going to either input level adjustment potentiometers (shown with orange tracks).




Questions
Am I right in thinking that the 4580 IC in the bottom right hand corner (which uses an opamp for each input) at fault?

What is the best way to test to see if it's a faulty IC?

Just work from the definition of an op-amp. DC measurements: Power supplies correct? Inputs within a few millivolts of each other? Output isn't locked at a supply voltage?

AC measurements: Give it about a dozen millivolts of 1KHz sine wave, output greater than input?

Thanks for getting back so quickly...

Power supply voltages were measured at -15.3volts and 15.3volts. So no problems there.

I attached my function generator to it with a 10mV output and the hissing is louder than the the input from the function generator.

Uh...1KHz doesn't sound like a hiss. It sounds like a sharp, penetrating, tone, about as high as an electric guitar can go. As far as the way you worded this, you didn't prove anything about gain.

You can purchase ten RC4580IDR from Farnell for £0.38 each. What's there to lose?

I would go ahead and replace IC1 in the bottom right corner.

I meant to add to my reply last night but a friend popped round and got diverted.

Anyway, I got a replacement through the post today for the op amp in the bottom right.

Unfortunately it hasn't cure the problem... damn!

I checked all the surrounding components and there doesn't appear to be any faulty ones.

Measurements then....

Between VEE and Input A (microphone side) - pin 2 = +15.2volts
Between VEE and Input A (microphone side) - pin 3 = +15.2volts

Output microphone side Pin 1
Between VEE = 15.2volts
Between VCC = 15.2volts
======================


Between VEE and Input A (microphone side) - pin 6 = +6.4volts
Between VEE and Input A (microphone side) - pin 5 = +6.4volts
Output Line in side (faulty side) Pin 7
Between VEE = 6.8volts
Between VCC = 24volts

So it looks that it is where the problem lies, would you agree?

I'm going to have another look at the board in the mean time and follow it round testing each component.

Output from VCC = 24 volts? That one is locked at the rail voltage. Problem in the feedback circuit.

So far all the components seem to work.

I have mocked up a draft version of the PCB...

There is no DC path to ground on pin 3. There has to be a leakage path somewhere, even if it is a million ohms. Bad design? You missed a part?

bum... I think I made a mistake with R6.
Also tidied it up a bit more...

OK. Now you have a ground reference. That means the input pins and the output pins must be centered around zero volts. So, how is the positive input being pulled down to negative 8.4 volts? And please quit measuring from Vcc and Vee. Use ground for your measurements.

Ok no problem.

So between ground and inputs of opamp....

Both have inputs of -8.4v and the output is -8.4v too

That's the problem. Find out where that voltage is coming from.
With the + input grounded through R6, it should be zero volts.
The feedback path has no DC reference so the opamp will put -8.4V on the negative input. It's only following instructions.
Find the voltage supply of -8.4 that is getting to the + input.

Literally just found the problem...

I removed the two diodes D1 and D3 which go to -15V to either inputs and hey presto it's working perfectly.

Checked the inputs into the opamp and it's sitting at 0V and the output is at 0V.


I tested the diodes and they seem to be both working perfectly as well.

Measured them both at VF 0.69V IF 4.60mA but no idea what model number they are for replacement.

So I'm just wondering why it's causing a problem.

Just put D3 back and it's still ok. So problem seems to be D1.

Maybe it's a zener that got in the wrong drawer.

bluebrakes said:

Literally just found the problem...

Measured them both at VF 0.69V IF 4.60mA but no idea what model number they are for replacement.

So I'm just wondering why it's causing a problem.

Click to expand...

Diodes can fail noisy and/or leaky .. same as many other components, and still test Ok on some meters using the diode test.

Those diodes are there for over-voltage protection on the inputs.
With 10k's for current limiting a small-signal type like a 1N4148 will be fine.

cool, thanks. I have ordered a replacement.

Your basic, one milliamp, diode test position is a fairly weak test. It pretty much only tests to see if it's a diode. "How good" of a diode isn't in the range that the meter will examine.