Distorted Sound From Wharfedale EVP-X12PM Monitor

[paul_g_787]

I am hoping someone can give me some ideas with this one.

I have here a Wharfedale EVP-12PM Active Monitor and the sound coming out of it very distorted!!

I have found the service manual on the internet and uploaded it to ElektroTanya here:
https://elektrotanya.com/wharfedale_pro_evp-x12pm_sm.pdf/download.html

So far I think I have narrowed it down to the power amplifier. (Schematic on Page 7).
Please see attached page 7 of the service manual with the schematic for my model outlined in red

I have probed the signal from the input of the power amp (pin1 of the connector) and listened to it using my headphones. At this point the signal is not distorted and sounds perfect! 

I have checked the power supply voltages:
I have confirmed that both the + and - 15V rails are stable (both reading 14.8V DC) on both the pre amp and power amp PCBs.
And also the + and - voltages of the power amp (both at 66.2V DC) are nice and stable.
And the MUTE pin of the power amp connector is steady at 13.8V.

So I am pretty certain it is the power amp circuit at fault somewhere.

Any tips of what I might look at next?

Could anyone point out to me which type of power amplifier this is? e.g. (Class A/B).

My knowledge of power amps is quite limited so any help is very much appreciated. 

[Audiorepair]

1.  Are you sure it's not the speaker that is distorting?
2.  Check in circuit for shorts on all the output transistors, and measure all 6 white ceramic 0.22 ohm emitter resistors, they should all read the same.

3.  If not the above then suspect the JFets Q1 & Q2,  You could cut the legs of these that attach either side of R111 to remove them from the circuit.

[Audiorepair]

The Mute pin should be at -12v, not positive as you state, so you should get the green power LED lighting.

The mute on the gates of Q1 and Q2 should also be negative for these Fets not to be muting the signal.
The anode of D11 is probably the easiest place to measure this.  Should be around -12v.


Edit: it's not the Anode, is it, its the "pointy end".
Cathode.

[bob91343]

If you can put in a sine wave and look at the output with an oscilloscope, you may find the answer.

Distortion takes many forms.  There is clipping, crossover, nonlinear operation, etc.  Once you identify the wave shape you will be halfway there.

Also, rubbing of a dynamic speaker cone over its travel causes distortion.

[paul_g_787]

1.  Are you sure it's not the speaker that is distorting?

Good idea, I just borrowed a speaker cabinet and connected it to the speaker output jack and when playing back both the internal and external speaker sound distorted.

So at least I know the speaker and filter are ok. So now the fault is definitely somewhere on the power amp board.

The distortion kinda sounds like clipping mixed with some very fuzzy noises.. sort of.. It is hard to describe.

The Mute pin should be at -12v, not positive as you state, so you should get the green power LED lighting.

The mute on the gates of Q1 and Q2 should also be negative for these Fets not to be muting the signal.
The anode of D11 is probably the easiest place to measure this.  Should be around -12v.

Edit: it's not the Anode, is it, its the "pointy end".
Cathode.

You are quite correct sir, twas just a typo. I just forgot the '-' on my original post! 

[paul_g_787]

2.  Check in circuit for shorts on all the output transistors, and measure all 6 white ceramic 0.22 ohm emitter resistors, they should all read the same.

Just tested the four power transistors and also all the other transistors for good measure. There are no shorts across any of these which is a good sign.
Also the four (not 6 on this model) emitter resistors all measure 0.22 ohm (+- 0.01 ohm). So these seem fine.

3.  If not the above then suspect the JFets Q1 & Q2,  You could cut the legs of these that attach either side of R111 to remove them from the circuit.

I can't cut them off as they are so crammed in I cannot get my wire cutters on them. I will have to de-solder them, (typically requiring all the transistors to be un-bolted from the heat-sink) so this is a later tonight job.

However measuring them in circuit I noticed that on both Q1 and Q2 there is 40 ohms between D and S on both of them (in both polarities). Does this sound correct?

[Audiorepair]

Ha, the reason I suggested cutting the Jfet legs is that to unsolder them requires taking everything off the heatsink.

Messy, time consuming, and prone to damaging stuff as you will no doubt find out.

And testing a power amp with the output transistors in free air will make those transistors very hot very quickly even with no load, and liable to thermal runaway and magic smoke.
So you are going to have to bolt everything back on the heatsink again after desoldering the Fets.


I don't think you can get any kind of reliable info from measuring the Jftets in circuit, other than a short.



You haven't given any indication of what this distortion actually is.

I'm guessing you don't have a scope.





Edit:  Ah, sorry, I missed a post back there where you did describe the distortion.

[paul_g_787]

Ha, the reason I suggested cutting the Jfet legs is that to unsolder them requires taking everything off the heatsink.

Messy, time consuming, and prone to damaging stuff as you will no doubt find out.

And testing a power amp with the output transistors in free air will make those transistors very hot very quickly even with no load, and liable to thermal runaway and magic smoke.
So you are going to have to bolt everything back on the heatsink again after desoldering the Fets.

Yes but that's part of the fun is it not! I will be able to do it just fine though.
I will take off those two JFETs you suggested and then put it back together and test if the distortion is gone.

I don't think you can get any kind of reliable info from measuring the Jftets in circuit, other than a short.

Can you give me any pointers regarding testing the JFETs? I have never actually tested one before.

I only have a multimeter at the moment. Perhaps I should invest one of those little ATMega 328 component testers.

You haven't given any indication of what this distortion actually is.

I'm guessing you don't have a scope.

Edit:  Ah, sorry, I missed a post back there where you did describe the distortion.

Would a recording of the audio be useful in any way? I can always record it on my phone and link it here.

I don't have access to a working scope at the moment. I used to use the one at my old job, and my HeathKit scope I have inherited has not been turned on since 1981 to my knowledge. Something I will have to have a look at soon I reckon.

[Audiorepair]

I don't think there is much point testing a Jfet.

What seems fine at low voltage is not what might happen in circuit.


Easiest way to test it is to take it out of circuit and thus let the circuit test it.

Or just replace it, they are very cheap.

[paul_g_787]

Well before removing those JFETs I decided to do some more probing using my headphones.

To do this probing I am connecting the input of a portable 9V battery guitar amp to the points on the board using my multimeter probe and a crocodile lead so I can hear the signal from the probe, and I am using headphones from the guitar amp.

I tested one side, then the other side of R11 (where the two suspect JFETs are connected) and the signal is not distorted at these two points (see attached image).

So I would think this rules out the JFETs would you agree?

[paul_g_787]

Some strange behaviour just now.

While I was probing I disconnected the speaker from the output so I could hear my headphones without a big 12" speaker in my face.

I connected it back up and tried running it again and the distortion was gone for about 20 seconds!

Then it started to hum and the hum started getting louder and the distortion faded back in.

Very odd. Perhaps a clue? 

[SilverSolder]

Sounds like a bad capacitor somewhere?  Have you looked at the ripple & voltages on the supplies?

[paul_g_787]

Sounds like a bad capacitor somewhere?  Have you looked at the ripple & voltages on the supplies?

Yes possibly! Perhaps power supply capacitor?

I will measure the ac on the different rails to check the ripple.

Edit:
Just taken some readings from the power supply using my multimeter.

These are the DC readings from my power supply rails:

+70V: +66.8V DC
-70V: -66.8V DC
+15V: +14.8V DC
-15V: -14.8V DC
-12V: -13.8V DC

These are the AC readings from my power supply rails:

+70V: 0.06V AC
-70V: 0.05V AC
+15V: 0.06V AC
-15V: 0.06V AC
-12V: 0.04V AC

So the ripple looks quite low to me.

I also noticed that if I turn up the volume on the monitor really loud there is a very quiet but prominent 100Hz (to my ears) hum.

[Audiorepair]

You didn't measure the AC on the mute circuit.

Are C6/C7 OK?

[paul_g_787]

You didn't measure the AC on the mute circuit.

Are C6/C7 OK?

Yes the -12V is the mute circuit. Perhaps I should have stated it more clearly.

I will check those two capacitors with my ESR meter before removing the JFETS.


I have confirmed that after leaving the monitor unplugged for 15 mins or more to allow the capacitors to discharge, I can repeat the process every time:
I turn it on and the sound is not distorted for about 20 seconds then the distortion fades in.

[TimFox]

That evanescent behavior may be due to an oscillation that starts after something warms up.  In analog amplifiers, a really good oscillation (even at supersonic frequencies) will totally mess up the performance in the audio range and DC.

[Vicus]

The scope will tell the true. So easy and fast.

[Audiorepair]

You didn't measure the AC on the mute circuit.

Are C6/C7 OK?

Yes the -12V is the mute circuit. Perhaps I should have stated it more clearly.

I will check those two capacitors with my ESR meter before removing the JFETS.


I have confirmed that after leaving the monitor unplugged for 15 mins or more to allow the capacitors to discharge, I can repeat the process every time:
I turn it on and the sound is not distorted for about 20 seconds then the distortion fades in.

The mute -12v goes to R7 and C7 , and then to the mute fets which you still haven't disabled to rule them out.

R7 is 1 Meg.   There is plenty of scope here for failing components to AM modulate the signal via the mute fets.
Hence the hum.


Take out the fets, and then progress from there, having at the very least removed the mute circuit as a possible candidate.

[Audiorepair]

The scope will tell the true. So easy and fast.

Well actually it won't.

Audio amplifiers have feedback, so any fault anywhere will turn up everywhere, you will not find the point where it all goes wrong with a scope.

[paul_g_787]

The scope will tell the true. So easy and fast.

Well actually it won't.

Audio amplifiers have feedback, so any fault anywhere will turn up everywhere, you will not find the point where it all goes wrong with a scope.

Interesting point you say that actually. It has made me think; what are the chances that perhaps the output section is struggling to provide enough current and then the feedback is trying to compensate buy upping the voltage therefore causing distortion?

Just an idea. Will try the capacitors and JFETs in the meantime.

[Audiorepair]

Amplifiers do not usually fail in a distort mode.

They either blow up or go DC.


I think it would be better to look for reasons external to the amplifier itself why this is not working.

[SilverSolder]

Have you looked at the DC level of the amp output?  Maybe it isn't zero...

[Vicus]

The scope will tell the true. So easy and fast.

Well actually it won't.

Audio amplifiers have feedback, so any fault anywhere will turn up everywhere, you will not find the point where it all goes wrong with a scope.

In this case I think it will be pretty clear.

[paul_g_787]

Have you looked at the DC level of the amp output?  Maybe it isn't zero...

Yes first thing I checked was DC across the load. There is no DC at all. 

[paul_g_787]

Just tested the voltage on D11 cathode and it is -11.6V

So I just removed Q1 and Q2, C6 and C7.

The distortion fault remains with these components out of circuit and so does the hum.

Tested C6 and C7 and they are both spot on 9.8µF and working fine.