Stereo Repair

[staze]

Okay, so a friend purchased a stereo receiver, largely untested, from a Goodwill here in town, and came to find out when he got it home that the left channel is largely gone (distortion, etc). Knowing I've taken up electronics, he handed it to me and asked me to fix it.

So, brought it home, hooked it up to the FG and Scope, and attached are the results. Bet you can guess which channel is which. My thinking is most stereos are push pull, so one set of transistors handle the positive side of the wave, and another set handle the negative. Obviously, the positive side is gone, and it appears to be oscillating. My thought would be the rail to those transistors, or maybe just a blown one, and the "oscillation" is just noise. I also note that the left channel has a negative DC offset relative to the right channel.

Any thoughts? I did manage to find the service manual online, so that helps. But holy cow, it's complex. For info's sake, it's a Pioneer VSX-452. Obviously it's not WORTH a whole lot of effort money wise, but experience wise...

Thanks!

[David_AVD]

Always check for the obvious causes first; blown rail fuse, cracked joints, etc.   

[Psi]

Does it have a amp chip or is it built from discrete components?

[staze]

Does it have a amp chip or is it built from discrete components?

Most of the circuitry is discrete. There are a few big ICs that mainly handle things like Dolby, etc. But the power amps, etc, are transistors.

[dr_p]

apply the 1st rule off ee troubleshooting: Thou shalt check voltages! 


I mostly mean the power supply, maybe load it up a bit, see how it copes with that.
I'm by no means an expert at this, but IMHO you can probe for that sine wave from the input towards the output stage and see where the shit hits the fan. That should narrow it down.


le: also, shorted semiconducors are a dead giveaway

[staze]

apply the 1st rule off ee troubleshooting: Thou shalt check voltages! 


I mostly mean the power supply, maybe load it up a bit, see how it copes with that.
I'm by no means an expert at this, but IMHO you can probe for that sine wave from the input towards the output stage and see where the shit hits the fan. That should narrow it down.

Yup. I picked up a couple 8R load (20W) resistors today to load down the channels. And yeah, the hope is to figure out where the signal goes wrong. As it appears to be, like most receivers, push-pull, and it splits the signal between positive and negative, it seems the positive side of the left channel is dead. But from what I've read, a shorted transistor should pull the rest of stuff down to ground (maybe). So yeah, guessing it could be the positive rail for that channel.

Will post back... haven't gotten to it yet since I've got a new baby in the house. =) Future Engineer (or the like), I'm hoping. =)

[smashedProton]

I would measure the bases of the power transistors first to see if it is a preamp problem or not.

[dfmischler]

... you can probe for that sine wave from the input towards the output stage ...

In my experience, the classic approach after checking power supply voltages is to inject the signal starting from the output stages backwards all the way to the input until you find where it stops working.  Make sure you are AC coupled if you have any doubts.  That is what my uncle (TV/radio repair business) taught me when he gave me an audio/RF signal generator in my teens.  I still have that old Paco G-30.  I never actually saw him work with a scope.

[staze]

... you can probe for that sine wave from the input towards the output stage ...

In my experience, the classic approach after checking power supply voltages is to inject the signal starting from the output stages backwards all the way to the input until you find where it stops working.  Make sure you are AC coupled if you have any doubts.  That is what my uncle (TV/radio repair business) taught me when he gave me an audio/RF signal generator in my teens.  I still have that old Paco G-30.  I never actually saw him work with a scope.

okay, wait... so are you saying you force the signal to run backward through the system, or just that you're isolating each stage and inputting the signal at the start of each stage, and if it leaves that stage intact, the stage is "okay" so move on?

Line level is 1Vpp... what signal do you feed into each stage in that case?

Thanks!

[David_AVD]

Safest and simplest way is to feed a sine wave into the aux input and look (with the 'scope) at the waveform at various points along the path from input to output.

I'd actually check the pre-out / main-in point first as that cuts your search in half straight away.  This is sometimes called a binary search method.

[dfmischler]

so are you saying you force the signal to run backward through the system

No.

... or just that you're isolating each stage and inputting the signal at the start of each stage, and if it leaves that stage intact, the stage is "okay" so move on?

Line level is 1Vpp... what signal do you feed into each stage in that case?

You've got it.

The old fashioned approach is to "inject" a signal into each stage and have it come out through the speakers so you can isolate the stage that isn't working.  When it doesn't come out you know you have found a problem.  You will probably be reducing the signal at each stage if they are all working.

Don't get me wrong.  This is the classic approach used by guys who had a signal generator, a VTVM and maybe not much else.  If you have a nice function generator and a scope you shouldn't be afraid to use them if that is easier/faster for you.  But with a little experience injecting a signal and working your way backwards through the amplifier will be quick and easy.  I remember fixing my brother's stereo that way when the only test equipment I had was a signal generator, and a cheap moving coil multimeter.

[Anks]

This looks more like a loss of drive to one of the rails on the faulty channel so check the prefianls. Normally the amp would shut down if the rail was shorted and would also pull on the supply for the other channel. Check the transistors if there is any shorted then you will also have bad emitter resistors. But this sounds like a poor design due to it not shutting down.

[staze]

Okay, so, I've verified that the signal is making it to the power amp "board". My biggest issue now is understanding how the circuit works. Attached is the schematic of that section (the front left is the top of the schematic). My understanding is that Q501 and Q503 should basically split the signal into the positive and negative halves of the wave for the left channel. I can see the signal on the base of Q501, I can't seem to see anything of note on Q503... but I'm not sure if that's normal or not.

Could someone shed some light on this?

Thanks!

Oops, forgot the schematic. There ya go!

[David_AVD]

The simplest way to check if something is normal is feed both inputs with the same signal and compare the same point on left vs right channel.

[edavid]

There's no splitting.  The positive and negative paths are both noninverting, so they are driven by the same signal.  Q501 and Q503 are a differential amp: input applied to Q501, feedback to Q503.

What do you see on the base of Q1?

[staze]

The simplest way to check if something is normal is feed both inputs with the same signal and compare the same point on left vs right channel.

That's what I'm doing. Right looks great. Left, only seeing the negative part of the wave, and it's pretty weak (relative to the right).

[staze]

There's no splitting.  The positive and negative paths are both noninverting, so they are driven by the same signal.  Q501 and Q503 are a differential amp: input applied to Q501, feedback to Q503.

What do you see on the base of Q1?

The negative half of the signal. Full amplitude. So it must be getting lost before there... further back the chain, Q511, I see only the half wave on the emitter and the half wave on the base with some capacitive coupling (think this is due to C513). Further back than that I'm not sure where to look... Q509's emitter, collector, and base all look identical. And Q505's collector shows the half wave (with some capacitance) but +55VDC rails on the base and emitter (which seems like it matches up with the schematic).

Where should I look further back?

I am seeing the full wave on Q3's emitter though.

Oh... or I could look at the schematic and notice the arrows. Looking, Q507 is the next further back, not 505 or 509 (not sure what they do in the grand scheme).

Q507 shows 55V on the emitter and base, with the half wave (with capacitance) on the collector. Which... now I'm at a complete loss...

[staze]

Nevermind... just reread answer.

[David_AVD]

If there are no missing rails or cracked solder joints, check the output transistors.  They are the thing that cops the brunt of it when the output is shorted.

Most of the time you can test them in circuit (power off of course) with your DMM on the diode test function.  Compare readings with the good channel.

[edavid]

Oh... or I could look at the schematic and notice the arrows. Looking, Q507 is the next further back, not 505 or 509 (not sure what they do in the grand scheme).

Q507 shows 55V on the emitter and base, with the half wave (with capacitance) on the collector. Which... now I'm at a complete loss...

Q505 is a current source load for Q507.  Q509 is a V_BE multiplier that provides the class B bias for the output stage.

There's something odd about your Q507 measurements.  If Q507's emitter is really at 55V, R519 must be open (or smoking?).  So, check your measurements, and also check R519.

[staze]

R519 looks good (100R). Q507 does indeed have -55VDC on it's emitter and base, and 0V on the collector. Resistance between emitter and base is about 2.75KR.

[edavid]

R519 looks good (100R). Q507 does indeed have -55VDC on it's emitter and base, and 0V on the collector. Resistance between emitter and base is about 2.75KR.

OK, -55V makes a lot more sense (you said 55V before).  At this point I'd be desoldering Q1 and probably Q511 and checking them out of circuit.

[staze]

R519 looks good (100R). Q507 does indeed have -55VDC on it's emitter and base, and 0V on the collector. Resistance between emitter and base is about 2.75KR.

OK, -55V makes a lot more sense (you said 55V before).  At this point I'd be desoldering Q1 and probably Q511 and checking them out of circuit.

Yeah, sorry about that, I was taking readings and posting in the other room.

Damn, okay. They don't make it easy to do with the whole heatsink, etc. guess I can just unsolder the legs from Q1 and test without removing the board. Q511 on the other hand. =P

Any chance you can explain how the circuit actually works? Obviously the signal comes in, gets decoupled by C501, and hits Q501, which amplifies it a bit from the low level it came in at. heads over to Q507, which...?

The big question in my head is if Q511 doesn't show the whole signal on it's base, but just the negative side, how could it or Q1, or Q511 be the issue?

Should Q503 show signal on it's base if it's providing feedback?

Is the point of Q1 and Q3 just to split the load of amplifying the signal? One is NPN and one PNP, which is where I was getting the whole push-pull thing. But since Q3 shows the whole signal...

[edavid]

Any chance you can explain how the circuit actually works? Obviously the signal comes in, gets decoupled by C501, and hits Q501, which amplifies it a bit from the low level it came in at. heads over to Q507, which...?

The big question in my head is if Q511 doesn't show the whole signal on it's base, but just the negative side, how could it or Q1, or Q511 be the issue?

Should Q503 show signal on it's base if it's providing feedback?

Is the point of Q1 and Q3 just to split the load of amplifying the signal? One is NPN and one PNP, which is where I was getting the whole push-pull thing. But since Q3 shows the whole signal...

Q501 and Q503 are a differential pair, used to measure the difference between input and output.

Q507 does all the amplifying.

Q505 provides the load for Q507, and drive for Q511.

Q509 provides the class B bias for the output stage.

Q511 and Q1 are the positive output stage.

Q513 and Q3 are the negative output stage.

(If you want to understand the circuit in detail, look for a copy of Doug Self's great book "Audio Power Amplifier Handbook".)

Now to answer your big question, either Q505 isn't providing drive, or Q511 and/or Q1 are shorting it out.  It's more likely that the output transistors are blown than Q505.  You could do some more probing around to convince yourself, but there must be some dead transistors there somewhere.

If the amp were working correctly, Q501 and Q503 would have the same signal on their bases.  You can see that Q503's base is connected to the amp output through R513.  But, at this point, I don't think you should worry about Q503.

[staze]

Q501 and Q503 are a differential pair, used to measure the difference between input and output.

Q507 does all the amplifying.

Q505 provides the load for Q507, and drive for Q511.

Q509 provides the class B bias for the output stage.

Q511 and Q1 are the positive output stage.

Q513 and Q3 are the negative output stage.

(If you want to understand the circuit in detail, look for a copy of Doug Self's great book "Audio Power Amplifier Handbook".)

Now to answer your big question, either Q505 isn't providing drive, or Q511 and/or Q1 are shorting it out.  It's more likely that the output transistors are blown than Q505.  You could do some more probing around to convince yourself, but there must be some dead transistors there somewhere.

If the amp were working correctly, Q501 and Q503 would have the same signal on their bases.  You can see that Q503's base is connected to the amp output through R513.  But, at this point, I don't think you should worry about Q503.

Okay, that helps. So I just need to get it out of my head that there's any negative and positive amplification, since Q3 sees the full wave?

I'll unsolder the legs of Q1 and see if it's shorted at all. I should just check between legs with ohms measurement, then diode? It should like I should just be able to go negative to base, then diode check collector and emitter (both should be the same), then move positive to base, and check collector and emitter with negative? One set should be high, the other low?