M1.0t died with a pop

Ok Rich, did the check for R172/R180= 145mV, R171/R179= 58mV
R174/R182= 144mV R173/R181= 59mV

Ok Rich, did the check for R172/R180= 145mV, R171/R179= 58mV
R174/R182= 144mV R173/R181= 59mV

Let's call that running twice and a half as hot as the good channel, and move back about halfway again.

Check the anode of D106 (non-striped) and cathode of D108 (striped side)
compare to the anode of D105 and cathode of D107

Rich, These were alittle tough to get. some were all over the place, but I think it was just from corrosion on the leads.

D106(nonstripe) 50mV continued to drop
D105(nonstripe) 33mV continued to drop
D108(stripe) 25mV and drops
D107(stripe) 5mV and drops

These don't look right at all, perhaps oxidation on the diode legs.

Let's remember (for me) how things are set up.

No input
feedback loop broken (input board connectors off)
bad channel's bias pot set in the same visualy position as good channel

Is this how she's set up still?

Let's ignore those readings and get some from the top and bottom of the bias network:

bad channel:
those three wire ribbon cables that go up to the bias xistor on the heatsink? pin 2 is the top of the network and pin 1 is the bottom (pin one should have a white stripe on the wire)

good channel:
Same thing

Rich, I did scratch the legs to get a good contact. Yes, still no input. No, the bias pots are not the same, the bad channel has been turned all the way down, or Q128 & Q130 get very very hot. Want me to set it to the same as the other channel?

I did turn up the bad channel pot to the same position as the good one and took readings again, basicly the same. However one thing I did notice since the input board is disconnected, Q128 and Q130 are not heating up, staying as cool as the others, and no loud buzzing from the tranformer even with the pot turned up.

These don't look right at all, perhaps oxidation on the diode legs.

Let's remember (for me) how things are set up.

No input
feedback loop broken (input board connectors off)
bad channel's bias pot set in the same visualy position as good channel

Is this how she's set up still?

Let's ignore those readings and get some from the top and bottom of the bias network:

bad channel:
those three wire ribbon cables that go up to the bias xistor on the heatsink? pin 2 is the top of the network and pin 1 is the bottom (pin one should have a white stripe on the wire)

good channel:
Same thing

Rich, did you want me to disconnect these ribbon wires for the readings? I'm getting some funky readings again! I have left the amp turned on now for over an hour and Q128 & Q130 are still cool.

....Rich, did you want me to disconnect these ribbon wires for the readings? I'm getting some funky readings again! I have left the amp turned on now for over an hour and Q128 & Q130 are still cool.

No, don't disconnect the ribbons, you can read them from the connector that's on those little boards bolted to the heatsink.

I did turn up the bad channel pot to the same position as the good one and took readings again, basicly the same. However one thing I did notice since the input board is disconnected, Q128 and Q130 are not heating up, staying as cool as the others, and no loud buzzing from the tranformer even with the pot turned up.

Hmmm....yes, I was assuming that both pots were in the same position when it looked like the bad channel readings were 2 1/2 times the good channel.

Remember how you were getting +100V and -100V to the inputs of Q108 and Q110? (pins 2 and 3 of MP102)

that's the kind of symmetry I was expecting throghoug this stage....a positive voltage on the top, a negative voltage on the bottom. but you're getting positive top and bottom, even on the good channel.

When you are taking these measurements do you have the negative probe on chassis?

Ok, did the readings and these are as funky as the others. Read it at the little board and down at the main board, both were similar.

Bad side
Pin 2= neg .12
Pin 1= .013v

Good side
Pin 2= .140.
Pin 1= neg .0149

I did turn up the bad channel pot to the same position as the good one and took readings again, basicly the same. However one thing I did notice since the input board is disconnected, Q128 and Q130 are not heating up, staying as cool as the others, and no loud buzzing from the tranformer even with the pot turned up.

Hmmm....yes, I was assuming that both pots were in the same position when it looked like the bad channel readings were 2 1/2 times the good channel.

Remember how you were getting +100V and -100V to the inputs of Q108 and Q110? (pins 2 and 3 of MP102)

that's the kind of symmetry I was expecting throghoug this stage....a positive voltage on the top, a negative voltage on the bottom. but you're getting positive top and bottom, even on the good channel.

When you are taking these measurements do you have the negative probe on chassis?

Yes I do, that's why these readings have me scratching my head!

Yes the neg probe is on the chassis, I've even moved it to different places to see if it would make any differince, didn't.

yeah, that's a puzzler. At least your good channel has a positive and negative reading....it's not balanced at all around zero, but it's there somewhat.

At those pins, the good channel has a span of 155mv (+140 and negative 15)
the bad channel has a span of 7mv (+20 and +13mv)
so we have something to go on here.

The whole point of this bias network is to keep the output transistor pair Q128/Q130 from turning off at zero crossing; this reduces crossover distortion ( a little 'notch' as the signal changes polarity) by keeping both polarities Xistors in their 'on' state. That rectangle of components between Q108/Q110 pair is the bias network. The next pair fed by Q108/Q110 is Q118/Q120. Q118 is a NPN (the little arrow points to the N type, in this case the emitter, so the base is P) and Q120 is an PNP, so it's base is P.

Q118 will be 'satisfied' by positive voltage on its base, the more P the more satisfaction (conducting).
Q120 will be 'satisfied by negative voltage on its base

The bias network is there to take that +100VDC and -100VDC Q108 and Q110 are seeing at their input, and adjust the span (difference between the top of the bias network and bottom) to satisfy the Q118/Q120 just enough to tickle Q128 and Q130 into 'idling'

BTW, when i get into what's happening here, it's half talking aloud, half 'in case you're interested' and half sanity check.

so, looking at SVR102, when you adjust for more resistance, the point at pin 3 of the ribbon cable should be more positive (further away from the negative rail), which would satisfy Q112 more, she'll conduct harder (in this case you can think of the transistor as a variable dimmer switch), and the span between pins 1 and 2 will decrease (2 will drop lower, 1 will rise).

This doesn't satisfy Q118 and Q120, it feeds down the chain and turns off Q128/Q130.

Similarly, adjust SVR102 for less resistance and pin 3 will decrease (closer to the negative rail) which doesn't satisfy Q112, which increases the span between pins 2 and 1. Pin 2 goes higher and pin 1 goes lower, satisfying Q118/Q120 and down to turn on Q128/Q130

SO, what you are experiencing, if the problem is here and not on the input board (get back about that third cable and the heating, etc) is a VERY small span at the bad channel between pins 2 and 1. Could be a bad Q112, for sure, but a bad R148, D102 or R150 could be the culprit too.

Let's turn power OFF and do resistance checks on Q112 as compared to Q111. Emitter to base should act like a diode, base to collector as well, and emitter to collector should be open (infinite reading, although in circuit you're liable to get SOME reading)



Also, you said it stopped heating the outputs and buzzing the coils when you unplugged the input board. Is the third ribbon cable (the power) disconnected too, or just the two sides?

The power is disconnected also.

I noticed your reading changed, and are now:

Bad side
Pin 2= neg .12
Pin 1= .013v

Good side
Pin 2= .140.
Pin 1= neg .0149


did you edit the post?

regardless,
the good span is now about 285mv centered roughly around zero
the bad span is now about 26mv centered roughly around zero

This is more what I'd expect, a positive and negative polarity, and it looks like the bad channel has a very small span across the bias network. I'd still check out Q112

I noticed your reading changed, and are now:

Bad side
Pin 2= neg .12
Pin 1= .013v

Good side
Pin 2= .140.
Pin 1= neg .0149


did you edit the post?

regardless,
the good span is now about 285mv centered roughly around zero
the bad span is now about 26mv centered roughly around zero

This is more what I'd expect, a positive and negative polarity, and it looks like the bad channel has a very small span across the bias network. I'd still check out Q112

Rich, yes I did, missed that negative sign again.

John