Fender M-80

[pchilson]

Alright, just so I know if I'm following your instructions correctly. I hooked a jumper from the center leg of Q3 to the center leg of Q5. With both rails hooked up the fuse blows. Is this the correct way of jumpering it?

[Enzo]

NOOOOOO!!!!

COnnect the bases together.

The center legs of the transistors are their collectors and are connected directly to the power rails. What your clip wire did was essentially what the transistors were already doing and we are trying to stop them. You had the collectors connected together. Fortunately that didn;t stress the transistors, only tyhe powr supply... and of course the fuse.

On these tab transistors, looking at the front, the legs are left to right Base, Collector, EMitter. So looking at the board, the four power xstrs are across the rear. The end ones are easiest to get at I think. So clip the left leg of the end ones together

The only purpose of this test at the moment is to see if our hard wired bypass of the bias makes it stop taking fuses. If it does, that tells us the bias circuit has problems still, but if it doesn;t help, then the next blow fuse tells us there are bad part out there.

[pchilson]

Connected the "left" leg of Q3 to the "left" leg of Q5. Fuse blows.

[Enzo]

Then the transistors have been compromised. On each one, check continuity from the center to the right legs - that is colletor to emitter. ANy that show continuity are shorted.\

STart again: power off - not like it can be on anyway - measure resistance at the transostor tabs. Tabs and center legs are internally wired together in these xstrs, so the tabs are also the collectors. Tabs of Q3, Q4 should be shorted together, but not to either ground, heatsink, or output bus. And Q5,Q6 should have tabs shorted to each other but not to ground, heatsink, or output bus. If CR19,20 were shorted, then the tabs of Q3-6 would show shorts to the output bus.

[pchilson]

On the 2 TIP142s Q3 and Q4 the resistance readings are 102 from the center lug to the right lug on both. On the 2 TIP147s Q5 and Q6 the resistance readings are 147 from the center lug to the right lug on both. One of the DMMs I have beeps on continuity but it does not beep on any of these tests. It does beep from center legs to tabs. No continuity from center legs to right legs on any of the xstrs.

Q3 and Q4 - tabs shorted to each other - true - to ground - false - output bus - false
Q5 and Q6 - tabs shorted to each other - true - to ground - false - output bus - false

[Enzo]

Here is a data sheet for a TIP142.
http://www.fairchildsemi.com/ds/TI/TIP142.pdf

It is a darlington, which means the part has two transistors in it along with a couple resistors and a protective diode. The internal diagram is on the front page.

You are getting 100-150 ohms from collector to emitter. That doesn't sound right to me. Extract them and see what they measure out of circuit. I expect to see the 8k or so between the outer legs, base to emitter, but not to the collector.

Let's see if the collectors are leaking to the emitters with that low resistance, or if the circuit board holds the mystery resistance.

Or do those resistances charge up to higher and higher and higher numbers? The filter caps charging in other words.

As i said, they could have been damaged when the diodes were open.

[pchilson]

Here is a data sheet for a TIP142.
http://www.fairchildsemi.com/ds/TI/TIP142.pdf

It is a darlington, which means the part has two transistors in it along with a couple resistors and a protective diode. The internal diagram is on the front page.

You are getting 100-150 ohms from collector to emitter. That doesn't sound right to me. Extract them and see what they measure out of circuit. I expect to see the 8k or so between the outer legs, base to emitter, but not to the collector.

Let's see if the collectors are leaking to the emitters with that low resistance, or if the circuit board holds the mystery resistance.

Or do those resistances charge up to higher and higher and higher numbers? The filter caps charging in other words.

As i said, they could have been damaged when the diodes were open.

All xtstrs out of circuit, read from collector to emitter
Q3 150 resistance
Q6 150 resistance
Q4 no reading
Q5 no reading

[Enzo]

OK, is that bias string straightened out now? It seems like one each on the two sides xstr failed, leaving the other two OK. REeinstall the two good ones where they belong - the amp will run with just one output xstr on a side. By no reading I assume you mean open - as desired. You have one remaining 142 good and one remaining 147 good, right?

With all the output xstrs out, what voltages now appear at their base connections? This checks the bias string.

If that works try it with the remaining Ok transistors. if that works, get some more new ones.

[pchilson]

OK, is that bias string straightened out now? It seems like one each on the two sides xstr failed, leaving the other two OK. REeinstall the two good ones where they belong - the amp will run with just one output xstr on a side. By no reading I assume you mean open - as desired. You have one remaining 142 good and one remaining 147 good, right?

I don't know if the bias string is straightened out now...???
Yes, one on each side failed leaving as you say the other two ok I guess.
from collector to emitter on Q4, Q5 no reading on DMM. Thats open? ok
If this situation makes them good then yes I have one good one of each.

With all the output xstrs out, what voltages now appear at their base connections? This checks the bias string.

Are you saying to fire it up without the xstrs in and read the base trace?

If that works try it with the remaining Ok transistors. if that works, get some more new ones.

Are you saying the amp will work with two of the xstrs "out of the amp"? Can you elaborate on this a bit?
Are you feeling this is all down to replacing these two xstrs?

Thanks

[Enzo]

An open reading E-C on a transistor is a good sign. It means not shorted at least. The transistor could actually be open, but open is a real rare fault in a transistor, especially when they are shorting all around it. So I'll bet on it being good.

The bias string is those four diodes, plus the couple resistors at each end to connect them to the power rails. You found the blown apart one or ones. And we needed to check the remaining ones for open. You know where they are on top, and you can;t get at them on top, so flip the board over and set your meter to diode test and measure the voltage drop across each one from the underside. Test one example of the diodes just sitting on your bench so you know about that junction drop to expect. I usually expect about .4-.5 volts on the diode test. I am not concerned over little differences, but any that are open need replacement, and if one has a real low drop, like .2v or something, pull it and test it. There shouldn't be any parallel circuits to screw up the readings I don't think.

Once we know those four diodes want to behave and those four resistors are close to value, we would expect about 2v more or less across the whole row of diodes (4 x .5v)

With the power xstrs reemoved, then there would be the resistors and diodes and not much else going on. Oh, the 1436 might try to move the whole string up or down towards one end to compensate for the missing outputs, but what we want to see is that 2v across the lot of them. Once that is there, we can worry if the whole thing needs to move, but as long as that 2v relation between bases exists, they won't turn on hard together.

Look at the output section, Q3,4 are wired in parallel with the exception of the .47 ohm ballasts for each. Those resistors are there to insure they share current. And Q5,6 are in parallel on the other side. The parallel connection just allows more current to be handled. REmoving one per side reduced the current output capacity - so instead of 50 watts, we'd be limited to 25...or something. But the amp will work with just one per side. This is exactly the same as pulling two power tubes from a set of four in a tube amp.

And remember, that clipping the left leg of the power xstrs on the two ends together does work, It just can't overcome shorted transistors. And those 150 ohm transistors were shorted just as sure as if they were zero ohms.

Small comfort to you, but in large power amps where there might be 6 or 8 power xstrs in a row on each rail, it is usually just one per side that shorts out when they blow. And in fact when one shorts, all the current flows through it, so it actually REMOVES the stress on the other parallel ones. And the flip side of the coin is this: when I work on some odd amp with rows of expensice xstrs trying to burn up, I will purposely remove all but one per side so if it burns up on my bench, at least it only burns up two expensive xstrs instead of 10.

[pchilson]

Enzo, the amp now operates with only Q4 and Q6 in circuit. Q3 and Q5 are removed from the board. Sidebar, the Normal channel has no output but the Drive channel does. Is this related to the missing xstrs or indication of something else?
I will place and order for the xstrs and replacements for the BYV26Ds. I'll just get all new replacements.

[Enzo]

As long as you are ordering, chose the TO218 ir TO3P version of those TIPs.

great, itseems to be coming together.

The power amp has no idea where the signal comes from, so losing half the output xstrs would have no bearing on preamp channel function, any more than pulling two power tubes from a Fender Twin Reverb would make the Reverb channel go dead. SO a dead preamp channel is a separate issue to resolve.

Might as well find out what that preamp channel needs before making an order.

[pchilson]

I agree. It looks to be a somewhat tough find on the xstrs I need and the BYV26D is an obselete product. I don't want to pay the NTE price and since you have warned me off them. Do you know a good source to find these items?

Discovery. When I had it on the bench I had all controls (pots) turned to zero. I was playing around with the amp again and reached down and turned the volume all the way up on the Normal channel and then started turning the TMB knobs and I started to get volume from the Normal channel. Mind you not loud, maybe equivalent to about 3 on the dial but the channel is not dead. So, when the all the TMB knobs are zeroed out there is no output on Normal but turn any of them up off zero and you start to get output...

[Enzo]

Look on the schematic at this classic tone stack. Your assignment: you tell me what we should expect through that stage with all three tone controls at zero.

Mouser has the TIP142/147 for $1.06-1.36, and they have plenty of fast recovery diodes cheap. The BYV has a slightly higher junction drop, but frankly a plain old UF4007 would run your xstrs a little cooler.

Allied has the 142/147 for $1.30 each, and they have in stock the actual BYV26D for 8 cents apiece.

Digikey has the 142/147, and lots of other diodes.

Make sure to get the larger transistors - TO218 or TO3P - not the TO220 as in your photo. ST has them in the larger still TO247. I'd avoid those as there may be a mounting issue.

www.mouser.com
www.alliedelec.com
www.digikey.com

[pchilson]

My homework assignment.
If all TMB controls are "zeroed" the signal is sent to ground but actually it looks as if the signal is sent to ground if only the treble pot is "zeroed". I humbly await my grade...

I have placed the parts order with Allied.
I oredered the TO218 type TIP142/147
I ordered the BYV26Ds and the UF4007s
Should have parts by mid week.
Anything else I should be aware of?

Looking forward to having this amp operational again. I have to thank you Enzo for sticking with me on this. I have learned things I had no other way of learning and I appreciate this.