Fender 212R power amp blown

[MarkusBass]

Q11 should turn off because there is zero current going through Q10 collector.

Yes, this is the most important statement. Q11 should not be on when the amp is muted (because Q10 should be also off). Since, when the amp is unmuted, everything seems to be OK, I would look more precisely at the base of Q11 or going further; is Q10 conducting when it shouldn't?
Q13 "pulls down" the output voltage despite the fact that Q12 is off. It's because it gets positive voltage through R91-95. Also some problems with D28-29 are possible (I suggested to disconnect them a long time ago) but I would rather look at Q10 (is it conducting or not).

It's hard to use a voltmeter on things like this because even the input impedance of a simple meter is low enough to affect the voltages on a thing like a transistor collector being held off.

This shouldn't be a problem with a digital multimeter which has input impedance in range of 10 MOhms.

I'm also deeply suspicious of those resistors marked "FU47" and "FU470". I'm hoping that means "fusable" instead of some other connotations I can think of. If they are fusable, they may also be non-obviously damaged. They're worth checking.

The are "fusible" resistors. Just look at google or here: Vishay - Resistors, discrete - Fusible

But.....
I wouldn't bet my lunch money that this is not by design. And we are trying to fix Fender's mistake . Are phone calls really that expensive in the States?

Mark

[R.G.]

This shouldn't be a problem with a digital multimeter which has input impedance in range of 10 MOhms.

I agree - it *shouldn't* be. However, I'm just relating my experience from other times. Instruments have fooled me enough times that when I read things that don't make sense, I also remember to check whether the instruments are the issue. High impedance things are one of those situations. May not be here, but then some DMMs have quite low (appx 1M) input impedance too.

The are "fusible" resistors.

Cool. So if the amp had a trauma, one or more might be partially fused?

But.....
I wouldn't bet my lunch money that this is not by design. And we are trying to fix Fender's mistake .

I agree here. The rest of the circuit is an odd enough design that I'm forced to kibitz.

If the whole idea is to mute the output at times, why didn't they use one of the JFETs they used liberally in the preamp section and cut off the signal into the power amp when muted instead of messing with turning off the front end half of the power amp? I notice that they couldn't be bothered to turn off any current into the power section, leaving it fully armed if there happens to be noise or transients. Mucking with the DC feedback section of a power amp and interrupting/restoring DC feedback in the middle of a power amp is asking for trouble.

I have held up promotions for engineers working for me who brought stuff of comparable clumsiness into design reviews who could not defend the silliness. It may be there is some good technical reason other than an urge to profundicate and profusticate the design, but it is not apparent.

[J M Fahey]

Dear friends, I *hate* seeing you waste so much time on this poorly designed amplifier.
The basis of the problem apparently lies on the muting system chosen.
I tried to add it to my amps as an output protection, spent some time on that project (I know there are some amps who use it successfully) but ended up using a relay.
What makes it so difficult to implement is that the differential input pair does many things; one of the most important is DC balancing the amplifier.
Turning it off kills it (duh) so it balances nothing any more.
I got exactly the same results, a few volts offset which I didn't like, so I opted for a Relay (They are *very* cheap now). End of the story.
But this designer insisted, so he:
1) added a servo (U6-A) ... that "solves offset problems", doesn't it? ... but it didn't , so:
2) to balance that Op Amp driving one side of the Diff Amp he used the other Op Amp (U6-B) driving the other half of the Diff Amp, so both would see the same impedance yada yada yada ... but it didn't work,
so he borrowed a page from Hi Fi designers, straight from the Hitachi Mos Fet amp application note, and
3) "symmetrized" the class A gain stage (Q12) not only by using an active load (Q13) but by turning it into a current mirror, by adding D27 and R89, which of course were current driven, out of phase, by Q11 (the villain-of-the-day) which is driven by "the other half" of the Diff Amp , Q10.
4) I forgot to say that it also got turned into a current mirror, by adding D24 and R84 .
Still didn't work, so he went one step further, and
5) symmetrized Bias, by splitting it into two symmetrical bias circuits in series, Q14 and Q15 ... but ... you know the end, it still didn't work , so he
6) put un-muting contacts on every signal jack.
7) Besides, he put lots of diodes everywhere, either regular ones, Zeners, and those modern BATs (Schottky?), trying to compensate junctions here and there.
That's why I refrained to trying to polish this turd, but now I am intrigued.
Lowell, would you please cut the track that, on the schematic, goes from the lower leg of R95 to the left, separating it from the collectors of Q12 and Q13? (and C51/52).
Everything to the right is symmetrical and has no gain, and *should* have 0V offset if

the string of R91, 92, 94, 95, 96, 97,99, 100.

is symmetrical, as RG pointed.
You have effectively split the amp in two.
If the right side still has offset, at least it can be troubleshooted with a reasonable amount of doubts (or variables).
If it does not , then the left side is guilty, and, really, beyond reasonable repair.
In tehory, *Everything* "can" be repaired, but, as Enzo said, is it worth it?, (beyond certain reasonable effort); he also said that the job is basically to repair, not reengineer.
As to those FUs, by nomenclature they should be Fusistors, but they don't come in that many values or dissipations, *maybe* they mean "flameproof". Might be wrong, of course.
As a side note, I don't understand the use of words such as "adjast", "resistiye", "minmum", "contorols", "PDB", and "capcitors".
Maybe I'm nitpicking here, but find strange they were not corrected.
A *lot* of people must have read these schematics carefully.
PS: if it works properly, either remove the mute and call it a day, or, if worried about it, leave the amp muted anyway and use Q7, Q8, and RT2 to trip a Fet or Led/Ldr mute or just a relay which cuts the speaker off. Anything that does not mess with the DC balance of the amplifier.

[Enzo]

And leave us not forget that this amp is on the do not repair list at Fender. If it comes in under warranty, they will simply replace the amp, NOT repair it. That has nothing to do with poor design, it is due to low cost, but nonetheless, I won't hold my breath on troubleshooting hand-holding on an amp they wouldn;t bother repairing.

[R.G.]

Once you get started watching a train wreck, it's hard to look away. 8-)

[R.G.]

I think this does the mute without the silliness.
Remove Q7 and put a wire between the emitter and collector holes. This permanently undoes the silliness in the power amp; it's always unmuted.

Tie the drain of a J111 JFET to the junction of R65 and R66. Tie the source to the ground side of R66. Connect the anode of a diode to the gate lead of the new JFET and tie the cathode end to a 0.1uF ceramic capacitor to ground and a 1M to -16V.

If I'm reading the jacks correctly, "Mute" goes open circuit when a plug is inserted. So finally, tie a 10K from "Mute" to the junction of the diode, 0.1uF cap, and 1M.

Now when there is a plug inserted, Mute is opened and the 1M drags the new JFET gate to -16, turning the JFET off. When all plugs are removed, Mute ties the JFET gate diode to ground, letting the JFET turn on to Rdson, about 30-50 ohms.

If I got the sex of the mute wrong, you can do the same thing with R65 and a series JFET - If Ididn't make any mistakes as I wrote this down, not even sketching the circuit.

[lowell]

I'm not with the amp at the moment. But have a quick question. Is mute circuitry important? What's the point?

[R.G.]

I'm not with the amp at the moment. But have a quick question. Is mute circuitry important? What's the point?

It's a good idea - but it's usually done by grounding the input jack with a tip shunt contact to ground.

The fact that this amp has so much junk in there to do muting must mean that for some reason, muting at the jacks themselves was not enough. Could be the "designer" had one of those mail-order degrees, and having expended so much cleverness on the power amp, had to do something funny with muting to keep it from self destructing. Or was - as sometimes surfaces here - a legend in his own mind, and saw a different and more beautiful reality hidden... just... behind... this clot of parts, and had to keep pasting things in to make it work as the ship deadline approached. No good way to tell.

But Fender making it a DNR (Do Not Resusc...er, repair) amp is instructive.

[lowell]

Ok. It seems to me though that the mute just shuts off transistors when nothing external is plugged into either of the 3 related jacks. How does that protect the power amp from destruction?

[J M Fahey]

This mute has two visible functions (maybe more that I'm not aware of)
1) What mutes do in guitar amps, give an illusion of noiselessness.
To avoid unexperienced kids saying (I've read it many times in various forums) "this amp is sh*t!!, I put volume, treble, boost, more gain, ninja mode, presence, master, 10 band EQ, my flanger, ... and it hisses as an angry cat (or sounds like Niagara Falls) ... and with no guitar !!!!!!!!!!!!!). So the mute ... mutes the noise.
If he plugs his guitar unmuting the amp, the feedback scream he'll get *can* be traced to his guitar, so *he* mutes his mouth/keyboard.
Marshalls do it in a simple and elegant way, just use a free contact on their classic Cliff jacks to ground some point in the distortion channel (by definition the noisiest one), some 2 or 3 gain stages upstream. Works like a charm. Most everybody copied that.
This Fender designer went overboard, wanted *every* input jack to turn the mute off.
2) It has a more important function here: it mutes the amp if it detects that the heatsinks are hot, thanks to the thermistor.
The mute trigger circuit is not bad, what's troublesome is the system of muting, by starving the very important input differential pair.
SWR has a working version of it, in the LA12 and LA15 power amps, but that one physically cuts power to the differential pair, and to the "Class A gain transistor" and to the bias chain. To boot, it's simpler !!!
It's not very intuitive, but I think I traced the output transistors to be grounded through a couple resistors and diodes, and unbiased too, so they are "off" ... and with no offset.
Of course, we are talking Stewart Rabe here, lots of history on his back, since the last days of Acoustic.
Back to the mute, when it's (thermally) triggered, it makes the amplifier silent, letting it cool off some minutes.
RG's idea of using a FET can be combined to this "mute trigger" but just shunting input signal to ground, not messing with the DC balance.
A suggestion: just disable it and turn it into a "usable" amplifier, remember it's already dead (and declared "unrepairable")
Just as a curio: post a couple pictures of its guts, specially the power amplifier section and heatsinks.

[MarkusBass]

This mute has two visible functions (maybe more that I'm not aware of)

There is another visible function: avoid "thump" when the amp is turned on. Many companies do it in a similar way (by switching off the input stage of the power amp) and I wouldn't replace it with a simple JFET. Of course, if there is a DC on the speaker, it does not work as expected.
Also if you look at the schematic, you will notice that the input jacks in this amp are connected to the ground (so the preamps are muted on their own) when there is no guitar plugged into any of these jacks.

Mark

[R.G.]

There is another visible function: avoid "thump" when the amp is turned on.

]
That is a good and valid thing to use a mute for, in concept.

However, a well designed example of the diffamp-input, CCS-loaded VAS, follower-output power amps generates almost no thump at all if it's powered from a bipolar power supply. Thump was a serious issue where amps were powers from a unipolar power supply. Thankfully, this is now almost unheard of for solid state power amps.

But it is still possible to mal-design a power amp to thump. This could well be one of those, given the oddities in it anyway.

Many companies do it in a similar way (by switching off the input stage of the power amp) and I wouldn't replace it with a simple JFET.

I would, given (a) a simple JFET worked and did the job and (b) it needed de-thumping in the first place; see above.

Frankly, mute on power up is probably more needed where you power up with something already plugged into the input. Powering up into a working input can be bad if you happen to have an unlucky input voltage at that instant. But this circuit doesn't even do that. A plug inserted in the input jacks means unmuted, so you have all the potential bad-ities.

I suspect that this amp including the preamp circuitry has some problems with thumping on power up, but it's likely that they come from the preamp's signal switching, not from the power amp. Maybe they are covering over those with the mute. Can't tell.

Of course, if there is a DC on the speaker, it does not work as expected.

There *is* that too! Even if they needed it, they got it wrong.

Also if you look at the schematic, you will notice that the input jacks in this amp are connected to the ground (so the preamps are muted on their own) when there is no guitar plugged into any of these jacks.

Yep. Maybe the mute is intended to soften the blow when something is plugged in by the dynamic nature of it not coming on suddenly, but with some time constant. I don't know.

[lowell]

Lowell,

I'm not asking you whether Q11 is on. We know that it is on for at least a week when you specified the voltage on R87 = 20 mV. I'm asking you to check why it is on.
Again, few remarks:
1. When the amp is muted, the negative DC feedback is not working. So it cannot correct the offset.
2. Due to reason that you haven't found yet, Q11 is conducting.
3. Diodes D25-27 are conducting because Q11 is conducting.
4. Q13 (connected to -V rail) is conducting because diodes D25-27 are conducting.
5. You get negative offset on the output because Q13 is conducting and Q12 is not. You can tell this because there is 0V on R88. I assume that when you unmute the amp everything is OK. It means that most probably R88 and Q12 are also OK.

So the main question is still: why Q11 is conducting. But please don't tell me that it has 0.5V drop on Vbe. You have to find out why there is voltage drop on R84 and D24.
If you look at the schemaitic, you can see that Q11 and Q12 are connected in almost exactly the same way: a resistor from emitter to +V rail, a diode and a resistor from base to +V rail and their bases are connected to Q9, Q10, which both should be off when the amp is muted. So why they are working differently?
I suggest that you measure voltages on R85, R86 (I'm asking about this for the third time) and possibly replace Q10. If this does not help, I would replace Q11. If this also does not help, I'd say that this is by design and you should call Fender support and ask about it.

Mark

There is 0v across R85/86.
I replaced Q10/11 and now they are both officially OFF. No voltage across D25-27. Problem is, there's still DC offset on speaker jack, granted it's less than before at about -2.2.v now.

Q18 Vb: .94v
Q19 Vb: 3.35v
WOW

Q20 Vb: 1.6v
Q21 Vb: 2.7v
WOW again

JM I did not scratch that trace connecting the resistor string to the center of Q12/13 because now Q12/13 are off so I figure THEY are no longer playing a part in the offset.

Voltages across the resistor strings to the center of Q12/13 are imbalanced.

R91:33v
R92:10.4v
R94:670mv
R95:600mv
R96:600mv
R97:670mv
R99:9.18v
R100:28.5v

So it seems that from the top of D30 to the bottom of D31 all is balanced.

I disconnected D28/29 and it didn't change. I then replaced D34-37 with no change. My next thought was that C56/56 may be leaking so I disconnected them and nothing changed. I'm gonna call Fender support and see what they say. Aside from that I may just disconnected the mute ground connections so the amp is ON. The only issue here is that there is a 1-second-long hum at power-on with the amp unmuted so I'm not sure if that's going to be a destructive issue over time.

[lowell]

Called Fender and they said that they cannot provide this kind of tech support.

[J M Fahey]

The only issue here is that there is a 1-second-long hum at power-on with the amp unmuted

Well, NOW we know why they put that mute here, and were so obsessive about its switching.
Do me a favor, post a couple pictures, I want to see the "real estate" available inside that chassis, so as to put a new power board there and call it quits.
I'm also interested in the heatsinking they used.
Is it a long bar of solid aluminum, joining the transistors to the aluminum chassis? Or a "regular" finned heatsink?