Fender deluxe 112 Plus. Really hot but sounding fine?

[allante666]

Hi. Anyone had a 112 in and noticed how hot it ran? Had one in with really low volume. Quickly saw that it was because someone had had the jack sockets for the power amp in and the phones out to pieces and reassembled them wrongly. (they are the big ones with a clear top). One had a contact trapped against the earth ring. I took both of them out and fitted new ones. Huge amount of volume but I believe this in normal for these? Like 80 percent of the volume in the first 15 percent of the volume knob!
anyway, played through it and both clean and dirty channels sound good, no distortion or hum so left it switched on to make sure it was ok. Half an hour later it still plays and sounds fine but I could fry my bacon on it! The heatsink is heating up th e aluminium casing and if I feel under the bottom where the transistors are bolted to it I cant bear my fingers on it. Just seems way too hot to be healthy. the thermal trip isnt cutting in but Im not keen on sending it out like this. I took it out and checked the TIP trannies and all 4 measure ok. Their bias resistors (R70-73) measure ok. The TIPS may have been replaced previously or not, soldering looks pretty original but theres big thumbprints on th eheatsink which usually means someones got messy with th e grease!
Any ideas anyone or are they just like this normally?

[MarkusBass]

In such a case you should rather check the current through the output transistors. Is it possible that someone replaced any of C5-C8 diodes and the quiescent current is set to high? I would check voltage on emiter resistors and calculate the current (and the power dissipated on output transistors).

Mark

[allante666]

Hi thanks for the reply. All 4 diodes are original and measure ok (not oc or sc). Voltage between collector and either side of emitter resistors is around 48volts assuming thats where you meant?

[MarkusBass]

Emitter resistors are in emitters of output transistors (they are also "balast resistors"). Each resistor is between emiter of output transistor and output of the amp. Definitely, they are not connected to collector of the output transistors. And the voltage on them should be in a range of milivolts. So for sure it is not 48V. Knowing the voltage and the resistance (0R47) you can calculate the current. This will be the same current that flows through the output transistors.

Marek

[allante666]

Sorry I didnt explain that very well. I thought you meant the voltage across the transistor.Yes, talking about the R47 resistors R74 75 103 105 in the emitter, the voltage measured across them is 0.013 volts which doesnt seem excessive?

[MarkusBass]

I told you to calculate the current. With voltage you cannot tell whether it is excessive, or not. It all depends on the resistance. In this case you have 13mV/0.47=27.7mA. This is per transistor. This is still not much but sometimes I've seen amps with lower quiescent current. Does it stay the same after half hour?

Mark

[allante666]

Its still as hot after half an hour, it had probably been on half an hour when I measured it. i did try shorting C5 to bias it cooler and it did work. However when I did that there was a slight amount of distortion on the clean channel, too much to let it go.
I have noticed that if i play a note through it on clean and switch it off while the note is still playing, in the half a second after switching it off the volume and distortion increases before stopping. dont know if that means anything?

[Jazz P Bass]

Are you able to measure the mains current that the amp is pulling from the wall at idle?

That will tell a lot.

[allante666]

Yes 360mA at idle

[52 Bill]

I have noticed that if i play a note through it on clean and switch it off while the note is still playing, in the half a second after switching it off the volume and distortion increases before stopping. dont know if that means anything?

I think that you might be hearing the sound of the distortion channel kicking in as the power supplies die off. Turn down the distortion controls and see if it goes quiet when you power off.

These normally do not get that hot when just idling. Is the heat sink assembly bar correctly installed and greased? You said that somebody was in there before and if the metal spacer bar is not installed correctly the heat will not transfer away nor to the temp sensing diodes.

Do you have a ultrasonic oscillation happening? Check the small value feedback caps in the preamp stages. Sometimes I find them broken off at one end.

Even though the bias diode string is original, one could be off enough to raise the bias too hot.
Diode test the bias diodes and read the conduction voltages, any higher than normal?

[MarkusBass]

It's still as hot after half an hour, it had probably been on half an hour when I measured it.

I think that in the first post you said that the amp gets hot after half an hour (so at first it is cold). Now you say that it is hot all the time. I suggest that you measure the voltage on ballast resistor when the amp is switched on, and then after 10 minutes and later after 30 minutes. We want to know whether the current increases over time, or it stays the same.

I did try shorting C5 to bias it cooler and it did work. However when I did that there was a slight amount of distortion on the clean channel, too much to let it go.

The problem with this amp is that it has fixed bias and it cannot be easily altered. The quiescent current greatly depends on voltage drop of the bias diodes. There is still a chance that some component in the amp has failed but it seems to me that voltage drop on bias diodes is to high (due to e.g. tolerance of the diodes). Is voltage on each diode the same?
Instead of shorting C5 you can put a trimmer across it (e.g. 10k) and check what happens when you decrease its resistance. You can check what change in current is caused by adding the trimmer.

Mark

[J M Fahey]

Yes, I'm not sure now whether it was fender or Peavey amps which had a similar "fixed bias" diode string and one of them had , say, a 100 ohms resistor or similar low value in parallel with *one* diode.
You might add another resistor in parallel with another, as an extra "trimming down" biasing.

[Enzo]

PV certainly "tuned" a diode with a parallel resistor in some models, but I can't say Fender didn't as well.

[J M Fahey]

Yes, very probably both did

Fact is, in an amp having fully complementary Darlington outputs (say, TIP142/147) you have 4 BE diodes to bias forward, and in theory a 4 diode string "should" be what the Doctor ordered ... problem is that a regular PSU diode (think 1N400x) is not exactly the same as the diode created when manufacturing a transistor BE junction, don't know why but plain diode voltage is *slightly* too high, so the standard solution is to use "1 diode less than needed" and replace the missing one with a trimmer resistor which can drop 700mV or less, check clasic schematics for that, or, when high volume must be automated and they don't want to rely on a Human measuring and adjusting abilities, some fixed trimming is used, hopefully it will work for the full production run.
Which it does, as soon as all parts come from the same batch.

Somebody might have replaced perfectly working diodes by new, perfectly working ones but slightly different spec ones and created a problem where there was no one before.
That's why in general we avoid and discourage shotgunning.

[mozwell]

28mA looks like a perfectly reasonable figure
28mA through each transistor with a 40V rail is 1.1W
Unlikely that 4.4W total idle dissipation will get the heatsink up to frying bacon temperatures at idle
I am not familiar with the amp, does this bracket with the power transistors on it bolt to the chassis, if so, the chassis will act as an extra heatsink, where just the pcb mounted bracket alone may be insufficient
If the heatsink is way too hot at idle, it will just blow up when we have music played through it, so we have to get the heatsink temperature rise at idle under control
I would guess a 10 degrees C rise (or less) at idle will do

We could try a 47 ohm 1/4W resistor in parallel with one of CR5, CR6, CR7, CR8 to reduce the bias slightly

If someone has been into this amp before, double check Q1 & Q10 are TIP142, and that Q2 & Q11 are TIP147 (maybe someone changed them & got them wrong)

Disconnect the speaker, do the power amp transistors still get hot at idle
If you have a different speaker, try that, do the power amp transistors still get hot at idle

What is VDC and VAC on the power amp output at idle with & without the speaker connected
Double check the voltage drop across each of R74, 75, 103, 105, check milli-volts DC and AC

Please check VAC at VDC on each of the power supply rails with the amp at idle. +/-15V rails, +/-40V rails

With power off & power supplies discharged, do a quick check of ALL the diodes in the amp, make sure these are all ok

With power off, measure between each terminal of the thermal switch TS1 to the heatsink, both these should measure open circuit

There is no Zobel network on this amp
Maybe a 10 ohm 5W and a 0.1uF 100V in series from the output of the amp to the "top" of R76 (or just across CP1 & CP2) will help here. The amp could be "marginally stable" & oscillating at idle (check out R61 & C39 on the fender champ schematic attached)