According to the sim, increasing this resistor dramatically increases the quiescent current in the drivers and outputs. The drivers are already cruising along at 25mA with the oem part. I can check in real life though.
So, if I replace the 2.43k resistor with a 10k trim pot with the wiper tied to one end (so that the maximum value would be 5k//5k=2.5k), then I think that would work pretty good.
It should either be mounted directly on one of the output devices or on the heatsink, not next to it. Bad thermal contact there is the most likely cause of the thermal runaway problem you have. Maybe the person who worked on the amp before didn't put it back where it belongs?
It would be safer to replace R254 with a trimpot, so that if/when the trimpot fails open-circuit, the output devices will be switched off. If R250 (the 2.43K resistor) is replaced with a trimpot and the trimpot fails, then the output devices will burn fast.
As has been discovered, the resistors are indeed 1% at the stated value, part number for them in the part list and everything.
I don;t consider these Fender amps poorly made, but they are mass produced. No one in the plant would be trimming resistances.
"It looks a pretty decent design, but having a fixed bias requires EXACT transistors fitting if you have to replace them - your problem 'may' just be that you didn't get the correct transistors direct from Fender?."
No, the transistors are not selected or matched. You'd have to have exact transistors if the bias circuit was designed to be right on the edge. But this is a guitar amp. I'd be more inclined to view this as when it works properly, the bias currents are well within what any stock outputs would draw.
I'd agree with the suggestion of making R254 variable if you are going to install a trimmer. Then if the pot opened, the transistor would turn on hard. If you have the pot open in R250 position, the bias transistor turns OFF. Not good.
Q16 is nestled snug in a small notch in the heat sink right next to Q30.
I don;t consider these Fender amps poorly made, but they are mass produced. No one in the plant would be trimming resistances.
"It looks a pretty decent design, but having a fixed bias requires EXACT transistors fitting if you have to replace them - your problem 'may' just be that you didn't get the correct transistors direct from Fender?."
No, the transistors are not selected or matched. You'd have to have exact transistors if the bias circuit was designed to be right on the edge. But this is a guitar amp. I'd be more inclined to view this as when it works properly, the bias currents are well within what any stock outputs would draw.
I'd agree with the suggestion of making R254 variable if you are going to install a trimmer. Then if the pot opened, the transistor would turn on hard. If you have the pot open in R250 position, the bias transistor turns OFF. Not good.
Q16 is nestled snug in a small notch in the heat sink right next to Q30.
Yeah, Q16 is in case type ATV, mounted on the board in a notch in the bar stock that serves as a heat sink, but it is not mounted on it, nor bonded to it in any way. That's the way it is from the factory. I suppose it's just proximity to the heat from the bar stock that is supposed to provide some compensation. With some customization, it's possible to use a TO-220 transistor there instead, drill and tap a hole in the bar stock and mount it directly on the heat sink that way. I'm not sure I want to do that, but it's possible.
I will replace R254 with a trimmer then. Thank you very much!
I talked with the owner tonight and he's all in favor of modifications. I'm intending to extend the bass and treble response of this amplifier, for example. As it stands, it appears to be -3dB at 80Hz, and rolls off right above 10kHz. I know that low E is about 80Hz, so that should be reproduced clearly, at least, but even more so, if someone uses an electrified acoustic guitar on it, and does some drumming and slapping on the guitar, it would be great if this amp reproduced that cleanly as well. I have a CD of a couple playing electrified Spanish guitars and the woman does a lot of slapping and drumming on hers, with the result that my subwoofer is very busy when I play their music!
I will replace R254 with a trimmer then. Thank you very much!
I talked with the owner tonight and he's all in favor of modifications. I'm intending to extend the bass and treble response of this amplifier, for example. As it stands, it appears to be -3dB at 80Hz, and rolls off right above 10kHz. I know that low E is about 80Hz, so that should be reproduced clearly, at least, but even more so, if someone uses an electrified acoustic guitar on it, and does some drumming and slapping on the guitar, it would be great if this amp reproduced that cleanly as well. I have a CD of a couple playing electrified Spanish guitars and the woman does a lot of slapping and drumming on hers, with the result that my subwoofer is very busy when I play their music!
Last edited:
If Q16 doesn;t seem bonded enough to the heat sink, blob some heat grease around it and fill the gap.
Well, go ahead and experiment, but before expecting a lot more bottom end, as far as I know, this amp has an 8" woofer in the front, and a 6" one in the bottom, and that bottom baffle is sitting out in the open.
You could probably clean up the sparkle by replacing the piezo tweet with a crossover and something nicer.
Well, go ahead and experiment, but before expecting a lot more bottom end, as far as I know, this amp has an 8" woofer in the front, and a 6" one in the bottom, and that bottom baffle is sitting out in the open.
You could probably clean up the sparkle by replacing the piezo tweet with a crossover and something nicer.
It will increase dramatically, that's why it's best to make it adjustable.
NO!!!! - the maximum value would be 10K - FAR too high.
If you're replacing R250, you should wire the slider to one end (so if the slider goes O/C the resistance is limited to the value of the preset). a 2.7K pot would be fine.
However, as already suggested, you're better off replacing the bottom resistor for the reasons stated.
So, I looked at this a little more. I measured about 1.2V across R248, 47ohms which is about right according to the sim. What is way off is the emitter resistor voltage drops, which are in the range of 20-30mV, meaning that the outputs are conducting up to 200mA each. At first turn on, the values across the resistors went up beyond 30mV, and then settled below 20mV, and then slowly crept up to above 20mV to 30mV. The other amp was showing lower voltages of between 8 and 12mV. The voltage across the Vbe multiplier Q16 was about 2.2V, but it varied slightly over time. The heatsink gets hot. Too hot to keep my hand on it. I'm a bit stumped.
I went around the board and checked nearly every diode and resistor and they are all fine. ,<scratches head>
As we've repeatedly told you, adjust the bias - it's hardly something to be 'stumped' over.
Passively checking components is rarely a workable method of servicing, and neither would be using a simulator 😀
Hi,
when I was an EEE student I did my placement servicing and repairing
gas lasers and their power supplies. On a slack day I asked about a
power supply stuck in a corner. "Oh that one, it been their years,
nobody can fix it and everyone who's worked here has had a go.".
Needless to say, I fixed it and found out why nobody else could.
They assumed if all the parts are working it would work, and they
were looking for something broken that caused the problem.
That wasn't the case, it had been "modified", by the customer.
I found this out by working out what is going on. Your "head scratching"
is caused by simply not working out what is going on by applying some
basic circuit measurement and circuit theory.
What resistance do each of the 4 emitter read across them ?
What resistance does each output transistors emitter leg to ground read ?
What voltage reading to you get for the above 8 cases with it turned on ?
Split your measurements into the NPN and PNP pairs and post them.
It should be simple to describe exactly what is happening.
rgds, sreten.
Last edited:
I already posted most of that information:
"What is way off is the emitter resistor voltage drops, which are in the range of 20-30mV, meaning that the outputs are conducting up to 200mA each. At first turn on, the values across the resistors went up beyond 30mV, and then settled below 20mV, and then slowly crept up to above 20mV to 30mV. The other amp was showing lower voltages of between 8 and 12mV. "
Tell me how long after I turn it on that you want me to read the voltage across each emitter resistor. I told you it changes with time. The values for the resistors are within specification, as are the base blocking resistors and every other resistor on the board.
The output trannies have been replaced ..OK
Maybe the driving trannies have been stressed and are not now giving there expected Vbe drops , ie there Vbes are just that little bit lower than they should be .
Maybe the driving trannies have been stressed and are not now giving there expected Vbe drops , ie there Vbes are just that little bit lower than they should be .
I have not replaced Q16, and, based on the soldering, it looks original. It's an ATV case and fits into a notch in the bar stock that serves as a heat sink. It's located right next to one of the output transistors. I should take some photos and post them. I have some Arctic Silver I could squish in there, since Q16 is not thermally bonded to the heat sink.
I suppose the changes in the various voltage readings over time is just a sign of thermal runaway due to the bias being too high. But as I recall, the voltage across R248, the resistor between the driver transistors Q26 and Q28 was very constant with time, but on the other hand I guess it wouldn't take too much variance to radically alter the current in the outputs.
I don't have a pyrometer to check, but based on how long I could put my finger on the transistors, I seem to think that one of the drivers got hotter than the other. So, there could be something wrong with that one, or since these are not matched devices, they could be very far off on their respective Hfe values. I could buy some kind of complementary replacements and match them for Hfe. Do you think that would be a good idea? All of the outputs and drivers are obsolete components. Also, the voltage drop across the resistor between the drivers is very close to the predicted value in the sim. I don't know how to do the math to figure out what it should be otherwise. I let the software do the math for me.
Last edited:
I would compare the Vbes to the good amp .
Ideally the Hfes would be matched but i doubt that happens very often.
Ideally the Hfes would be matched but i doubt that happens very often.
Thanks. That's a good idea. I have not checked that very closely. I did compare the emitter resistor voltage drops in the finals for each amp though, as I noted. I should compare other voltages also I think. I'll do that when I have the time.
- Status
- Not open for further replies.