There you have it... The Mosfets must be isolated from each other and the heatsink.
edit: Keruskerfuerst was faster....
Last edited:
When I got this amp, the primary side of the power transformer was incorrectly wired and half the connector tabs on the PCB were broken from their solder pads. The bridge rectifier was missing and there was a massive amount to thermal compound all over the place. The lack of thermal insulating pads on the mosfets would have definitely guaranteed their failure when the previous owner replaced them but I still want to make sure there's not something else that may have caused their failure in the first place. Thanks to everyone for getting me this far. Now I need to get ahold of Mouser and add mica pads before they ship off my order.
OK, I'm back. I just went ahead and replaced all four power amp mosfet's(Q2,5,7,8), Q4 and Q1 as well as Q3. I also replaced Q6, Q9 and U5 for good luck. All four power amp mosfet's are now attached to the heatsink plate with a mica insulator with thermal paste on both sides. I hooked everything back together and when I turned it on the light bulb in the limiter seemed to be pretty bright so I immediately shut it off. My question is how long is it safe to let it run on the light bulb in order to take rail voltage and output DC voltage measurements?
Update-I got anxious so I decided to check the bias voltage across R72 and R74 by turning the amp on and off real quickly. I was getting around 50 MV when it should be between 15-20. So I turned the bias pot and the lightbulb dimmed. I then set the bias to around 16 MV and the lightbulb stayed somewhat dimmed. To me, this sounds like I have found my culprit but I just wanted to verify it with you guys before I call it good and plug it straight in minus the limiter.
After replacing power transistors you should always set the bias trimmer to the lowest level which should be CCW if the schematic is correct.
If you have a scope and an audio generator apply a signal to the input and check the output without load.
Check all DC values over R72, 73, 74 and 75 and ensure they are similar. They never will be exactly the same because of resistor and MOSFET tolerances, but should be within about 20...30 percent difference. I f you find one that differs significantly consider using a different transistor there. I would set bias according to the highest reading anyway.
What rail voltage do you get with the bulb installed? Maybe it's better to set bias to its lowest level before removing the bulb and repeat setting then.
If everything is fine I think you can give it a try.
If you have a scope and an audio generator apply a signal to the input and check the output without load.
Check all DC values over R72, 73, 74 and 75 and ensure they are similar. They never will be exactly the same because of resistor and MOSFET tolerances, but should be within about 20...30 percent difference. I f you find one that differs significantly consider using a different transistor there. I would set bias according to the highest reading anyway.
What rail voltage do you get with the bulb installed? Maybe it's better to set bias to its lowest level before removing the bulb and repeat setting then.
If everything is fine I think you can give it a try.
So my rail voltage is at +/-36V and with the bias set as low as possible, I get around 15MV across R73, 16MV across R72, R74 and 18MV across R75. This to me would at least indicate that the bias voltage can be set within the 15MV-20MV it needs to be but I will need to crank it down before I plug the amp straight into the wall since my rails are about 14V under what they should be. Is it bad to under bias mosfet's even if it's only for a short period of time? I do have a signal generator but only an o-scope plugin that runs on my DAW. What should I be looking for?
Underbiasing will do no harm. In extreme cases it only will introduce crossover distortion.
Overbiasing is dangerous as you know, the transistors may run too hot.
The difference between the DC values over the resistors is fine.
But I understand you can't go below 15mV even with the bias pot turned fully left? If so, you could make R43 a little higher (maybe 390R).
Overbiasing is dangerous as you know, the transistors may run too hot.
The difference between the DC values over the resistors is fine.
But I understand you can't go below 15mV even with the bias pot turned fully left? If so, you could make R43 a little higher (maybe 390R).
Last edited:
No, I can drop the bias voltage to 0 if I want. I was just saying that if I try to keep all my bias readings above 15MV will result in R75 reading 3 MV higher. So do you think it's safe to eliminate the limiter(dropping the bias prior to doing so and then setting it correctly after) without performing an O-scope check first?
I would test with a scope and the limiter installed to see if the amp is working basically without load and make sure there is no significant DC on the output. If everything looks fine, remove the limiter, turn bias to zero, power up again and readjust bias. Then you can test the amp with a dummy load or the speaker.
3mV difference is nothing to worry about. I have seen Ampegs with more.
If you do not have a scope you have to rely on the fact that you can obviously adjust the bias of the power stage, which is definitely a good sign.
3mV difference is nothing to worry about. I have seen Ampegs with more.
If you do not have a scope you have to rely on the fact that you can obviously adjust the bias of the power stage, which is definitely a good sign.
Last edited:
You're welcome!
Just use the multimeter to check DC. A few millivolts are ok, but if you find a value of a few hundred there could be a problem. If it gets into volts there is definitely something wrong.
A scope just gives you more insight about what's going on. In your example, when setting the bias on Ampeg power amps you can clearly learn that they are quite sensitive about bias setting - too low and you will be punished with crossover distortion that is clealy visible on the screen. I always cross-check the current set by the potentiometer with the result on the scope. This ensures that amp is running with bias as low as possible and as high as necessary. But the 16 mV (about 50 miliiampere per transistor) are certainly ok.
If you do not have a scope, you can use the multimeter to estimate whether the amplifier is actually amplifying. Again with no load, set the meter to AC Volts and measure the output. Apply a signal to the input. If you have nothing else at hand, plug a guitar cable into the input and touch the tip of the plug with your finger.
You should cleary see a reaction on the meter.
Anyway, as you obviously have come to a point now where there is no more shorting and the circuit is reacting on the bias pot, you are probably through
with it. Make the final settings and fire it up. Good luck!
Just use the multimeter to check DC. A few millivolts are ok, but if you find a value of a few hundred there could be a problem. If it gets into volts there is definitely something wrong.
A scope just gives you more insight about what's going on. In your example, when setting the bias on Ampeg power amps you can clearly learn that they are quite sensitive about bias setting - too low and you will be punished with crossover distortion that is clealy visible on the screen. I always cross-check the current set by the potentiometer with the result on the scope. This ensures that amp is running with bias as low as possible and as high as necessary. But the 16 mV (about 50 miliiampere per transistor) are certainly ok.
If you do not have a scope, you can use the multimeter to estimate whether the amplifier is actually amplifying. Again with no load, set the meter to AC Volts and measure the output. Apply a signal to the input. If you have nothing else at hand, plug a guitar cable into the input and touch the tip of the plug with your finger.
You should cleary see a reaction on the meter.
Anyway, as you obviously have come to a point now where there is no more shorting and the circuit is reacting on the bias pot, you are probably through
with it. Make the final settings and fire it up. Good luck!
very low dissipation capacity devices could be harmed by long term usage through the Mains Bulb Tester (MBT).
But in my experience, I have never damaged any parts in an incorrectly assembled amplifier when powered up via an MBT. Not even the fuse.
Here's why:
assume the nearly bright bulb filament is dropping 50% of the mains voltage.
for a 60W bulb that 120Vac is dropped across the filament and 120Vac is dropped across the 230Vac mains transformer.
The output from that transformer must be less than half the rated VA and approximately half the rated Vac.
the amplifier is thus being supplied with rails at ~ +-50% of rated voltage.
If the filament becomes bright, the voltage dropped across the mains transformer is only 5Vac to 10Vac. The output is < 5% of rated voltage.
Although the voltages will be lower in North America, the same protections apply.
IMHO, go ahead.
Bias setting should not affect any DC level on the output, so from that point everything is fine.
0,3V are not brilliant, but I do not think this is something to worry about.
Have you checked if the amp works as I suggested?
I know that powering up a repaired amp for the first time is a can be a moment of fear.
Especially without a scope. Check DC levels on pins 1&7 of the opamp, they should be near zero.
The power stage does not have any offset correction, and any offset created by the opamps will be amplified and appear on the output.
I really do not want to put you into trouble, but if this would be my amp I would say that everything I could think of has been done and just give it a try.
I'll keep my fingers crossed.
Good luck!
Bias setting should not affect any DC level on the output, so from that point everything is fine.
0,3V are not brilliant, but I do not think this is something to worry about.
Have you checked if the amp works as I suggested?
I know that powering up a repaired amp for the first time is a can be a moment of fear.
Especially without a scope. Check DC levels on pins 1&7 of the opamp, they should be near zero.The power stage does not have any offset correction, and any offset created by the opamps will be amplified and appear on the output.
I really do not want to put you into trouble, but if this would be my amp I would say that everything I could think of has been done and just give it a try.
I'll keep my fingers crossed.
Good luck!
Last edited:
I did the AC test with no load and the amp was amplifying signal when I touched the end of the guitar cable so then I backed down the bias pot and plugged it straight in, bypassing the limiter. It fired up and the fuse did not blow. It did make a pretty loud energizing(hard to describe) sound when it turned on which kind of scared me but otherwise it operated fine. I set the bias so that my lowest reading was 15 MV which made my highest 19 MV. I did notice that the +/- rails were reading 72V when the schematic states that they should be 50V. Should I be concerned about that?
Yes.
72V??
That's strange. Is the transformer wiring correct? How many wires does your transformer actually have on the primary (mains) side?
Disconnect the secondary wires and measure AC voltages between the 3 leads.
What color do they have?
Take care, the amp and the power supply probably won't stand this for a longer time.
Last edited:
To be more precise, the components that are most in danger are the power resistors and Z-diodes (R66/68, D11/12) of the low voltage supply due to current overload. The IRFP(9)140 mosfets are rated for a maximum Ugs of only 100V. Pushing the amp may kill them.
If you have an export transformer with 7 primary leads fitted check their connections to the terminals on the board. What is your mains voltage?
Could it be somebody put in a wrong transformer? The originals are rather *small* standard (non-toroidal) types with usually red tape covering the windings if I remember correctly.
Maybe someone read "50V" in the schematic and thought he needed a 50V transformer for that, which is about the AC voltage the transformer delivers at the moment.
To get the +/-50V supply the schematic reads you can expect about 2x35VAC on the output of the transformer.
If you have an export transformer with 7 primary leads fitted check their connections to the terminals on the board. What is your mains voltage?
Could it be somebody put in a wrong transformer? The originals are rather *small* standard (non-toroidal) types with usually red tape covering the windings if I remember correctly.
Maybe someone read "50V" in the schematic and thought he needed a 50V transformer for that, which is about the AC voltage the transformer delivers at the moment.
To get the +/-50V supply the schematic reads you can expect about 2x35VAC on the output of the transformer.
I measured 53 VAC on each half of the secondary. AC mains are 122V. There is only one method shown on the schematic for wiring this thing for 120 Vac and I double checked and it is wired as it should be. I was under the assumption that these amps are all made with the same transformer and you just have to wire it according to where it’s being used in the world. This transformer is pretty big and has yellow tape wrapped around the windings and the sticker on the top of it reads “SLME 94-015-40 Federal 0834”. All the color coding in the wires matches the schematic and I found a forum online where someone else was trying to get info on this same model transformer for the exact same amp, so It should be the original.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.