Ok, I adjusted the bias until the P- side came up to 0.424V but now the P+ is 0.528. Also, the output offset is now +032mv
thanks
Hi Terry, I'm not sure there is a problem with the fets. You are able to vary the bias voltage ok and both ends of the pot change voltage. Another thing that can be done is to reduce the load on the collector of Q4. On the p chan. side lift one end of the gate drive limit network D3-ZD2 and see if the voltage changes. Also lift one end of the gate stoppers one at a time to see if the voltage changes.(R22-26). Make sure the filter caps are discharged before lifting any leads.
Steve.
Steve.
Just looked at the Soundcraftsmen and I was wrong. The MOSFETs are J50/k135. Not going to work. I did find some J52/K176 for sale but he wants $25 a pair with $28 shipping
Hi Terry, the gate current check is not a definitive one. It is just one of a number of tests that can be done before you have to pull the fets and test them on a jig. Don't forget to check the diodes D3, ZD2 for faults. It is difficult to troubleshoot using only a dmm when there are no major discrepancies in the voltage readings. There is no substitute for a dummy load, a signal source and a scope when it comes to fixing amps ime. As I have suggested previously, running it under load and watching the waveform symmetry is the easiest way of seeing if all the fets are functioning.
Steve.
Steve.
32mV offset is not a problem.
Try to listening to it in this state before pulling MOSFETs.
Ok, I'm still trying to understand where the voltage at RV1 is coming from.
1) Does it come from the collectors of Q4 and Q5? If so, could a fault in Q3, Q4 or Q5 cause the voltages to be off?
2) What is the function of D3-ZD1?
3) The Schematic calls for 0.6V across RV1. The good channel was set at 0.8V when I got it and it plays fine. Is this due to the higher rail voltage?
Thanks, Terry
1) Does it come from the collectors of Q4 and Q5? If so, could a fault in Q3, Q4 or Q5 cause the voltages to be off?
2) What is the function of D3-ZD1?
3) The Schematic calls for 0.6V across RV1. The good channel was set at 0.8V when I got it and it plays fine. Is this due to the higher rail voltage?
Thanks, Terry
1) Yes - the voltage comes from the current generated by Q4 and Q5, and the current through Q5 is created by Q3. So technically speaking - yes a faulty transistor in one of these positions would of course be a bad thing...as well as faulty transistors in Q1 and Q2 since they also influence the current through Q3 and Q4. However it is not likely that any of those are defective.
By looking at you measurements in post #7 both channels looks quite balanced and should work. Of course there is a difference of some 7-8Volt some places, but remember that the supply lines are at 117Volt each ... that is less than 10% meaning it is probably just component tolerances and age. Without an osciloscope and sinus generator to prove differently it is hard to think that any of these transistors should have gone.
Tolerances and age is probably also the reason why the voltage over VR1 has moved below the safe MOSFET-on-state base-source voltage-drop, starving the output stage and resulting in a true class B performance with lots of crossover-distortion. You should really just give the VR1 one notch more until the MOSFETs are conducting safely, and the sweet sound exits at the speaker terminal.
2) D3 - ZD1 are there to protect the bases of the MOSFETs for voltages over 10V between base and source, wich could destroy them.
3) The higher the voltage over VR1 - the higher the idle current through the MOSFETs. As I recall it (hard to find the datasheets) 0.7-1.0 volt seems right for these audio MOSFET pairs. Now that you can't measure the idle current then you can maybe feel how warm the heatsink gets in the two channels. I bet the good channel gets warmer - because it carries the 170mA more or less, whereas the bad channel will be as cold as the cabinet because no current is running through the output. If that is indeed the case, then that is another proof that you can safely adjust VR1 a little up.
Thanks for the great explanation! I have a much better grasp on it now. I ordered a sinewave generator and will get the scope out and see if I can remember how to hook it up. Turned up the bias to get the p- to .424. I was going to give it a listen but can' find my XLR adapters. After I do I'll report back to how it sounds.
Still, it seems suspicious that an amp that played beautifully for years just all of a sudden goes to distortion in one channel without some kind of component failure.
Thanks again, Terry
Still, it seems suspicious that an amp that played beautifully for years just all of a sudden goes to distortion in one channel without some kind of component failure.
Thanks again, Terry
Great.
Looking forward to read about your measurements.
By the way: The schematic shows some fuses in the supply lines. It is a bit strange that there are none in real life.
But for the final adjustments it would be nice to be able to measure the idle current somehow.
Would you be able to insert a 0.1 ohm/5W resistor in one of the supply lines?
Looking forward to read about your measurements.
By the way: The schematic shows some fuses in the supply lines. It is a bit strange that there are none in real life.
But for the final adjustments it would be nice to be able to measure the idle current somehow.
Would you be able to insert a 0.1 ohm/5W resistor in one of the supply lines?
Perreaux Mosfet Amps 3000,6000B,8000C etc
This is my first post and I joined to see the schematic from still4given Thank you for the link to the 8000C/9000 schematic...
The advise given here is very good.
Tips
Yes clean & dry long parallel tracks and around the Gate pins.
Replace C10 & C20 47pF caps, use High Voltage rating as these go leaky and cause problems to gate drive bias.
To check bias current (Bias Voltage to gate is similar to BJT amp) from the rail fuse I use Current DC on my Fluke 75 on 300mA range.
*Caution, when using a MM on Current it act like a short piece of wire!!!
Securely Hook this up across either of the rail fuse holders with the fuse still in, the switch it on you should see about 1mA on the meter, then remove the fuse (in parallel) and now see the full bias current. i.e. 150-250mA depending on number of output pairs and mains voltage variation. Adjust to spec!!
I hope this helps.
This is my first post and I joined to see the schematic from still4given Thank you for the link to the 8000C/9000 schematic...
The advise given here is very good.
Tips
Yes clean & dry long parallel tracks and around the Gate pins.
Replace C10 & C20 47pF caps, use High Voltage rating as these go leaky and cause problems to gate drive bias.
To check bias current (Bias Voltage to gate is similar to BJT amp) from the rail fuse I use Current DC on my Fluke 75 on 300mA range.
*Caution, when using a MM on Current it act like a short piece of wire!!!
Securely Hook this up across either of the rail fuse holders with the fuse still in, the switch it on you should see about 1mA on the meter, then remove the fuse (in parallel) and now see the full bias current. i.e. 150-250mA depending on number of output pairs and mains voltage variation. Adjust to spec!!
I hope this helps.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.