I'm reluctant to take the output transistors out as I'm not confident I'll be able to do this neatly and don't want to lift any traces.
Both Pots P1&P2 measure 0 ohm's
If I take out the ZTX's, should I take out resistors R15, R16...,R20, for testing purposes?
I'm planning to rig up a 15watt test bulb tomorrow (I'm using a 100watt now) and take some voltage measurements at several point to ground and report back.
I'm also going to dail in my left channel board. If this goes well, I know I didn't make a systematic error.
Thanks everybody for all the input sofar.
do not try to set the bias of the outputs while testing via the bulb tester.
Keep the P1 & P2 pots at minimum using the 15W bulb.
Check the supply voltage is near what you expect.
If this is too low you may have to use a higher wattage bulb, try 40W.
Check all the voltages CAREFULLY. A slipped probe can damage semiconductors instantly. Check Vdrops across resistors. These tell you operating currents.
When you are sure that all the wiring is correct, then power up direct off the mains with the correct mains fuse fitted.
Start to bias up the output stage. Take it in small steps reading the source resistor Vdrops and reading the output offset.
If you can clip on three voltmeters simultaneously then better.
Get your head and hands to "remember" what direction the pots move in to increase the output device current. AND mark the PCB with a guide for next time.
Keep the P1 & P2 pots at minimum using the 15W bulb.
Check the supply voltage is near what you expect.
If this is too low you may have to use a higher wattage bulb, try 40W.
Check all the voltages CAREFULLY. A slipped probe can damage semiconductors instantly. Check Vdrops across resistors. These tell you operating currents.
When you are sure that all the wiring is correct, then power up direct off the mains with the correct mains fuse fitted.
Start to bias up the output stage. Take it in small steps reading the source resistor Vdrops and reading the output offset.
If you can clip on three voltmeters simultaneously then better.
Get your head and hands to "remember" what direction the pots move in to increase the output device current. AND mark the PCB with a guide for next time.
I made the mistake of having the pots at the halfway point during testing, this set the bias unwittingly to about 3.6A.
All I lost were the output MosFETs and their source resistors.
With the pots at minimum you should get 0V at the output as both MosFETs should be OFF, if you don't you might have lost one of the four components mentioned above.
The magic smoke out of mine was one of the source resistors.
All I lost were the output MosFETs and their source resistors.
With the pots at minimum you should get 0V at the output as both MosFETs should be OFF, if you don't you might have lost one of the four components mentioned above.
The magic smoke out of mine was one of the source resistors.
Good news. Today I tested just the left channel board it worked as expected and biased up nicely, which confirmed I hadn’t made a systematic error and my limited understanding was at least sound. I then took the right channel board out, checked the backside which I hadn’t seen with the cabling installed, and re-installed it. I then started testing it with a 15watt lightbulb inline with the mains cable and low and behold it seemed to be working fine !?
I still don’t know what was the root cause of my earlier issues, but they aren’t there no more. Pffuu..
Both board have now warmed and biased up and settled down and it’s been playing for a few hours.
The healthstats are :
Thanks for all the feedback. I couldn’t have build this amp without some of your encouragement and feedback.
Next project, the complementing DAC.
I still don’t know what was the root cause of my earlier issues, but they aren’t there no more. Pffuu..
Both board have now warmed and biased up and settled down and it’s been playing for a few hours.
The healthstats are :
- 0.57Vdc (slightly conservative) across the R7&R8
- Voltage across speaker terminal is hovering around 0.001 Vdc
- Heatsinks are a constant 40Celcius, and power transistors are similair so heatsinking is good.
- Mains draw is 125Watt (very respectable)
- It’s very dark and silent, smooth and easy sounding with a nice full body, extended highs and lows, very relaxed.
Thanks for all the feedback. I couldn’t have build this amp without some of your encouragement and feedback.
Next project, the complementing DAC.
Last edited:
Congratulations!
That is also how I fix most things - take it apart and put them back together. I have never really understood why it helps...
Thanks!
Law of attention?
I just got back to working on my F5 and somehow I hooked up one of my boards to the power supply backwards. I didn't power it up the full way on the variac but now it is passing DC on the outputs. I am guessing (praying) that I fried a jfet or two and not my outputs. It is possible to get DC on the outputs due to bad jfets right? If I did fry the jfets, exactly how important is it match the replacements?
I know it's unlikely, but nobody could offer me another possible cause for my problem
The amp lives. It was a burned out jfet. Luckily I had another matched pair around from a different F5 board I "practiced" on the first time I tried to build the amp.
I'm curious, has anyone ever tried using a microcontroler to control bias? You could use PID to keep bias perfectly at what ever you want to set it at. Would there be any real world benefit to such a system?
I'm curious, has anyone ever tried using a microcontroler to control bias? You could use PID to keep bias perfectly at what ever you want to set it at. Would there be any real world benefit to such a system?
I did some work in this direction. Look for a thread called F5 Micro. I never finished the project. My goal was mainly to have a way to rump up the bias at various level in very short time to compare the difference.
D.
