The current mirror stage, would you be using mje350 and mje340..
Many may years back I used mje340 bias servo track temp difft in the output stage as the stand a loan trim pot needed to much triming.
Many may years back I used mje340 bias servo track temp difft in the output stage as the stand a loan trim pot needed to much triming.
Last edited:
one pair of outputs will drive an 8ohms load without any damage, but one will get better performance using 2pair into 8ohms.
If you're using 4ohms, then I suggest 3pair.
Yes, it's jsut that I have only one pair per channel... a redesign for me means using the existing devices with mounting bracket and heatsink. And possibly the PSU too (2x30V AC transformer)
Regarding the psu design..a nice saftey mod is to add two 6k8 resistors say 3 watt across the filters caps, 1 from the + rail to centre ground and 1 from -- rail to centre ground and add 100nf cap across each resistor this will act as bleed resistors ie aid descharging the caps when the amplifier is powered down.
I've never quite understood the point of that, the bias current would do the same
If I recall the dave Goodman module used 35-0-35 @200va giving plus and minus 55vdc per rail.
Using 30-0-30 would be about 45 +/- dc.
For a twin channel amp aim for 300va or more.
Using 30-0-30 would be about 45 +/- dc.
For a twin channel amp aim for 300va or more.
Most high end pa amplifiers use bleed resistors, take peavey they have done this also.
Look at it this way, the wondering hand would get a big zap if the psu was not designed to bleed away high voltages hinding inside equipment!
A crude way to decharge caps would be to short it with a screw driver mmm big spark..so you could use a quick bulb of 40w to 100w in a holder and some fly leads much safter all round.
The old mosfet amplifiers would decharge slow ie..with the speaker still connected plus music still playing, if the amp was turned off, you'd hear the sound slowly fade away.
Last edited:
Wow, been away a few days and this thread has really taken off, Interesting to see all the different Ideas, to give the original design a make over.
So now I'm caught between diving in and building the Pass F5, or wait and see what crops with this re-design wise, with the wrinkles of the original design ironed out, and suggested improvements incorporated.
Also nice to see in (Jamesfeline's) build thread that someone noted that these amps would give the Naim and Quad 405 (I Think) a run for their money.
I can vouch that under the right conditions these amps can sound very good indeed.
I need to go back and re-read the thread where the Bias is set to 100ma, as it seemed to be the sweet spot, but as my amps use the 2SB48 and 2SK133, I'm not sure that would be a good idea, would need to do some tests to see how hot they run.
At the moment, at a normal listening level, and the Bias set to 50ma, they don't even break a sweat.
So now I'm caught between diving in and building the Pass F5, or wait and see what crops with this re-design wise, with the wrinkles of the original design ironed out, and suggested improvements incorporated.
Also nice to see in (Jamesfeline's) build thread that someone noted that these amps would give the Naim and Quad 405 (I Think) a run for their money.
I can vouch that under the right conditions these amps can sound very good indeed.
I need to go back and re-read the thread where the Bias is set to 100ma, as it seemed to be the sweet spot, but as my amps use the 2SB48 and 2SK133, I'm not sure that would be a good idea, would need to do some tests to see how hot they run.
At the moment, at a normal listening level, and the Bias set to 50ma, they don't even break a sweat.
Last edited:
This circuit improves on the Maplin mosfet amp.
1/ Constant current source on front end.
2/ Proper 2 transistor current mirror.
3/. Front end decoupled to eliminate hum.
I have built this and tested it and it was very good.
1/ Constant current source on front end.
2/ Proper 2 transistor current mirror.
3/. Front end decoupled to eliminate hum.
I have built this and tested it and it was very good.
An externally hosted image should be here but it was not working when we last tested it.
I built a 300w per channel bipola amp using current dump at the tail pair plus vas current mirror followed v limiter in case the o/p was shorted. This pulls the drive voltage to deck and the output don't over heat!