My MOSFET amplifier designed for music.

[Mooly]

Hi,
Depends what you hang on the end of it really, but yes I never had any problems during it's development driving "normal" speakers as a load. The series 0.22 usually provides all the isolation needed and this should not be ommited.
It was really to help Space get it all up and running, and as he likes experimenting to see if he could tell any difference.
Technically it should be there. 6 microhenry there or there abouts, air spaced.

Did not quite know what to say to Lumba Ogir.
Perhaps the "enlightened one" could indeed post a design that would lift us to another state.
Regards Karl

[jerluwoo]

Mooly, I believe the issue he is refering to with the drivers is the very large currents they will have to provide to the gates at higher frequencies. This can be a matter of amperes. If your driver transistors can not supply this current it will cause considerable distortion in the high frequencies.

[Mooly]

Hello Jerluwoo,
Gate capacitance is one of the "issues" with FET's it's true (The lateral FET's are quite good in this respect ---600 pf and 900 pf). You have to put it in perspective though. When you talk of "Amperes" of gate current that implies swinging the gate voltage at extremely high slew rates. What would be a sensible value ?
Although nothing to do with the amp design I have attached a shot of a 1 k h z squarewave as reproduced from CD. That kind of puts it into perspective really. Any player will reproduce this the same. Many many years ago Nelson Pass did some tests with slew rates off vinyl which is a system without the absolute bandwidth limitations of CD. I think I am correct in saying the figure was around 2 Volts / microsecond at 100 watts. (Hope I have got that figure correct Nelson )
Personally I am not a great fan of super extended bandwidths for audio amps, much prefering to see a nicely tailored and controlled response with no "nasties" on squarewave testing.
Regards Karl

[nigelwright7557]

Originally posted by Mooly
Hello Jerluwoo,
Gate capacitance is one of the "issues" with FET's it's true (The lateral FET's are quite good in this respect ---600 pf and 900 pf). You have to put it in perspective though. When you talk of "Amperes" of gate current that implies swinging the gate voltage at extremely high slew rates. What would be a sensible value ?
Although nothing to do with the amp design I have attached a shot of a 1 k h z squarewave as reproduced from CD. That kind of puts it into perspective really. Any player will reproduce this the same. Many many years ago Nelson Pass did some tests with slew rates off vinyl which is a system without the absolute bandwidth limitations of CD. I think I am correct in saying the figure was around 2 Volts / microsecond at 100 watts. (Hope I have got that figure correct Nelson )
Personally I am not a great fan of super extended bandwidths for audio amps, much prefering to see a nicely tailored and controlled response with no "nasties" on squarewave testing.
Regards Karl

In my designs I always filter out frequencies not required.
This helps stop oscilation and RF demodulation !

One of my first Maplin amps picked up the local police radio a treat !

[Mooly]

You know the old saying, "Amplifiers often oscillate, oscillators often fail to do so"

[AndrewT]

Originally posted by Mooly
.......I have attached a shot of a 1kHz squarewave as reproduced from CD.

and they can't get close to a 20kHz squarewave.

The filters on the front end of the amp, both low pass and high pass, should be adjusted to pass the wanted signal and attenuate the unwanted signal.
As for slew rate, 20 to 30V/uS is sufficient for most listeners.

[Mooly]

Hi Andrew,
Got the camera out, and took some real screen shots.
NOTE, The top trace is the input to the power amp but this signal has been through the preamp first ( it is an integrated amp after all ), where the preamp also has a "tailored response". The amp was delivering around 1.6 watts R.M.S. into 5 ohms.
I would suggest that anyone viewing these looks back at post #153 for a 1khz signal from CD which if that is what you use as a source is as good as it gets.
This shows the response at 1 Khz

[Mooly]

This is the rising part of the signal expanded --- no "nasties" whatsoever.

[Mooly]

This is at 10 Khz

[Mooly]

And at 20 Khz