IRF540/9540 will give you about 1-1.5V more output voltage on account of slightly better transconductance. That being said, IRFP240/9240 will be better for cooling which is something you need to think about. On a nominal 4 ohm load from 40V rails, with about 6V loss, the peak power dissipation will be quite high. IRFP devices have a thermal resistance of about 1.1K/W while the IRFs have 1.5-1.6, AND have a lower rated dissipation, which certainly makes things far worse compared to the IRFPs.
One other thing: the 9xxx part will be the problem one, as it's current rating is lower and transconductance also, which makes for a higher loss on the P side, and more heat. Not a good combination.
To make matters worse, you will be getting asymetric transconductance in the output stage, which will not do good for distortion, especially into 4 ohms.
You might want to plug a IRF9540/IRF640 pair into your simulator and see the distortion fall
Also, IRFP parts were engineered for better analog performance and will offer better linearity. You may want to look into that...
One other thing: the 9xxx part will be the problem one, as it's current rating is lower and transconductance also, which makes for a higher loss on the P side, and more heat. Not a good combination.
To make matters worse, you will be getting asymetric transconductance in the output stage, which will not do good for distortion, especially into 4 ohms.
You might want to plug a IRF9540/IRF640 pair into your simulator and see the distortion fall
Also, IRFP parts were engineered for better analog performance and will offer better linearity. You may want to look into that...
OK, let me give you a bit longer answer then...
If I was making an amp with your conditions (+-40V rails, 4 ohm load), I would use IRFP240 and IRFP9140. This are much more complementary than 240/9240 even though the P channel part if 100V and the N is 200. In either case, with 40V rails, both will do fine, and besides 540/9540, which were your alternatives, are 100V also.
If you only have 40V rails (no auxiliary rails for the driver stage of the MOSFETs), the power with 540/9540 would be about 7W more, assuming no rail sag (no drop from 40V under load).
If you design your driver stage well (so it can swing to voltages close to power rails), the maximum power will be about 130W into 4 ohms (again, assuming no rail sag - but this is a large assumption). The fact is, this will not work reliably with 540/9540. It would if you could assume a resistor load, but not on a realistic speaker load. Also, it is doubtful they would survive a short circuit when hot. Although the power rating for these is 100W, in practise, it is nearly impossible to come even close.
The IRFPs would be a far better choice because they are more robust. Although they are rated 150W, which does not seem dramatically better than IRFs, they have a much more massive metal tab which alowes them to better survive short current transients much higher than rated current, and also short power dissipation transients much higher than rated power. They are also easyer to fit on a heatsink while assuring uniform distribution of force - the TO220 tabs on IRF transistors bends too easily.
Also, simulation and real world testing confirms much better distortion figures on IRFP types - sometimes even 10x better!
If I was making an amp with your conditions (+-40V rails, 4 ohm load), I would use IRFP240 and IRFP9140. This are much more complementary than 240/9240 even though the P channel part if 100V and the N is 200. In either case, with 40V rails, both will do fine, and besides 540/9540, which were your alternatives, are 100V also.
If you only have 40V rails (no auxiliary rails for the driver stage of the MOSFETs), the power with 540/9540 would be about 7W more, assuming no rail sag (no drop from 40V under load).
If you design your driver stage well (so it can swing to voltages close to power rails), the maximum power will be about 130W into 4 ohms (again, assuming no rail sag - but this is a large assumption). The fact is, this will not work reliably with 540/9540. It would if you could assume a resistor load, but not on a realistic speaker load. Also, it is doubtful they would survive a short circuit when hot. Although the power rating for these is 100W, in practise, it is nearly impossible to come even close.
The IRFPs would be a far better choice because they are more robust. Although they are rated 150W, which does not seem dramatically better than IRFs, they have a much more massive metal tab which alowes them to better survive short current transients much higher than rated current, and also short power dissipation transients much higher than rated power. They are also easyer to fit on a heatsink while assuring uniform distribution of force - the TO220 tabs on IRF transistors bends too easily.
Also, simulation and real world testing confirms much better distortion figures on IRFP types - sometimes even 10x better!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.