Best MOSFET for output stages

Status
Not open for further replies.
Jump to Latest
#61
Mooly said:
Electrical damping of speakers is misunderstood too as the speaker is a mechanical device. An amp with a "high" damping factor is no better able to control a speaker than one of "low" damping within reason.
Click to expand...

Yeah, damping factor makes a difference but not as much as people tend to think based on damping factor numbers. The total electrical damping is a function of the amp's output resistance plus the cable resistance plus any passive crossover series resistance and finally, the addition of the voice coil resistance. In this summation, the voice coil resistance usually dominates by an order of magnitude or so.
 
#65
soekris said:
I will do that and report, currently have a testbench (physical hardware) ready designed for both bipolar and mosfets.... The bipolar results are very good, hope the mosfets will be even better as mosfets are my preferred output devices....

Soren
Click to expand...

Bjts will have less THD at open loop, but the mosfets allow
for more open loop bandwith, therefore , there s more
available negative feedback , reducing this THD more efficently.
 
#66
That's exactly what my test setup is supposed to find out...

The bipolar setup currently gives about 600 Khz power bandwidth though, so it's not like it's slow 🙂

And the new Toshiba power mosfets have Yfs at typical 12S, so given a reasonable bias I hope for lower THD than typical for mosfets....

I also have sets of 2SK1530/2SJ201, BUZ900P/BUZ905P and IRF5210/FQP18N20 to test with, I'm building exchangeable output "modules" on small heatsinks....

Fyi, the input stage is jfet with folded cascode and the overall design is with no NFB at all, pretty simple but high performance....

Soren
 
#67
soekris said:
And the new Toshiba power mosfets have Yfs at typical 12S, so given a reasonable bias I hope for lower THD than typical for mosfets....
Click to expand...

The Yfs is considerably better than the older K1530/J201 parts but the plots don't show the P-channel one hitting 12S until an unreasonably high drain current. Even though the N-chan device is around 30% higher Yfs, the 'headline' values are the same for both. Methinks someone did a paste job on the datasheet... Nice parts and not too expensive either, even from Digikey.😛
 
#68
The only way to get reasonable linearity from the HEXFET's is to run them at high current... as in Nelsons class A designs... but to me while certainly effective and capable of very good results, it's a bit of a brute force approach.
Click to expand...

Or you can use local error correction. It really works quite well if done properly.
 
#69
Yes, the Toshiba 2SK3497/2SJ618 are not that expensive and Digikey have them. But not in stock, it took something like 7 months for my 100 sets from Digikey, I was surprised when they finally shipped....

But if they turns out to be as good as I hope then I will purchase a large qty of them, also for others to enjoy....

Soren
 
#70
abraxalito said:
The damping factor is not measured when the FET is fully on as its not a realistic condition - if its fully on your amp is clipping. To do a damping factor measurement with the FET hard on, the amp would have to be presented with a DC signal sufficient to drive it close enough to one or other of the supply rails. Most amps will have response roll off meaning that such a DC signal will potentially overload the input common mode voltage (assuming its servoed or has unity gain at DC), making the test invalid. If the amp is AC coupled, a DC input will fail to drive the output anywhere at all.
.
Click to expand...

But the hard on MOSFET is the best case for damping factor.
The worst case is when the MOSFET is least on and so at the highest resistance. At a low voltage the MOSFET has a resistance on around 50 ohms which is disasterous for the damping factor.

I was always taught that a high damping factor was very important and it is very often quoted for amplifiers.
 
#71
I wouldn´t agree completely about this, the damping factor is more a result from the amount of feedback (local or global) than the resistance of the FETs themselves. The output impedance of the amplifier approaches zero if there is sufficient feedback no matter what FETs are used. But of course, once you reach the maximum output current of the FETs, there is no longer any way to control the back induced force from the speaker.
 
#73
P/N Mosfets like Hitachi and so on did take audio Designer in the early 80s circuits, often P/N Mosfets are not really complementary, the beneficary is that you pay a lot of money for nothings compare to Vertical Switching N-Channel or Vertical Linear N-Channel Mosfets in Quasicomplementary Mosfet Circuit

for excellent sounding very rugged Mosfet Amp use N-Channel Quasicomplementary Mosfet Circuit

This Mosfets working with execllent results

APL502L or APL 602L or IXTK46N50L or IXTK110N20L2

if you want save cost for HI-FI choose

IRFP460 or IRFP460N

Dont use output relay for turn on Delay and DC Protection, take a crowbar

Trust me we have test it, benefit from P/N Mosfets is only theory.....

You can use lateral mosfets (BUZ9xxD, Exicons) in N-Channel Quasicomplementary Mosfet Circuit too
 
Last edited:
#74
okkyn said:
Could someone tell me, which one is better, using double die lateral mosfets (BUZ9xxD, Exicons) or use two single die? Can we assume those dies inside one package are matched? Thanks. 🙂
Click to expand...

I have been told that the double die mosfets are more prone to oscillation.
The double die mosfets also can be difficult to cool, if you are going to run them at high power outputs.
I think though that they have their uses, like increasing current capability in an existing amplifier design.
 
#76
Try this simple (well, maybe not so simple) experiment. Either build or simulate an output stage with no NFB and a reasonable (~100 ma) amount of idle current. I did that with a number of complementary MOSFETs pairs and concluded that the BUZ5xx/9xx lateral devices yielded the lowest distortion over a wide range of power levels. Yes, they are expensive, but I designed and built a pair of 400W monoblocks about three years with these devices, and they have performed flawlessly.
 
Status
Not open for further replies.
Top Bottom