SemiSouth SiC Power JFETs

[AKSA]

Kees,

Your circuit is ingenious and clever, but perhaps a little too clever.

Try a simple plate loaded triode (12AT7) followed by a concertina phase splitter (12AU7), ditch the SS devices you have used to improve linearity.

From the schematic, I cannot see why you have such an abrupt phase shift around 500Hz, but you could probe a frequency sweep with the CRO at various points along the circuit since there is no (global) feedback. That should reveal it.

Hugh

 

[kees52]

Thanks, it was a quick idea to let see that it can work, I have to put fasesplitter and such, but as example that quasy can work like this yamaha idea, the modern jfets has no -40 volt Vgso but -5 volts, so everything get different.

I will also try a allfet version, with al are jfets, cascode with these devices give liniarity.

good stuff to try, over a couple days I receive 6J9P and 6J49P penthodes for testing.

these schematic I use now see post 99, and sound here, some harsh because of youtube compression I can nowhere post pure wav.

6C4N-EB hybride - YouTube

it is much fun testing, the faseflip is only when simulate, not in real multisim has bad models in them, I need to find better ones.

have a good day.\ evening.

 

[kees52]

What do you mean ditch the ss devices, you mean the dn 2540 devices? the drive I need for the semisouth is to big for tubes, therefore I use
them, now as test I search for others also, I have removed the CCS and voltage of 200 volt because it is not needed with modern jfets, the old 2sk77 has Vsgo of -40 volts these
days it is +1 to -5.

Therefore the B1 can not be used anymore, al there fets are so special.

 

[kees52]

some change in the schematic, the amp has in in it a source follower and a drain follower in one power stage, that is quasi, therefore I have need of a fasesplitter who is also drive it with one output much lower then the other, I hope you understand me, the cortina fasesplitter has also to be one what do not amplify otherwise it hase to much and then I need feedback.

a long tail fasesplitter is the one I need I think.

First I need a model of the new power jfets, and some in real to test it like this, there a a lot more possibillities, I am a little curieus about them this moment

one picture without feedback the other one with, with feedback the distortion was higher, but it is simulation, it is not perfect.

regards

 

[kees52]

ahaaa the Constellation Audio uses two balanced single ends, do we not call this circlotron for a part, difference only, voltage on speaker cable to ground. and afcourse
both sides have the same way of amplifying, quasy do have not, there is one part a amp, the other a follower..

I think a circlotron is very balanced with only one kind of mosfet or transistors, expecially higher frequenties in audio is prone to imbalance, making sound very harsh.

regards

 

[kees52]

Hi all

I have found a model here of a semisouth J-Fet, I have simulate a quick drawn amp topology to test it.

for as far the mosfet go, it is faster, but has more distortion as used in ab biassed amp, in class a it get less then a normal mosfet.

I have not the idea that if I made a quasy only Nfet amp that using semisouth is better, I need a beefy driver that is for shure.

I have some pics with sqare waves to see what is differend.

The prise fopr them is high, it is not better if used without class A idle currents, it need big current to go liniair, so in the nelson pass amps it go better then normal mosfets, I have to say, it is simulation, I have not listen yet, I will do if $ are down.

squares are picture 1 =10Khz SJEP120R100A picture 2 =100Khz SJEP120R100A picture 3 =100Khz 2SK1058, jou see slight slewing with 2sk1058 that is the only differency, last picture the schematic for testing.

regards and nice weekend.