Qusp, It's interesting that you were able to get the data sheet. I don't know that TranSiC sampled many devices. The issue is moot now that Fairchild owns them - we'll have to see what they do with the IP. Probably not much until they get repeatability and yield issues sorted out (it's SiC - there will be yield issues).
Semisouth apparently tweaked their process and "lost the recipe". They for some reason couldn't get back to the yields they had before the tweak. This seems to indicate that they didn't quite understand what they were doing to begin with. Since SiC is largely unexplored territory anyway, this is sort of understandable, but very unfortunate when you have business that's supposed to be making money, a foundary that eats money by the truckload, and no yield to show for it. Even silicon fabrication can have some screwy problems 50+ years down the line. The whole semi fab process is an arcane witch's brew of chemistry and physics, and something can slip up at any step.
Semisouth apparently tweaked their process and "lost the recipe". They for some reason couldn't get back to the yields they had before the tweak. This seems to indicate that they didn't quite understand what they were doing to begin with. Since SiC is largely unexplored territory anyway, this is sort of understandable, but very unfortunate when you have business that's supposed to be making money, a foundary that eats money by the truckload, and no yield to show for it. Even silicon fabrication can have some screwy problems 50+ years down the line. The whole semi fab process is an arcane witch's brew of chemistry and physics, and something can slip up at any step.
Supplies of parts still available
I am a new reader of this board. From what I've read, it looks like folks have an interest in knowing more about the SiC FET business and what happened at SemiSouth. I'll post some info to this thread if I see an expression of interest. I am a professor of electrical and computer engineering at Mississippi State University, which spun out SSL, and am an inventor on some of the original patents. I will be happy to share some thoughts from personal experience of the matter.
I have been impressed with the interest in the vertical-channel SiC JFET by the DIY Audio scene and would like to help keep the faith if possible. I am learning fast thanks to the many excellent blog posts concerning audio amp designs. Naturally, I'm most interested in those designs that use SiC. The innovative thinking on this site is pretty amazing.
It appears from reading some posts that there may be an interest in obtaining a few SemiSouth parts, such as the SJEP120R100 or 100A and equivalents, as well as -063's, -550's etc. Contact michael.mazzola@impowersystems.com (iMPOWER Systems) for more info.
I am a new reader of this board. From what I've read, it looks like folks have an interest in knowing more about the SiC FET business and what happened at SemiSouth. I'll post some info to this thread if I see an expression of interest. I am a professor of electrical and computer engineering at Mississippi State University, which spun out SSL, and am an inventor on some of the original patents. I will be happy to share some thoughts from personal experience of the matter.
I have been impressed with the interest in the vertical-channel SiC JFET by the DIY Audio scene and would like to help keep the faith if possible. I am learning fast thanks to the many excellent blog posts concerning audio amp designs. Naturally, I'm most interested in those designs that use SiC. The innovative thinking on this site is pretty amazing.
It appears from reading some posts that there may be an interest in obtaining a few SemiSouth parts, such as the SJEP120R100 or 100A and equivalents, as well as -063's, -550's etc. Contact michael.mazzola@impowersystems.com (iMPOWER Systems) for more info.
nice to have you here ( even if , in first moment I was ROFL , because of your moniker , but after reading your post it is logical )
in any case , if you have any question regarding use of SJEP and SJDP in audio circuits , NP is right man to be addressed , at least if you're searching for both short and clear answers
Thanks for the welcome! The more I read, the more I have to agree, NP is the man. I had heard about him from the SSL CEO back in the day, but I just recently made the time to check all this out. Pretty cool.
P.S. It was no laughing matter in Starkville. I was just over at SSL yesterday to sign for some things as the last employees were working on moving the business offices out of the building. I knew everyone by first name. Kind of sad. Everyone knew what a great technology they were making.
P.S. It was no laughing matter in Starkville. I was just over at SSL yesterday to sign for some things as the last employees were working on moving the business offices out of the building. I knew everyone by first name. Kind of sad. Everyone knew what a great technology they were making.
sorry - I didn't wrote that I specifically meant on Pass SIT-1 three-legged-fuses , as almost proper substitution for 300B + OPT 
edit : I'm not finding anything wrong with nerdville ........
on the other hand - will you reveal catch behind "A" suffix , for some of parts sold via big houses ?
are those rejects , slightly critical for inverters (say hifreq) usage , as I very cleverly assumed ?
there is no way that some specimens of same type are more suitable for audio , only the other way around ..... audio being less demanding field .........
of course - your'e not obliged to answer directly
edit : I'm not finding anything wrong with nerdville ........
on the other hand - will you reveal catch behind "A" suffix , for some of parts sold via big houses ?
are those rejects , slightly critical for inverters (say hifreq) usage , as I very cleverly
assumed ?there is no way that some specimens of same type are more suitable for audio , only the other way around ..... audio being less demanding field .........
of course - your'e not obliged to answer directly
Last edited:
Welcome aboard, Semisouthfan. Such a sad state of affairs. I can say that
I enjoyed working with people at SemiSouth, and I found that they
operated with integrity.
It would be nice if some of those parts could be obtained, particularly the
R100/R100A's as the performance is superior to the other N channel
Mosfets or Jfets that I've tried.
I think I previously have mentioned that it was explained to me that the
"A" suffix denoted parts that didn't meet the -15V Vgs leakage spec. They
were put into distribution at a discount for audio guys.
AFAIK, this means that they are perfectly suitable for audio, given that you
would not see much use for a negative Vgs in linear operation.
I enjoyed working with people at SemiSouth, and I found that they
operated with integrity.
It would be nice if some of those parts could be obtained, particularly the
R100/R100A's as the performance is superior to the other N channel
Mosfets or Jfets that I've tried.
I think I previously have mentioned that it was explained to me that the
"A" suffix denoted parts that didn't meet the -15V Vgs leakage spec. They
were put into distribution at a discount for audio guys.
AFAIK, this means that they are perfectly suitable for audio, given that you
would not see much use for a negative Vgs in linear operation.
You are right on both accounts, Nelson. Obviously you have earned your reputation among your dedicated fan base as the "Fearless Leader," as one blog post I was just reading put it.
According to the company, the "A" suffix was created for parts that sorted into the -10V bin only, otherwise, they are identical to the regular 100's (and 063's, which are two 100 die in the same TO247 package). Folks might ask why it matters when the parts are perfectly capable of blocking their rated voltage (1200V) with zero gate bias. In fact, this fantastic property of being enhancement mode changed the meaning of what is possible with JFETs when the Normally OFF JFET first hit the streets about four years ago. I know, as I had the privilege of making some of the earliest technical presentations of this part's performance in power converters in applications forums such as the Power Electronics Specialist Conference. After that, my colleagues stopped telling everyone that the problem with the JFET is that they were normally on. What fun that was.
The need for a negative bias rating is a story for power switching: We need to be able to bias the gate negative to sink the Miller current during high-voltage, nanosecond switching. This is not a unique requirement of SiC FETs as it is commonly done in both silicon MOSFET and IGBT switched power converters.
But you Class-A audiophiles don't need negative bias hardly at all, hence a "100A" or a "063A" is no different from a "100" or a "063" in terms of use potential. I know that, with your help, the "A" suffix quickly came to mean "Audio" among many, but within the company they tried to make that an urban legend. "A" simply meant an amended data sheet. LOL! They knew who was buying those puppies!
At the end, however affectionately the SJEP and SJDP series was with the DIY Audio community, they were coveted even more by the power electronics community for which they were first intended. I've noticed some comments about SiC MOSFETs in this thread, but when SemiSouth was making JFETs, the SiC MOSFET was no threat. SemiSouth was a victim of its own growth. Too much, too fast. Not the first time that's happened in the world of entrepreneurship, I'm afraid.
Anyway, I can broker a significant (but not inexhaustible) supply of a number of popular SemiSouth part numbers. Just send me an email.
Regards and best wishes from a new fan.
According to the company, the "A" suffix was created for parts that sorted into the -10V bin only, otherwise, they are identical to the regular 100's (and 063's, which are two 100 die in the same TO247 package). Folks might ask why it matters when the parts are perfectly capable of blocking their rated voltage (1200V) with zero gate bias. In fact, this fantastic property of being enhancement mode changed the meaning of what is possible with JFETs when the Normally OFF JFET first hit the streets about four years ago. I know, as I had the privilege of making some of the earliest technical presentations of this part's performance in power converters in applications forums such as the Power Electronics Specialist Conference. After that, my colleagues stopped telling everyone that the problem with the JFET is that they were normally on. What fun that was.
The need for a negative bias rating is a story for power switching: We need to be able to bias the gate negative to sink the Miller current during high-voltage, nanosecond switching. This is not a unique requirement of SiC FETs as it is commonly done in both silicon MOSFET and IGBT switched power converters.
But you Class-A audiophiles don't need negative bias hardly at all, hence a "100A" or a "063A" is no different from a "100" or a "063" in terms of use potential. I know that, with your help, the "A" suffix quickly came to mean "Audio" among many, but within the company they tried to make that an urban legend. "A" simply meant an amended data sheet. LOL! They knew who was buying those puppies!
At the end, however affectionately the SJEP and SJDP series was with the DIY Audio community, they were coveted even more by the power electronics community for which they were first intended. I've noticed some comments about SiC MOSFETs in this thread, but when SemiSouth was making JFETs, the SiC MOSFET was no threat. SemiSouth was a victim of its own growth. Too much, too fast. Not the first time that's happened in the world of entrepreneurship, I'm afraid.
Anyway, I can broker a significant (but not inexhaustible) supply of a number of popular SemiSouth part numbers. Just send me an email.
Regards and best wishes from a new fan.
Thanks for dropping in here...I would love to get a hold of some depletion mode power jfets to play with. So which ones are these now...SJDP? The ones on the J2 specifically, if I'm correct?
I do have depletion mode parts, including the SJDP120R085 that I've seen on some schematics and blog posts. Send me an email at michael.mazzola@impowersystems.com.
The J2 uses the R100 enhancement mode parts.
I hope you have lots of all those parts, Michael. I have a small list of JFET
designs following the F6, most of them finished, and I would like to see
some of them in DIY circulation.
Not to mention ZM - God knows what he dreams about after sausages and
Serbian alcoholic beverage.
There may not be enough for commercial production, but we can certainly
tease the big boys.
Thanks Michael...I have your response.
The J2 uses the R100 enhancement mode parts...
Is the SJDP120R085 a good part for some Pass designs? Or is it better to go for the R100? Is it going to get some circuit time with you in your estimation? Thanks!
