Well, should only be a matter of time, here is one of their devices that you can buy already: GA04JT17-247 GeneSiC Semiconductor | Mouser
Yes, I think if Nelson postulate that these are good for A class amplifiers, these will be dead, before TBDs will be resolved.
At the end of doc there are SPICE parameters.
Whatever it is it ain't a jfet, nor is it a MOSFET. "Gate" current in the 100s of milliamperes (figures 1-3) hfe is specified in the datasheet. Capacitance seems low and linear. Looks quite interesting if you're willing to design to it.
It's different than what this forum uses. I suspect NP could design something worthwhile in minutes.
They make higher amp models.
It's an interesting idea - current gates.
Last edited:
Astute observation. This is a SiC Bipolar Junction Transistor (BJT). Ranbir Singh, the founder and chief cut up of GeneSiC has been trying to "disguise" this part as anything but a BJT for years. He seems to be trying to introduce maximum confusion. In the Mouser link to a similar part it is described as "MOSFET SiC Supr Jnctn Trans" where Supr Jnctn = "super junction" transistor.
I think Ranbir intended to associate his technology with the "super junction" class of silicon MOSFETs to deflect concern about the problems with bipolar transistors in SiC. But please don't be fooled, this part is NOT a MOSFET or a JFET. When I asked him about the name he didn't really have a good answer except "marketing." I watched Dan Kinzer, who is a well respected silicon power MOSFET expert that is now hawking SiC BJTs for Fairchild, joke Ranbir about "super junction" at a conference two years ago. The whole room broke into laughter. Moral of the story? Sometimes it's hard to come up with those clever brand names.
It's interesting that the data sheet for the GA20JT12-247, roughly comparable to the R100, is filled with TBD, but it includes an entry for DC Current Gain or "beta" (value also TBD but BF = 100 in the Spice model at the end of the data sheet). Nobody selling a MOSFET or JFET would EVER put beta in a data sheet. Beta is for BJTs, which Ranbir admitted this part is when Dan asked him in public. I haven't heard the super junction term used much since and I notice that the datasheet omits the phrase "super junction." But on p. 6 you can see "SJT" in the gate drive section.
Does that mean the GA20JT12-247 can't replace the fine SemiSouth JFET in Pass amplifiers? Maybe, maybe not. You can try it, but based on pricing on Mouser for other available parts, I speculate that the 70 mOhm part, when available, won't be cheaper than a SemiSouth JFET and I can still get you those.
That Ranbir, what a comedian!!
I remember Fairchild gobbling up a Swedish (?) firm that made SiC BJTs. They had some interesting claims for the devices, especially regarding storage time, but I never saw a data sheet. Maybe FCH wants to use them in motor drives, as they already have their tip-toes dangling somewhat in that market.
Last edited:
Yep, it was TranSiC. Two top SiC epitaxy people from SemiSouth (both former students of mine) are working for them in Sweden before transitioning back to Maine. Not sure which place is colder.
Hi Mike,
Thank you for your detailed explanation.
You know I'm not an electronics guru, so I thought this would be replacement for R100.
Maybe it is forgivable me, because I have found it as JFET on Farnell site:
SIC JUNCTION TRANS, 1.2KV, 20A, TO-247AB
Transistor Type: JFET
And yes, twice the price.
Regards,
Gyuri
Thank you for your detailed explanation.
You know I'm not an electronics guru, so I thought this would be replacement for R100.
Maybe it is forgivable me, because I have found it as JFET on Farnell site:
SIC JUNCTION TRANS, 1.2KV, 20A, TO-247AB
Transistor Type: JFET
And yes, twice the price.
Regards,
Gyuri
Hey Gyuri, your not to blame for this ridiculous subterfuge. I mean, Mouser and Farnell both have legitimate categories for BJTs, so why describe this part as a MOSFET or a JFET and list it in the search fields for these two?
Crazy but whatever.First line after the comments: .model GA20JT12 NPN
This means, of course, that the GeneSiC model uses the native NPN BJT model in SPICE.
Hmm, that's NOT the model I would select if I were selling a JFET or a MOSFET.
Nothing wrong I suppose with giving someone the benefit of the doubt. But this confusion was injected for marketing reasons understandable in the context of the reliability concerns of SiC BJTs and the history of Si BJTs in power electronics.
You are absolutely correct in saying that the forward bias voltage in the GeneSiC datasheet is unlike a silicon NPN BJT, but it is perfectly in line with a silicon-carbide NPN BJT. The difference is due fundamentally to the large separation in bandgap energy between silicon and silicon carbide, which is why the former is a "narrow bandgap" semiconductor while the latter is a "wide bandgap" semiconductor. The difference in bandgap energies is also why SiC makes such better performing switching transistors. But this difference also means that a typical forward-bias voltage of a 4H-SiC base-emitter junction is 3 V, much larger than the 0.7 V of a Si base-emitter junction. (You can use this difference to spot Si BJTs being used to fake SiC JFETs, see my post and article on that.) The base-emitter forward bias voltage of the native NPN model in SPICE was adjusted in GeneSiC's spice model to account for this (IS = 5E-47 A and VJE = 3 V). Try it and see.
An expert of the caliber of Ranbir Singh knows all of this. Even Dan Kinzer in a plenary speech to the Applied Power Electronics Conference was nearly over the top in characterizing the virtues of the SiC BJT in the furtherance of Fairchild's marketing objectives, and he was not shy about characterizing the silicon BJT as a poor cousin. But he called his device a BJT. In my humble opinion Ranbir should too.
Huh-did they really think that the engineers that actually have to evaluate the part would be fooled? Better to admit the actual nature of the part and find some good things to say about it for a given application (but that takes some work).
Humph - thinking about it, what can a 1200V SiC BJT do that a silicon IGBT or SiC Mosfet can't? I remember the silicon BJT devices of similar voltage rating - they were hilariously awful, and died deservedly when "jug" monitors were replaced with LCD panels (their niche application was for CRT horizontal deflection drivers). I have a few in my parts bins - they might as well be buried with me....
The only advantage I can see is if the SiC BJT device has a smaller device area than an SiC mosfet or jfet of similar ratings - the IGBT vs mosfet wars all over again, with more exotic materials, and before the technology is even mature...
Humph - thinking about it, what can a 1200V SiC BJT do that a silicon IGBT or SiC Mosfet can't? I remember the silicon BJT devices of similar voltage rating - they were hilariously awful, and died deservedly when "jug" monitors were replaced with LCD panels (their niche application was for CRT horizontal deflection drivers). I have a few in my parts bins - they might as well be buried with me....
The only advantage I can see is if the SiC BJT device has a smaller device area than an SiC mosfet or jfet of similar ratings - the IGBT vs mosfet wars all over again, with more exotic materials, and before the technology is even mature...
Is there a proof, that the company ever claimed, that these devices are MOSFET or J-FET ?
This link clearly says, that SJTs are SiC BJTs:
GeneSiC
Semisouthfan:
Telling, that a whole conference laughed about the President of GeneSiC, sounds for me like ego related babble with no proof.
If you are using names here and getting personal, what is your name ? Are you a (former) competitor of him ?
EDIT: I should have read your profile ... ;-)
Whatsoever: With a little bit of creative thinking some extremely simple topologies and circuit tricks should come into our mind. Such topologies, which can't be realized very well with J-FETs, SITs, MOSFETs or even Silicon BJT ...
I had a few of the 6A part on the bench and can only tell you, these are very special, at least on the second sight.
I'm pretty sure Nelson Pass is just preparing his next design with it ..
This link clearly says, that SJTs are SiC BJTs:
GeneSiC
Semisouthfan:
Telling, that a whole conference laughed about the President of GeneSiC, sounds for me like ego related babble with no proof.
If you are using names here and getting personal, what is your name ? Are you a (former) competitor of him ?
EDIT: I should have read your profile ... ;-)
Whatsoever: With a little bit of creative thinking some extremely simple topologies and circuit tricks should come into our mind. Such topologies, which can't be realized very well with J-FETs, SITs, MOSFETs or even Silicon BJT ...
I had a few of the 6A part on the bench and can only tell you, these are very special, at least on the second sight.
I'm pretty sure Nelson Pass is just preparing his next design with it ..
Last edited: