EEVblog Electronics Community Forum
Electronics => Beginners => Topic started by: pulsed_power on July 06, 2023, 07:58:56 pm
-
Yamaha 2SK76 & 2SJ26 Reproduction. How could these be reverse engineered to obtain production formulas?
KEN SHIRRIFF
-
While theoretically it is possible to replicate anything, it certainly is out of reach for you, me and mostly anyone. The best you can do is try to find some other complementatry pair with similar (or better) specifications which are still available to buy.
-
Can't read the specs.
Perhaps Toshiba 2SK135 and 2SJ50? (160 V, 7 A) or 2SK1530 and 2SJ201 (200 V, 12 A)? Both sets were used in the legendary Elektuur/Elektor "Crescendo" amp.
The odds of finding parts from 1968 are around zero.
-
Yamaha 2SK76 & 2SJ26 Are VFETS
-
http://www.amplimos.it/images/2sj26_2sk76_data.jpg (http://www.amplimos.it/images/2sj26_2sk76_data.jpg)
-
Parts list and photos:
It looks like everything is available except for the VFET's
-
A little more info
-
While theoretically it is possible to replicate anything, it certainly is out of reach for you, me and mostly anyone. The best you can do is try to find some other complementatry pair with similar (or better) specifications which are still available to buy.
So the question becomes how many Brinks Trucks to build? Then how many Brinks Trucks of parts can be reproduced and sold? How many units does someone like ONSEMI need to produce to gain their interest?
-
Can't read the specs.
Perhaps Toshiba 2SK135 and 2SJ50? (160 V, 7 A) or 2SK1530 and 2SJ201 (200 V, 12 A)? Both sets were used in the legendary Elektuur/Elektor "Crescendo" amp.
The odds of finding parts from 1968 are around zero.
Those are MOSFET, Yamaha's are VFETS
-
Yamaha 2SK76 & 2SJ26 Are VFETS
Sorry, the 2SK/2SJ part numbering led me to assume MOSFETs.
And the totally unreadable datasheet didn't help (your new post is no better).
-
Yamaha 2SK76 & 2SJ26 Are VFETS
Sorry, the 2SK/2SJ part numbering led me to assume MOSFETs.
And the totally unreadable datasheet didn't help (your new post is no better).
Sorry I tried
-
https://audio-database.com/YAMAHA/amp/b-2.html
-
the pair characteristics required for pure complimentary service FET were pursued, and the original manufacturing method and pair selection method were devised and developed. The device chip size of Pch and Nch, impurity concentration and gate mesh structure are carefully examined, and pair characteristics in a wide operating range are matched by strict control in the manufacturing process. Then, pairing is done by Yamaha's original pair selection method.
-
How many units does someone like ONSEMI need to produce to gain their interest?
Infinite.
I looked a bit into this, and the problem is that there are practically no wafer fabs in the world capable of doing vertical processing. There's a reason everything is produced using planar technology today. Vertical wafer processing is damn expensive and involved.
SemiSouth went broke 10 years ago, they were the only ones pursuing this (VFET/SIT).
Your best bet is NOS, either the Yamaha parts or perhaps Sony 2SK82/2SJ28.
-
Not all datasheets mention if they're VFETs I think.
I thought Hitachi would have made the originals but their package looks Sanken? 85V rails (edit: for driver stage) with two mosfet pairs, gee that's asking a lot. I think the thermal characteristics are very important for the bias not running away.
Consider Exicon (http://www.exicon.info) lateral power mosfets, they had TO-3 but now only have TO-247, TO-264 offerings. ECF10N20, ECF10P20 and there are double-die 250W puppies too. You could email them and ask. Their substitutes list has the '176 but not sure if it's related to the '76.
BUZ900/BUZ905 might be possible alternates as well.
-
Not all datasheets mention if they're VFETs I think.
I think you need to read up a bit here. The official name is SIT or "static induction transistor", VFET was the marketing name back then. They're much more like power JFETs than MOSFETs.
-
Yes I'll read more. I had them as enhancement mode devices until I looked at detailed specs. VFET's in 1976?! So they are extra rare birds then.
Rated max. 40V gate-source and VGS(off) of 20/25V.
-
How many units does someone like ONSEMI need to produce to gain their interest?
Infinite.
I looked a bit into this, and the problem is that there are practically no wafer fabs in the world capable of doing vertical processing. There's a reason everything is produced using planar technology today. Vertical wafer processing is damn expensive and involved.
SemiSouth went broke 10 years ago, they were the only ones pursuing this (VFET/SIT).
Your best bet is NOS, either the Yamaha parts or perhaps Sony 2SK82/2SJ28.
May not be a complete dead issue, looks like some effort continues:
n October 2010, in order to drive the expansion of its SiC fabrication facility, Power Integrations Inc of San Jose, CA, USA (which supplies high-voltage integrated circuits for compact, energy-efficient power conversion in electronic products for AC-DC, DC-DC and LED lighting applications) made a $30m strategic investment in SemiSouth (including an equity investment, a technology license and other financial commitments). The firms also collaborated on driving the adoption of SemiSouth’s SiC technology, aiming to speed the development of efficient power conversion devices for applications including solar and wind inverters as well as hybrid/electric vehicles.
And:
https://www.diyaudio.com/community/threads/semisouth-sic-power-jfets.148959/ (https://www.diyaudio.com/community/threads/semisouth-sic-power-jfets.148959/)
-
Yes.
But that was 2010. Power integrations dropped financing SemiSouth in 2012 and it went bust.
It's a dead duck, sorry.
-
https://en.wikipedia.org/wiki/Nelson_Pass
His SIT amplifiers are the first of a new generation of audio amplifiers using Static Induction Transistors in a single-stage, single-ended, Class A circuit without feedback or degeneration.[13] The SIT chip combines a square-law input character with a low impedance output to form the only solid-state gain device, which Pass claims, "behaves like a triode tube." There is no output transformer on the SIT amps. The point of the SIT is that it behaves like a triode but at lower voltages and higher currents, so it doesn't need a matching transformer to deliver power to 8 Ohms. Like tubes, SITs have soft overload clipping. When brief bursts of musical energy occur, SITs react with rounded waveform tops instead of sharp and hard clipping of solid-state. SITs have a curve which looks a lot like a triode vacuum tube; low at first and climbs steadily. The distortion curve is similar, a steady rise instead of a valley with high distortion at both ends.
Is it possible Mr. Pass has invested $60m to create his SIT's?
-
Based on the sources, it seems you copy pasted an ad. I would assume Mr. Pass was using new old stock parts (a decade ago) in what sounds like a limited production product.
This was a niche transistor type that never really took off, neato that it sounds like a triode in an audio amp, but that isn't a "support a special device" size market. Like all the other now discontinued FETs you're at the mercy of the salvage/NOS market unless InterFET (I guess you could inquire there) or some other niche transistor supplier is having them made already. That's not to say it's impossible to get a discontinued transistor back in limited production, it does happen, but that needs both demand (or equivalent pile of money) and extant production equipment/processes that can already do the job.
-
Pass Labs is a current production company that designs and builds very high end audio equipment. He currently uses the SIT FET in his products.
https://www.suncoastaudio.com/collections/pass-labs (https://www.suncoastaudio.com/collections/pass-labs)
-
Where does it say they're being used in a production product? I went back and read a little closer, I was wrong, he only ever used SITs in some prototype stuff.
Here's the actual website. https://www.passlabs.com/ (https://www.passlabs.com/) zero mention of VFET/SIT on any of the amps offered. It's something the dude messed with personally, hyped a bit and probably ran into the same problem you have, they're out of production, rare, and expensive, and assuming Benta is correct, built with a no longer available process.
-
May not be a complete dead issue, looks like some effort continues:
n October 2010, in order to drive the expansion of its SiC fabrication facility, Power Integrations Inc of San Jose, CA, USA (which supplies high-voltage integrated circuits for compact, energy-efficient power conversion in electronic products for AC-DC, DC-DC and LED lighting applications) made a $30m strategic investment in SemiSouth (including an equity investment, a technology license and other financial commitments). The firms also collaborated on driving the adoption of SemiSouth’s SiC technology, aiming to speed the development of efficient power conversion devices for applications including solar and wind inverters as well as hybrid/electric vehicles.
SiC stands for silicon carbide and has nothing to do with SIT. And while they made SiC JFETs "for audio" https://www.farnell.com/datasheets/1455139.pdf (https://www.farnell.com/datasheets/1455139.pdf), they are not the same beast. And from a quick glance into datasheet it seems this "for audio" is meant to be class D.
-
Where does it say they're being used in a production product? I went back and read a little closer, I was wrong, he only ever used SITs in some prototype stuff.
Here's the actual website. https://www.passlabs.com/ (https://www.passlabs.com/) zero mention of VFET/SIT on any of the amps offered. It's something the dude messed with personally, hyped a bit and probably ran into the same problem you have, they're out of production, rare, and expensive, and assuming Benta is correct, built with a no longer available process.
I have sent a RFI to Suncoast asking this question. Maybe they will reply.
-
Where does it say they're being used in a production product? I went back and read a little closer, I was wrong, he only ever used SITs in some prototype stuff.
Here's the actual website. https://www.passlabs.com/ (https://www.passlabs.com/) zero mention of VFET/SIT on any of the amps offered. It's something the dude messed with personally, hyped a bit and probably ran into the same problem you have, they're out of production, rare, and expensive, and assuming Benta is correct, built with a no longer available process.
I have sent a RFI to Suncoast asking this question. Maybe they will reply.
Anyone recognize this response: 240 / 9240's
-
Put IRF as prefix, then you'll understand. It's a middle finger.
-
IRFP240/9240 are a currently available MOSFET complementary pair with a DC safe operating area, allowing for linear operation.
https://www.mouser.com/ProductDetail/Vishay-Semiconductors/IRFP240PBF?qs=cvaI6Thkwxs4DuqF5TWfHQ%3D%3D (https://www.mouser.com/ProductDetail/Vishay-Semiconductors/IRFP240PBF?qs=cvaI6Thkwxs4DuqF5TWfHQ%3D%3D)
Only other place to look is Soviet surplus, if they ever made anything like that.
-
IRFP240/9240 are a currently available MOSFET complementary pair with a DC safe operating area, allowing for linear operation.
https://www.mouser.com/ProductDetail/Vishay-Semiconductors/IRFP240PBF?qs=cvaI6Thkwxs4DuqF5TWfHQ%3D%3D (https://www.mouser.com/ProductDetail/Vishay-Semiconductors/IRFP240PBF?qs=cvaI6Thkwxs4DuqF5TWfHQ%3D%3D)
Only other place to look is Soviet surplus, if they ever made anything like that.
Not sure why Pass Labs is using these, they have bet their company on them.