SemiSouth boiler room

Perhaps two years ago I first heard about the growing interest by the audiophile community in SemiSouth JFETs. My exposure was courtesy of Jeff Casady, a longtime friend, a cofounder of SemiSouth, and the inspirational leader of SemiSouth from the beginning. He told me about this guy named Nelson Pass, and that because of him the company had amended its specification for its enhancement mode JFETs to make available more parts for the audio community. Thus was born the "A" for audio campaign.

I am also a cofounder of SemiSouth. My role has always been tilted toward finding ways to make the company's products usable by potential customers. During my time as a fulltime employee of SemiSouth (August 2005 through August 2007) I was the Chief Technology Officer. While I was CTO, the channel design for the SemiSouth enhancement-mode JFET was born from which came the SJEP120R100.

Since the loss of SemiSouth as a producer of silicon carbide JFETs late last year, I have had time to contemplate the impact of SemiSouth. I have also heard from former customers seeking continued access to the company's products about which I hear almost universal admiration. Whether from the linear community (audio) or from the intended customer base of switch-mode power electronics, the product itself has been hailed as (so far) unmatched by alternatives.

Most of all I have enjoyed learning about the audio community. That motivated me, as mentioned in the thread sjep120r100 faking from China , to write some articles about SemiSouth products that might help DIY'ers understand the technology better, and thus use it with confidence. Since alternatives to SemiSouth technology are on the horizon, I am confident that access to new produciton will be made available to the audio community. Understanding a bit more about what you seek might help you make good decisions about those alternatives in the future. My next post will be my first article entitled "Is it SemiSouth?" in which I offer some insights into protecting yourself from fakes.

In addition, I hope to stimulate comments and ideas for new articles in the future. You are welcome to send me an email with comments or ideas, but looking at the spirited culture of posting on this blog, I expect to hear from most of you by posts on this thread. I look forward to your thoughts and inputs.

Mike Mazzola
 
  • Like
Reactions: 1 user
Is it SemiSouth?

As promised, attached is the first article entitled "Is it SemiSouth?" This is my attempt to contribute to the understanding of what is inside the TO247 package and how to feel confident that you have a genuine SemiSouth JFET.

I have tried to balance my professor's instinct to bathe you in technical details with my own realization that I would prefer you to find this interesting. Only you will be judge of that, but if it seems a bit corny, please have mercy. :eek:

Next up will be an article on what I have learned about the native linearity of SemiSouth JFETs and my personal thoughts on simulation, with the goal of making available relatively simple methods for estimating the performance of SemiSouth JFETs in real circuits.
 

Attachments

  • Is it SemiSouth.pdf
    355.1 KB · Views: 1,539
  • Like
  • Thank You
Reactions: 1 users
I wonder what would have happened if you and your friends had been able to find a suitable condition Amphicar? Would you have driven desperately into a raging torrent on the way home one day? would Semisouth still have formed in the way that we knew it? would future employment prospects have been damaged by you turning up in an amphibious vehicle?

thanks for the read, luckily I bought a small stash already, pity I couldnt justify more expense when you still had some, but very glad to have our very own tame SiC expert here at the forum.
 

Gyuri

Member
2003-11-03 6:47 pm
Budapest
Hi Mike,

Thank you for your effort what you carried out for our satisfaction.:)
I'm honestly say I'm sorry for what happened with SemiSouth.:(
We may have seen similar stories.
For example Tripath.
Or Mylex.
A company that deserve a better fate, fails.
Others who produce garbage, lives like a fish in water.:mad:
It's no good.:apathic:

I have four SJEP120R100A, from last Farnell batch.
I have checked them with DMM diode checker, according to your article.
But I don't get similar results.
I have three different DMM, but none of these read similar result, than yours.
Maybe because the measuring voltage too low?
Your Fluke how many voltages give in diode checker mode?
Mine DMM's gives 2.88V, 2.72V, 2.55V volts.
Actually these are shown open circuit, for all FETs, in either directions.

Sorry for my crappy english.

Best Regards:

Gyuri
 
Hi Mike,

Thank you for your effort what you carried out for our satisfaction.:)
I'm honestly say I'm sorry for what happened with SemiSouth.:(
We may have seen similar stories.
For example Tripath.
Or Mylex.
A company that deserve a better fate, fails.
Others who produce garbage, lives like a fish in water.:mad:
It's no good.:apathic:

I have four SJEP120R100A, from last Farnell batch.
I have checked them with DMM diode checker, according to your article.
But I don't get similar results.
I have three different DMM, but none of these read similar result, than yours.
Maybe because the measuring voltage too low?
Your Fluke how many voltages give in diode checker mode?
Mine DMM's gives 2.88V, 2.72V, 2.55V volts.
Actually these are shown open circuit, for all FETs, in either directions.

Sorry for my crappy english.

Best Regards:

Gyuri
Thank you Gyuri for reporting your results. I had a feeling that I would send all good DIY'ers to their DMM's with this article and, of course, I can't guarantee your results will match mine exactly with different equipment. With two different models of Fluke DMM I got the same results as reported in my article. I know that both DMM's have batteries with at least 6 V of open-circuit potential. I think you are correct to suspect the battery voltage as an explanation for your results. Using an external battery and a resistor to bias the diodes of your parts and then using the DMM to measure the resulting voltages across the terminals of the device should eliminate this ambiguity.

By the way, I noticed a typo in my article. The bandgap energy of 4H SiC is 3.24 eV, not 3.4 eV. I want to slip that errata in before one of my gallium nitride friends wags a finger. :eek: (GaN has a bandgap of 3.4 eV.)
 

Gyuri

Member
2003-11-03 6:47 pm
Budapest
Thank you for your answer,
You know, I'm not an electronics expert.
If I'm thinking of it, I'm not an expert in anything.
Maybe I'm just an expert of suffering.
But would you do it for me, measure your DMM's diode measuring output voltages with each other?
I'm just curious.

Thanks,

Gyuri
 
Thank you for your answer,
You know, I'm not an electronics expert.
If I'm thinking of it, I'm not an expert in anything.
Maybe I'm just an expert of suffering.
But would you do it for me, measure your DMM's diode measuring output voltages with each other?
I'm just curious.

Thanks,

Gyuri
I checked both of my Fluke 16's at home (a rather cheap DMM) and measured 2.66 V on one and 2.64 V on the other when in diode check. I have not checked the Fluke 87 in the lab yet. Both models were able to source adequate current to measure a cut-on voltage when I tried them with multiple parts. Other DMM's may read open if the leakage current with the available open circuit voltage is too low, at least that's my speculation. I'm assuming that if a meter has a diode check function, it was designed with 0.7 V in mind.
 
Thanks Semisouthfan!

good article!


I have three DMM at my house, all three did show "infinite resistance". So the fluke DMM must have another characteristic.....

Thanks for the feedback. Your not the only one seeing this with your DMM's. Since diode checkers are not expecting such high built in potential, it's not surprising. Bias either junction with a power supply or battery in series with a resistor and that same DMM on a voltage setting will measure what the diode checker won't. When doing the measurement, say gate-to-source, leave the drain pin open (or if gate-to-drain, leave the source open). Again, thanks for relating your experience.
 
very interesting Blues,

All my three DMMs have 9V supply....:p

Some DMMs might not have a 9V battery supply like Fluke DMMs have. These lower battery supply DMMs typically can only measure 0.6V silicon diodes (or bjt pn junctions) or the now rare 0.3V germanium diodes/transistors.

Thanks for the article, Mike.

I'll add to your list: Silicon Schottky barrier diodes which are not rare and cut on at about 0.3-0.4 V.;)