Semisouth goes Dodo ; what now ?

Jump to Latest
#105
Great question. I can say for sure that Cree makes a fundamentally different device technology than SemiSouth made (MOSFET vs. JFET: sparks flew during the competition at the time). But what matters, I think, is the shape of the output curves. And in that sense, there are some similarities between the two. But they are not the same, so it will require experimentation to answer your question for sure.

flg said:
:up::up::up:
There has been some mention of the devices like C2M0080120D from Cree. Any coments towards similarities to the SS devices we like or audio uses in general?
I would aquire a few and try but available $$$ are tight right now...
:Pawprint:
Click to expand...
 
#111
Dr. Mazzola,

I wonder if you and your team did ever generated Spice model for your devices ?
If so, would you care to share them ?

As far as I know, there is only model for one device around which was generated by a private person.


Many thanks and Merry Christmas,
Patrick
 
#112
EUVL said:
Dr. Mazzola,

I wonder if you and your team did ever generated Spice model for your devices ?
If so, would you care to share them ?

As far as I know, there is only model for one device around which was generated by a private person.


Many thanks and Merry Christmas,
Patrick
Click to expand...

Dear Patrick,

Thank you for the Christmas greetings! I hope the same for you.

The published SPICE model you refer to (I believe) is the SJEP120R100. That model is attached for those so interested. If the subcircuit model gives you trouble (it does for me in the student versions of PSPICE that I use) then use the text after the ".model SJEP120R100 njf" statement, which is the core enhancement-mode JFET model.

I think I have previously posted a good SJDP120R045 (45 mOhm depletion mode) JFET model that one of my students used in his dissertation. I'll have to look in my records to find it to repost. You might search this forum to see if my memory is true.

The same student developed a SJEP065R055 model for a class, which is also attached to this post. This part is a 650 V, 55 mOhm enhancement mode JFET that SemiSouth released as part of an effort to address the much larger market for 600-V FETs. This is an awesome part that folks should consider as an alternative to the ubiquitous R100. If anyone would like to try some I have them in inventory.

Regards,

Mike Mazzola
 

Attachments

#115
EUVL said:
Any models available for the SJEP170R550 ?


Thanks again,
Patrick
Click to expand...

Not that I know of. SemiSouth did not produce one and my research group has not needed to produce one either. A SJEP120R100 SPICE model could be modified with a little scaling.

But truthfully, none of these models do a good job with the curvature of the output characteristics (ID vs. VDS) near saturation, a problem no one has been able to fix using the core SPICE JFET model. So predicting amplifier linearity with the available models is not recommended. It is ironic that the gradual "knee" or transition to saturation of the output curves for the R100 was viewed by the power switching community as a curiosity indicating lack of maturity in the channel design. Silicon power MOSFETs have sharper transitions, which is a major reason, in my analysis, for why the IRF parts deliver higher THD than the SemiSouth JFETs. What is one man's poison is another man's nectar.

Gee, I could be a poet.

For power switching, modeling near saturation is not really needed so it just hasn't been a priority, despite a few tries with only partial success. I have a need for modeling the transfer curves at the corners of the safe-operating-area (i.e., at hundreds of volts on the drain) and a student will finish a Master's Thesis on this improvement soon. But DIY Audio folks don't operate in that part of the device SOA, so I would say the available models are mostly useful for checking the basic functionality of a dc bias circuit because threshold and transfer curve modeling at low drain voltage is pretty good, and maybe some general guidance on AC operation, but not much more.

If there becomes a bigger demand, I might arrange for work on the near-saturation model to see if we can make this a better representation of the true characteristics. Maybe as an undergraduate research project. 😱
 
#116
As stated "It is ironic that the gradual "knee" or transition to saturation of the output curves for the R100 was viewed by the power switching community as a curiosity indicating lack of maturity in the channel design. Silicon power MOSFETs have sharper transitions, which is a major reason, in my analysis, for why the IRF parts deliver higher THD than the SemiSouth JFETs. What is one man's poison is another man's nectar." is a favorite post! Has anyone had the huevous to say anything like that here, before? It's my favorite post too! 😀
Holiday cheer, and new year wishes Mike :cheers:
Back to your regularly scheduled channel....
 
#117
flg said:
As stated "It is ironic that the gradual "knee" or transition to saturation of the output curves for the R100 was viewed by the power switching community as a curiosity indicating lack of maturity in the channel design. Silicon power MOSFETs have sharper transitions, which is a major reason, in my analysis, for why the IRF parts deliver higher THD than the SemiSouth JFETs. What is one man's poison is another man's nectar." is a favorite post! Has anyone had the huevous to say anything like that here, before? It's my favorite post too! 😀
Holiday cheer, and new year wishes Mike :cheers:
Back to your regularly scheduled channel....
Click to expand...

Wow, a groupie! (Just kidding!😀)

A Happy New Year wish to you too. I will be spending my New Year's Eve at the Liberty Bowl drinking beer and watching Mississippi State take on Rice. Go Bulldogs!
 
You must log in or register to reply here.
Top Bottom