Sony SITs - 2SK82 E Rank With 2SJ28 F Rank

[Zen Mod]

Files from RKR

tnx RKR , for everything

here they are , for rest of Greedy Boyz

https://dl.dropboxusercontent.com/u...er Amplifier TA-N88 Technical Information.pdf

https://dl.dropboxusercontent.com/u... Premplifier TA-E88 Technical Information.pdf

 

[R-K Rønningstad]

 

[Nelson Pass]

Here's the AES paper on the Sony VFET switching amp. Simple enough as a concept.

 

[R-K Rønningstad]

SIT amp

A little afraid throwing this circuit into the ring to all you geniuses. Based on some Borbely front-end I once built. It has nearly 2 A of bias current. I have not tried to tweak it in Lt Spice, and most importantly, I have not built it. Thanks to M. Rothacher for 2SK82 model. (And 2SJ28, alternatively, I "inverted" the 82. I can not remember...)

 

[R-K Rønningstad]

Oooops!
Corrected to +/-24 V for the input stage. I experimented and accidentally left it at +/-30 V.

 

[Michael Rothacher]

Bear in mind the difference in vgs values from N to P. If you just flip my model to make a P-channel you may want to account for that by adjusting the parameters, i.e. the vgs shift, if you use the concise model from Linear Audio #6.

 

[Zen Mod]

me no habla SIT followers

2N3055 are good enough for that task

 

[R-K Rønningstad]

Bear in mind the difference in vgs values from N to P. If you just flip my model to make a P-channel you may want to account for that by adjusting the parameters, i.e. the vgs shift, if you use the concise model from Linear Audio #6.

Thank you, Michael. I used your model inverted - surely. Now I went to Jan's page and ordered Linear Audio issue #6.

/RK

 

[woody]

Because of the high cost of these devices perhaps a cascode output stage might be worth looking at. Similar to this from Nelson's classic article. Using 2 banks of mosfets to provide the cascode shelving for the sits. Where the sits take the place of the darlington pairs of bipolars.

 

[R-K Rønningstad]

Well, my idea was basically to do the following:

1. Use only JFET or VFET/SIT, not MOSFET or bipolar
2. Complementary throughout
3. Use a VFET/SIT follower output stage capable of handling my full range speakers (Quad ESL-63)

Of course one can discuss if this really is necessary, e.g. could the second stage be MOSFETs without any sonic degradation? But the concept is all-JFET...
Furthermore, I have not cascoded the input JFETs, but I expect I have to do something here. Because the VFETs have to be driven from something on higher voltage rails. And I am looking towards 2 x 60 W into 8 Ohms or in that vicinity.
I am also thinking along the line regulated power supplies, with pass transistors...
I cannot say cost is no object, but I strongly believe the bulk of the cost is power supplies, heat sinks and the enclosure. So there you have the background.
But thank you for the proposal of course!

 

[JohnW]

Does anyone have the full Datasheet for the 2SJ28 / 2SK82's or have any idea of the capacitance values for CDS, CGD & CGS for these parts?

 

[Scheff]

Seems , many experts around . I´m not, but I like experiments.
I wonder if I can replace a IRLI3705N in a VCCS by a sit transistor. And I ´d like to know how I can get a SIT... Can anyone help?