Srajan has the 1st SIT amp for Une dégustation

[wrenchone]

It would be interesting if you could swap some SIT devices with the guy who heads up Digital Do Main in Japan. Apparently, he's keeping some of the old Tokin SIT product line alive for his amplifiers.

 

[Nelson Pass]

Don't need to. I have a small stash of those.

 

[wrenchone]

Interesting. Have you compared curves?

 

[fireworks]

Can anyone post a link that explains what exactly are the SITs, how they work, etc. and why are they better than regular transistors ?
I saw on the Internet that they are supposed to be 1000 times faster than an ordinary transistors ? SAY WHAT ?

 

[Zen Mod]

naah

they're nothing special
just sand triodes

 

[Nelson Pass]

Can anyone post a link that explains what exactly are the SITs, how they work, etc. and why are they better than regular transistors ?
I saw on the Internet that they are supposed to be 1000 times faster than an ordinary transistors ? SAY WHAT ?

They are quite fast, but that is not the relevant item. They behave like
tube Triodes, but can do so at voltages and currents of interest to
loudspeakers, removing the need for transformers, heaters, and so on.

I am writing a nice little article on the subject which will make an
appearance after a bit.

 

[Tea-Bag]

In the meantime, I scoffed up a few of the acroman's stash.

 

[fireworks]

This is the link where it says that they are 1000 times faster than transistors:

New Scientist - Google Books

However that article is from the 80s so the regular transistors have progressed a lot since then.

The only useful link I found that explains how they work is this:

Junction field-effect transistors : SOLID-STATE DEVICE THEORY

It appears that they are JFETs with the gate inside the channel as opposed to next to it.

Any more info available ?

 

[VladimirK]

Speaking about a use of SITs, I would remind, that the working point they are usually implemented at is chracterized by relatively low Drain-Source resistance. What this can be useful for?
As in a case with NFB and THD, it is better to insure maximum linearity before the closing of NFB loop.
Same is true with the output impedance Zout. Usually it is lowered by the action of NFB. But, again, much better to insure quite low Zout, before the closing the NFB loop, and this could be done just due to triode-like output characteristics of SITs.

 

[rsdio]

They behave like tube Triodes, but can do so at voltages and currents of interest to loudspeakers, removing the need for transformers, heaters, and so on.

Forgive me for asking a potentially obvious question, but I assume that tube amplifiers generally operate in the range of hundreds of volts (300 V, 700 V) where loudspeakers are happier with tens of volts (25 V, 50 V max supply). What currents are interesting for loudspeakers? (excepting the exotics like planars, electrostatics - or include them if you'd like to be comprehensive)

P.S. Looking forward to the article!

 

[kevinh]

On the Semisouth site they have a couple of devices identified as being for audio use:

SJEP120R100A & SJEP120R063A The specs seem the same as for the devices without the audio designation.

Are the devices in the 6 moons article different from these FETs?

 

[buzzforb]

Yes. The ones in the article were made for Nelson, I believe.

 

[Nelson Pass]

Interesting. Have you compared curves?

Thumb-nailing the performance, they are pretty comparable in gain and
distortion for single-ended Class A in the 10 watt category. The 2SK77
has a bigger die and more capacitance, which is both an advantage and
disadvantage, depending on what you are looking for. Certainly that is
an advantage for higher power, not an advantage for single-stage amps.

A very nice part, though.

 

[Blues]

Page 4 is up...http://www.6moons.com/audioreviews/firstwatt12/4.html...looks like another exciting FW model to look forward to!

 

[Zen Mod]

naah ......

just go to first page , and read again :

6moons audio reviews: Nelson Pass' NewWatt

 

[Blues]

Sorry can't edit anymore, this is the correct link... 6moons audio reviews: Nelson Pass' NewWatt

ZM, naah can't afford those...I'm NP so an FW product will come of it...

 

[Nelson Pass]

Thanks, I had not seen the latest.

 

[wrenchone]

I'm curious - what sort of topology was used for this part to make manifest its character? I'm not asking for the schematic, mind you - just a broad-brush description of topology, From the 6 Moons pictures, there appears to be not a whole lot going on. I would guess a simple SIT stage with active loading. I may guess wrong...

 

[Zen Mod]

Okay, you've had your initial impression so here's the poop. 10 watts/ch into 8 ohms. 47Kohm input impedance, damping factor of about 3, bandwidth is 1Hz to 300KHz. Gain is 19dB. The distortion is 2nd harmonic at about 0.5% at 1 watt and a mix of 2nd and 3rd at about 4% at 10 watts. The amplifier circuit has a single stage consisting of the SIT biased by a Jfet set up as a mu follower. There is no feedback loop and the SIT has no degeneration. It draws slightly less than 200 watts. I hope you are enjoying it! - np

there is it ....

 

[wrenchone]

I stand chastened... I went back to the 6 Moons review and bored back through it, managing to find the tidbit quoted above. Food for thought, anyway.

I'm SIT-less at present, but the simple topology mentioned could possibly be duplicated with a pair of Ixys depletion mode fets or an Ixys fet plus Semisouth 085 (or a pair of Semisouths, if you're feeling flush). It wouldn't likely have the same characteristics as Nelson's SIT amp but it would have some sort of characteristic[/U], perhaps worth looking into.... A nudge, perhaps, to turn loose of papa's coat-tails for a bit (admitting freely that they're rewarding coat-tails) and forge off on one's own tangent. Even though the results might not be as quite as nice, they may still be pretty good.