diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Pass Labs (http://www.diyaudio.com/forums/pass-labs/)
-   -   SIT amp idea (http://www.diyaudio.com/forums/pass-labs/220963-sit-amp-idea.html)

VladimirK 4th October 2012 06:35 PM

SIT amp idea
 
1 Attachment(s)
I just have managed to get some russian SITs, KP801A and KP801G. The first is 60W 3A (at Vgs=0 and Vds=10V), and the second is 100W 6A part. I am not sure that they will do OK just at single common source stage, with my big PMC EB1i speakers. As an inevitable compromise, I am thinking about two stages and some GNFB. How would you comment and what could be advised to the enclosed schematics?
For getting the indicated 8mA+8mA currents at diff stage, I plan to use matched 4pcs of 2N5462 p-ch jFETs, with Idss around 6mA and 4mA operation point. Using of just 4mA+4mA at diff stage I guess could be not sufficient, because of the SIT Ciss and some gate leakage current.

juma 4th October 2012 08:10 PM

Looks like an honest SE amp :up:
One thing - you could save yourself some hustle by using something like IRF9610 or BD140 or even LM317 as a CCS for LTP...

Zen Mod 4th October 2012 08:28 PM

as Juma sez .....

my fave for that is ring of two bjt (say BC546) plus additional series resistor to eat some voltage

you can use J271 for LTP , in case that you insist on western parts

LED CCS with cheap BF245C ; eh , memory ........ around 510R for ~5mA

VladimirK 5th October 2012 04:25 AM

Thanks, Juma and ZenMod. Some people are very careful to PSRR issues, and said that I shall have no PSRR at the -16V PS rail. Because of this, better to use standard source follower at the output. Majority of people tend to bring the PSRR on the top of everything. I am, on the contrary, find from listening experience, that even with high PSRR, the PS e-caps adds substantially to the sound.
I tend to reject this proposal, since the low intrinsic Rds resistance of SITs allows to benefit from the common source output stage, and moreover amplification of this stage adds essentially to GNFB. Due to the same reason (low Rds), PSRR of the source follower variant will be not as good as with MOSFETs, and this will spoil SIT specific advantages.
With the common source output stage, possible cure for -16V PSRR could be just following the output cap design receipes.

juma 5th October 2012 07:20 AM

Quote:

Originally Posted by VladimirK (Post 3190481)
Thanks, Juma and ZenMod. Some people are very careful to PSRR issues, and said that I shall have no PSRR at the -16V PS rail. Because of this, better to use standard source follower at the output. Majority of people tend to bring the PSRR on the top of everything. I am, on the contrary, find from listening experience, that even with high PSRR, the PS e-caps adds substantially to the sound.
I tend to reject this proposal, since the low intrinsic Rds resistance of SITs allows to benefit from the common source output stage, and moreover amplification of this stage adds essentially to GNFB. Due to the same reason (low Rds), PSRR of the source follower variant will be not as good as with MOSFETs, and this will spoil SIT specific advantages.
With the common source output stage, possible cure for -16V PSRR could be just following the output cap design receipes.

I tend to agree with your conclusions about topology choice. Also, (IMHO) it's OK to take care about PS but it's better to handle the ripple and other nasty artefacts in the PSU (by using capacitance multiplier, various regulators...) than to compromise the amp design due to PSRR worries...

VladimirK 5th October 2012 07:31 AM

Quote:

Originally Posted by juma (Post 3190572)
I tend to agree with your conclusions about topology choice. Also, (IMHO) it's OK to take care about PS but it's better to handle the ripple and other nasty artefacts in the PSU (by using capacitance multiplier, various regulators...) than to compromise the amp design due to PSRR worries...

Thanks, Juma, for your supporting words. I shall start probably next week SIT measurements, in order to be sure with optimum voltages and currents through the SITs.

Salas 5th October 2012 07:37 AM

Looks like another good natural sounding amp idea from the ever skillful Vladimir. Good luck!

Michael Rothacher 5th October 2012 11:53 AM

I like it Vladimir. Go for it. :)

VladimirK 5th October 2012 12:29 PM

Thanks, Salas and Michael. For the T8 and T9 BJTs, I just have got nice p-n-p metal case transistors, 2N2905A. Hope some of them have hFE more than 200.

VladimirK 24th November 2012 11:41 AM

1 Attachment(s)
At the end, I have to confess, that I have not managed to convince myself to go ahead with the schematics indicated at the beginning.
I decided to implement SE version, that is well investigated by me, and I am sure in it. The SE schematics with SITs at the output is the following:


All times are GMT. The time now is 07:10 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2