Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 26th December 2005, 09:08 AM   #1
Bensen is online now Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Default Fully complementary FET-in FET-out design

Hi all,

For the last four months I've been simulating alot in LTspice, my knowledge about ampdesign has grown alot. But I'm still an amateur, and probably will stay also.

I wanted to desing an amp with MOSFET output devices and JFET input devices. I've used always MOSFET's in my previous diyamps, and I'm very pleased with them. Some people made me curious about using JFET's as input.

Because JFET diff pairs don't have much gain, I had to use a differential VAS also. The total closed loop gain is about 13000.
I used cascoded VAS'es. Because the voltage drop is big with a cascoded VAS and even bigger with a diff cascoded VAS (+R48 and R49), I use a +-80V regulated PS for the front stage.

I want to use the inverting input of the diff pair also for a balanced input signal. When I'm not using this, R33 has to be connected to ground.

Some numbers:
*current throught each leg of diff input: 1.7mA
*current throught each leg of diff VAS: 22.5mA
*current throught EF (MJE15030/31): 37.4mA
*idle current throught MOSFET's: +-140mA

*Slew Rate is +-100
*Closed loop Bandwith= +-110kHz

Please critizise the design, when you see error's of things that can be beter, tell me.

Greetz
Ben
Attached Images
File Type: jpg amp 26_12_2005__1.jpg (89.5 KB, 2320 views)
  Reply With Quote
Old 26th December 2005, 09:12 AM   #2
Bensen is online now Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
In this picture you can see the output stage, with 4 MOSFET's in parallel. I will use the 2SK1058 and 2SJ62, because they are the newer versions of the 2SK135 and 2SJ50 wich I use in another amp of mine.
Attached Images
File Type: jpg amp_26_12_2005__2.jpg (89.8 KB, 1904 views)
  Reply With Quote
Old 26th December 2005, 04:34 PM   #3
Bensen is online now Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Default PCB split

Is there any problem when I would split the PCB in three parts, namely;

1) the total front stage, power supply +-80V
2) 4x 2SK1058 with MJE15030
3) 4x 2SJ162 with MJE15031

I can't make it out of on big PCB. But I can assemble these three parts close to eachother with very short wires.

Is there any problem with osscilation or something else maybe.
This could save me alot of money when I would do it wrong I guess.

Cheers
Ben
  Reply With Quote
Old 26th December 2005, 05:37 PM   #4
diyAudio Member
 
Join Date: Aug 2003
Location: Prague,Czech Republic
Hi Ben, rail voltage at VAS must be little bit higher, cca 82-83 V... Serial resistors at gates should have lower walue ( 39 at P and 47 at N ) - amp stay stabil and SR will be much higher... Give some RC compensation from VAS to minus input and give out this one from output... And design of PCB will be very important...
  Reply With Quote
Old 26th December 2005, 05:44 PM   #5
diyAudio Member
 
Join Date: Aug 2003
Location: Prague,Czech Republic
And one important notice : rail voltage at VAS must " flow " on output's rail voltage, so all PSs must be connected at series...
  Reply With Quote
Old 26th December 2005, 05:47 PM   #6
diyAudio Member
 
Join Date: Sep 2005
Hello!

If I'm not mistaken, I think you may be able to delete transistors Q1, Q11, Q9, and Q10.


==> You're taking the differential output of your input stage into another differential input of the VAS stage, but one side of the differential VAS stage is doing nothing (it has its outputs grounded at the collectors of Q1 and Q9).

Unless I'm missing something, I think you could simplify the VAS stage to save on a little cost and board space for the PCB.
  Reply With Quote
Old 26th December 2005, 06:04 PM   #7
lineup is offline lineup  Sweden
diyAudio Member
 
lineup's Avatar
 
Join Date: Dec 2005
Location: the north
Looks like that, at first sight.

But if you look at power supply - to the left -
you see VAS is a differential stage.
With a common resistor from supply.
__________________
lineup
  Reply With Quote
Old 26th December 2005, 06:06 PM   #8
Bensen is online now Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Upupa Epops,

Do you mean to lower the gate resistors from the MOSFET's? (Not JFET's). That's the first time I see these low values for gate stoppers in lateral mosfet design's?? Are you sure? In my last project I blew up 10 x2sk1530/2SJ201 due to, to low gate stoppers -> oscillation.

I've experimented alot with this NFB network. Taken the feedback from VAS will lead to a more stabil amp, but THD will rise.
Do you mean by "and give out this one from output", that I have to omit this feedback from the output and take only feedback from the VAS? Have maybe a link to a good amp that is using this kind of NFB?

I'm aware that the voltage for the front stage must be floating round the same ground. But you say that they have to be in series, in my design I'll do this in serie. But is this really necessary?

rtarbell,
Im' sorry to say, but yes you are missing something. Take a good look at the shematic, R48 and R49 can be replaced by two current cources, than you will have the same as in the differential input. R48 and R49 are serving as current source. Without these transistor's, the closed loop gain would be +-3000 instead of +-13700.
  Reply With Quote
Old 26th December 2005, 06:49 PM   #9
diyAudio Member
 
Join Date: Aug 2003
Location: Prague,Czech Republic
Yes, I mean gate stopers at power mosfets - if you give them ASAP to gate, amp will be stabil. And I don't mean all feedback, but only compensation... And floating supplies are neccessary - if you make VAS's voltage regulated, by full load, when voltage at output rails will drop, voltage at VAS will be too high and amp will be not stabil... If you can, I will send you my schematic and you can take there inspiration...send me a mail, all is to large...
  Reply With Quote
Old 26th December 2005, 08:51 PM   #10
Bensen is online now Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Thanks for the mail.
You did me realise that the capacitor (C8, like by Leach http://users.ece.gatech.edu/~mleach/lowtim/fdbk.html)
to ground, better not can be used with mosfet's. With a mosfet design it is possible that the mosfet's are oscillating (or starting to), and with this capacitor the error correction won't be able to "see" this oscillation.
Am I correct?
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Another quasi-complementary design quasi Solid State 631 4th January 2013 09:05 PM
Quad 303 : Upgrade output stage to be fully complementary askbojesen Solid State 32 18th February 2012 05:12 PM


New To Site? Need Help?

All times are GMT. The time now is 07:56 PM.


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