2nd amp from scratch: Low NFB- FET front end - BJT OPS - diyAudio
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 24th December 2013, 01:50 PM   #1
Bensen is offline Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Default 2nd amp from scratch: Low NFB- FET front end - BJT OPS

Dear Diy'ers,

I'm feeling confident enough about the current state of this amplifier design to show it to you guys.

Please have a look into the attached document. I've tried to put all information into there including links to other threads. To encourage some interest I've attached already some pictures

This amp will be build, only it will take some time. I also have to put some time into my family offcourse (daugther of 1year old).
My previous amp kept me busy for about 1.5year.

If you guys see some opportunity's to make it better or if there are remarks, please let me know!
Attached Images
File Type: jpg status 20131224.JPG (214.9 KB, 743 views)
File Type: jpg SAM_1855.JPG (175.4 KB, 730 views)
Attached Files
File Type: pdf LowNFB - FET input vas and driver - BJE output - 20131224.pdf (273.4 KB, 124 views)
  Reply With Quote
Old 24th December 2013, 01:53 PM   #2
diyAudio Member
 
jacco vermeulen's Avatar
 
Join Date: Oct 2004
Location: At the sea front, Rotterdam or Curaçao
Send a message via Yahoo to jacco vermeulen
Wowza, some power supply (another trespa lover on the block)
__________________
The buck stops Here
  Reply With Quote
Old 24th December 2013, 02:09 PM   #3
Bigun is offline Bigun  Canada
diyAudio Member
 
Bigun's Avatar
 
Join Date: Jan 2009
Blog Entries: 2
Gorgeous amp !

This looks like a very nice design, great use of FETs and BJTs and a really well executed build with top notch parts.
__________________
"The test of the machine is the satisfaction it gives you. There isn't any other test. If the machine produces tranquility it's right. If it disturbs you it's wrong until either the machine or your mind is changed." Robert M Pirsig.
  Reply With Quote
Old 25th December 2013, 10:25 PM   #4
diyAudio Member
 
LineSource's Avatar
 
Join Date: Sep 2003
Location: SiliconValley
Very well engineered power supply.

It looks like the 2X59V driver regulator circuits are under the protection board.

It looks like the output voltage supply is a C-R-C topology using fuses for the R elements. An AC slow start circuit before the main transformer will remove repeated surge currents which can gradually degrade these short length fuse R elements. Long serpentine PC board traces that can distribute surge currents plus heat are often used as R's. A heatsink for the power rectifiers is often included, just to be cool.


If you "voice" your amp mainly from Spice simulation harmonic distortion data, then high hfe=500, low noise(nf=0.6db) bipolar transistors like the BC550C, BC560C, and VAS transistors like KSA1381 and KSA1186 can sim 20x-50x lower THD than MOSFET circuits using the irf610, irf9610, irf710. Most modern bipolar amps can LTSPICE sim to 0.00005% THD at 200 watts, and maintain this low THD over diverse loads through near-optimal Vbias thermal tracking circuits. A read (re-read) of Bob Cordell's book thread covers most modern bipolar circuits. Cascodes, current mirrors, tracking current sources, 3T outputs, dual slope Vbe bias, etc... Cheap to try a few LTSPICE sims before you commit to your MOSFET driver design. Bipolar+feedback philosophy= "It is better to significantly beat down THD than to worry if the TINY 3rd/odd harmonics are higher than the TINY 2nd/even harmonics."


If you "voice" your amp mainly from listening tests, then a few talented designers like Arye support your all-FET front-end preference. FET second order non-linear distortion vs. BJT odd order. I bet that Arye caliber companies perform extensive parametric testing of each MOSFET because of the wide production variation common in MOSFETS. The datasheets for irf610, irf9610, irf710 show a 2V variation in Vt, and only give a minimum gs spec. A Monte Carlo analysis quickly shows parameter variation effects.

Your 3pF Miller VAS compensation seems small, and worth a few stability sims.

A more advanced Vbe multiplier with two gain transistors as discussed by Jim Hagerman and others has shown good value. Have you run multi-temp simulations?
2SC2922 temp={t}
.step param t list 30C 80C


YOURS THD 20KHz 1Wrms = 0.000038%
BIPOLAR THD: 20KHz 1Wrms = 0.000002% 30C output stage
BIPOLAR THD: 20KHz 1Wrms = 0.000003% 80C output stage

YOURS THD: 20KHz 10Wrms = 0.00013%
BIPOLAR THD: 20KHz 10Wrms = 0.000007% 30C output stage
Attached Images
File Type: jpg Bipolar 1W 20KHz FFT.JPG (120.6 KB, 587 views)

Last edited by LineSource; 25th December 2013 at 10:52 PM.
  Reply With Quote
Old 26th December 2013, 01:00 AM   #5
diyAudio Member
 
LineSource's Avatar
 
Join Date: Sep 2003
Location: SiliconValley
Fig1.. CRC power supply with serpentine PCB trace R.
Fig2.. Vbe multipliers showing value of 2 gain transistors.
Attached Images
File Type: jpg CRC power.jpg (76.7 KB, 547 views)
File Type: png VbeMultiplier_DCSwee.png (96.8 KB, 547 views)
  Reply With Quote
Old 26th December 2013, 01:31 AM   #6
diyAudio Member
 
ostripper's Avatar
 
Join Date: May 2008
Location: Albany , NY (smallbany)
Where did you get those trafo's !!!

I want !

OS
__________________
Mongrel website , always current and updated :
http://www.fidelityforce.com/ostripper
  Reply With Quote
Old 26th December 2013, 01:58 AM   #7
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
Line Source, what sort of R can one expect from those traces? I'd expect <1.0 ohm, perhaps into the 100s of milliohms??

So, what is the benefit? Seems like minimal isolation between the cap stages.
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- [...2SJ74 Toshiba bogus asian parts - beware! ]
  Reply With Quote
Old 26th December 2013, 04:55 AM   #8
diyAudio Member
 
Join Date: Feb 2007
Location: Tashkent
Quote:
Originally Posted by LineSource View Post
Fig2.. Vbe multipliers showing value of 2 gain transistors.
Very interesting comparison of various "zener substitute" versions.
It would be very interesting, and many I guess would be thankful, for adding the Fig2 by transient responce simulations, when CS current jumps from 5 to 15mA, for instance. And you might exclude additional caps from the circuits, just interesting how transistors do themselves.
  Reply With Quote
Old 26th December 2013, 06:55 AM   #9
diyAudio Member
 
LineSource's Avatar
 
Join Date: Sep 2003
Location: SiliconValley
Quote:
Originally Posted by bear View Post
Line Source, what sort of R can one expect from those traces?
Hi Bear,

CRC is especially effective when the amp is operating in class A

1oz copper, 0.2inch_wide * 20inch long PCB trace ~0.05 ohms

One CRC Example.
60,000uF R=0.01 60,000uF CRC
11amps into 4ohms generates 0.56V ripple

R = 0.01 ohms generates 0.56V ripple
R = 0.1 ohms reduces to 0.230V ripple
R = 0.2 ohms reduces to 0.115V ripple
R = 0.5 ohms reduces to 0.054V ripple

VOLTAGE DROP DATA:
50.94V - 50.38V 0.01ohm 0.56V ripple
49.77V - 49.55V 0.1ohm 0.23V ripple
48.625V - 48.51V 0.2ohm 0.115V ripple
45.546V - 45.492V 0.5ohm 0.054V ripple

Last edited by LineSource; 26th December 2013 at 07:02 AM.
  Reply With Quote
Old 26th December 2013, 07:22 PM   #10
Bensen is offline Bensen  Belgium
diyAudio Member
 
Bensen's Avatar
 
Join Date: May 2003
Location: Belgium, Limburg
Quote:
Originally Posted by LineSource View Post
Very well engineered power supply.

It looks like the 2X59V driver regulator circuits are under the protection board.

It looks like the output voltage supply is a C-R-C topology using fuses for the R elements. An AC slow start circuit before the main transformer will remove repeated surge currents which can gradually degrade these short length fuse R elements. Long serpentine PC board traces that can distribute surge currents plus heat are often used as R's. A heatsink for the power rectifiers is often included, just to be cool.


If you "voice" your amp mainly from Spice simulation harmonic distortion data, then high hfe=500, low noise(nf=0.6db) bipolar transistors like the BC550C, BC560C, and VAS transistors like KSA1381 and KSA1186 can sim 20x-50x lower THD than MOSFET circuits using the irf610, irf9610, irf710. Most modern bipolar amps can LTSPICE sim to 0.00005% THD at 200 watts, and maintain this low THD over diverse loads through near-optimal Vbias thermal tracking circuits. A read (re-read) of Bob Cordell's book thread covers most modern bipolar circuits. Cascodes, current mirrors, tracking current sources, 3T outputs, dual slope Vbe bias, etc... Cheap to try a few LTSPICE sims before you commit to your MOSFET driver design. Bipolar+feedback philosophy= "It is better to significantly beat down THD than to worry if the TINY 3rd/odd harmonics are higher than the TINY 2nd/even harmonics."


If you "voice" your amp mainly from listening tests, then a few talented designers like Arye support your all-FET front-end preference. FET second order non-linear distortion vs. BJT odd order. I bet that Arye caliber companies perform extensive parametric testing of each MOSFET because of the wide production variation common in MOSFETS. The datasheets for irf610, irf9610, irf710 show a 2V variation in Vt, and only give a minimum gs spec. A Monte Carlo analysis quickly shows parameter variation effects.

Your 3pF Miller VAS compensation seems small, and worth a few stability sims.

A more advanced Vbe multiplier with two gain transistors as discussed by Jim Hagerman and others has shown good value. Have you run multi-temp simulations?
2SC2922 temp={t}
.step param t list 30C 80C


YOURS THD 20KHz 1Wrms = 0.000038%
BIPOLAR THD: 20KHz 1Wrms = 0.000002% 30C output stage
BIPOLAR THD: 20KHz 1Wrms = 0.000003% 80C output stage

YOURS THD: 20KHz 10Wrms = 0.00013%
BIPOLAR THD: 20KHz 10Wrms = 0.000007% 30C output stage


Linesource,

Underneed the softstart board is the DC-blocker, have a look here:
High performance dual mono PSU

The PSU board contains 4 diode bridges, 2 for the 36Vac secondaries of the 500VA R-core, 2 for the 9Vac secondaries of the 30VA R-core.

This enclosure will be connected to the amp using neutrik 8 pole Speakon connectors. Inside the amp enclosure there will be something like 20000µf added to each rail. The fuse and wires from PSU enlcosure to amp eclosure will create the "R" from the CRC.

I'm not chasing super low distortion figures. My goal is to have a nice harmonic spectrum using as less as possible feedback.
In my previous amp, Fet drivers did give much more detail to the music compared with 3 different pairs of bjt's as drivers.

The simulation tells me that I can leave the Miller caps out, I just put them into the LTspice file to make sure I do not forget room for these on the PCB whenever Miller caps are required. The VAS Fet's have a rather high input capacitance which takes the Miller cap function over.

Good tip about the temp simulation, I've not done this sofar.

Thanks for the info!
  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
J-FET & MOSFET VAS with current NFB VladimirK Solid State 30 11th April 2011 12:29 PM
Lower volume by switching from FET to BJT input op-amp? wwenze Chip Amps 5 15th December 2010 06:29 AM
Advantage of a FET-BJT CCS vs BJT only Bitrex Tubes / Valves 2 10th June 2009 03:01 AM
Front end for sub amp mashaffer Solid State 9 22nd July 2005 03:35 AM
Single device (BJT/FET?) amp? RichardJones Solid State 2 19th April 2003 10:06 AM


New To Site? Need Help?

All times are GMT. The time now is 10:22 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright İ1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2