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 July 2009, 07:57 AM   #1
diyAudio Member
 
audiopip's Avatar
 
Join Date: Nov 2006
Location: Melbourne, AU
Default New Cherry NDFL amp

Hi Folks, first post from a relative newbie and former lurker, so here goes...hope you can read the circuit....its long, bear with me.

I am actually a switch mode PSU designer who moved into analogue and mixed signal semiconductor sales... it pays more!My background is Linear Tech, Burr Brown, Unitrode, TI but I have always had a hobby of audio dating back to when I was 11 years old...a long time ago now. Now I am selling motors..more on that another day.

I first experimented with nested feedback by putting a 400W MOSFET amp inside the feedback loop of a OPA627. Fixes the DC drift just great...and actually works quite well. Why...the sales director wanted 'more bass'. I reckon he just needed to clean out his ears...

But it got me interested in nested feedback, and I read the Cherry papers. And wished I could remember more of the stuff I learned at college.....that's what sales does to you.

The original objective was, using the Cherry paper (with his permission for hobbist use incidentally), duplicate his results and make a judgement as to the subjective audio quality.
As time went on the project slowly mutated into a 'make it as best you can' then apply NDFL feedback.

The input stage was elaborated to a CFP type in an attempt to achieve best linearity with the minimum additional complication.

The Rush stage was retained but with the inclusion of R21 to limit the maximum stage gain to something sensible...I am a pom after all so good to have a UK circuit in a Monash Univ. paper!

The VAS transistor was chosen as a BF469, to preserve as much HF performance as possible, rather than use a MJE. I appreciate other much better (and easier to obtain) transistors exist now.

Likewise the current source for the VAS stage was elaborated to a cascode. Later this was found very handy as it was modified to dethump the amp at turn on and diodes were added to the supply rails to prevent positive feedback thumping the amp at turn off.

The output stage was eleaborated into a CFP driver with common emitter output. The object here was to present minimal loading to the VAS, maxmimised voltage swing and allow multiple pairs of output devices to be driven, if a high power version was required.

The prototype used Rod Elliots P03A as the driver...thanks Rod!

The output and driver transistors are the newish On Semi MJE15035 and 36 with MJL4XXX outputs.

The thermal tracking is split into two, one Vbe multiplier tracking the driver drift, the other tracking only the output temperature mounted smack on the face of the output transistor. This appears to work very well.

In the Cherry paper he suggests that including the output stage withn the CDOM loop is acceptable, it is certainly not using this more elaborate and possibly slower output stage, hence the position of C9 is conventional except that it is now part of a nested system.

The choice of transistors is not totally trivial in this design BTW, quite a lot of work was undertaken to prove the CFP stages were stable.

What does it sound like...well I think pretty good, no transistor 'edge' and very good resolution. It has powerful
but uncoloured bass too.

BTW some years ago I built the Self amps and verified his results..yes he's right THD+N circa <0.0005 in a 80KHz BW (APS1 noise floor stuff) but the amps sound diabolical on music, still have the pcbs if anyone interested. For anyone trying this, wiring up the amps and the test equipment is more critical than the choice of transistors...back to the Cherry amp..

Spice analysis (thanks LTC, great free tool!) shows the overall loop gain to be very similar to an amplifier with 2 pole CDOM feedback, very high gain at low frequency, followed by a 2 pole roll off reducing to a single pole at unity. But this is achieved in a more controlled fashion, so I suspect, but cannot prove, lower overall distortion.

What do you guys out there in DIY land think? or

And thanks to Hugh Dean for tea and sympathy...sympathy for trying to build this thing in the first place!
Attached Images
File Type: jpg cherry_1_1.jpg (98.8 KB, 1818 views)
  Reply With Quote
Old 26th July 2009, 08:32 AM   #2
jaycee is offline jaycee  United Kingdom
diyAudio Member
 
Join Date: Apr 2004
Location: Norwich, UK
You got a pretty good mentor there, Hugh's amps seem great to me

It is well worth looking at Fairchild's range of transistors. They have some excellent parts - in fact they cover power amps from pretty much small signal to power output now!

KSC1845/KSA992 - nice small signal parts
KSC3503/KSA1381 - make excellent VAS and predrivers
KSC2690/KSA1220 - similar to above but more current, good for small stage drivers/headphone outputs

FJL4215/FJL4315 - second sourced 2SA1943/2SC5200 of excellent quality, also available in TO3P and TO220 versions.

Best of all, you can order small quantities direct, so guaranteed no fakes

The drawing's a bit too small to make anything out I'm afraid. Better to use a PNG image with a low number of colours.
  Reply With Quote
Old 26th July 2009, 03:08 PM   #3
diyAudio Member
 
audiopip's Avatar
 
Join Date: Nov 2006
Location: Melbourne, AU
Hi,

I kind of found Hugh by accident and we have had lots of dicussions about various aspects of audio design. I want Hugh to listen to the Cherry circuit, and hoping that might happen soon

Yes the transistors are not the best although the output devices and the CFP driver seem to be well suited and stable.

I'll have another go with the schematic!

Thanks!
  Reply With Quote
Old 26th July 2009, 03:08 PM   #4
diyAudio Member
 
audiopip's Avatar
 
Join Date: Nov 2006
Location: Melbourne, AU
Default Cherry amp photo

Here is a photo of the prototype. Hope this is better than the circuit..

The little pile of bits is the de thump circuit...now hidden underneath.

The pcb was my first Protel effort, and I made the mistake of using the library TO92 outline, so the pads are impossibly small. A colleague etched the pcb from a ink jet tranparency master, and kindly drilled it for me. The layout was done specifically to be single sided, there are 3 links, all short.

Its all a real 'garage' job so not perhaps as professional looking as I would like.
Attached Images
File Type: jpg ndfl.jpg (96.5 KB, 1626 views)
  Reply With Quote
Old 26th July 2009, 04:36 PM   #5
diyAudio Member
 
tiefbassuebertr's Avatar
 
Join Date: Jul 2005
Location: D-55629 Schwarzerden
this is very interesting. NDFL is still the best way for all kinds of not hard running amps (i. e. ClassAB) - so I think. Can you upload your simulation results? For example see my pdf versions about this one
schematic variation and sim results of Pass "X" series US pat 5376899
AD797 discrete clone - comparison of seven variations
  Reply With Quote
Old 27th July 2009, 09:28 AM   #6
diyAudio Member
 
audiopip's Avatar
 
Join Date: Nov 2006
Location: Melbourne, AU
Hi,

Yes I'll try to post the plots. Might have to remake the Spice model as things have moved on a bit since I built the amp and the orginal correct model has become a little ...ahem...corrupt!

Should have saved it before I started changing it...!
  Reply With Quote
Old 1st March 2010, 06:34 AM   #7
diyAudio Member
 
audiopip's Avatar
 
Join Date: Nov 2006
Location: Melbourne, AU
A sequel to all of this, rather late yet again...as so common with low distortion amplifiers IMO, this thing does not sound very good, flat as a pancake sound stage etc etc. Hugh agrees, so I am looking at other possibilities. Its a shame because the thing is actually stable, very low noise, etc etc, but lacks that certain something.....bet you have heard all this before many times.
  Reply With Quote
Old 1st March 2010, 09:41 AM   #8
diyAudio Member
 
tiefbassuebertr's Avatar
 
Join Date: Jul 2005
Location: D-55629 Schwarzerden
Quote:
Originally Posted by audiopip View Post
A sequel to all of this, rather late yet again...as so common with low distortion amplifiers IMO, this thing does not sound very good, flat as a pancake sound stage etc etc. Hugh agrees, so I am looking at other possibilities. Its a shame because the thing is actually stable, very low noise, etc etc, but lacks that certain something.....bet you have heard all this before many times.
Unfortunately, you don't mention a little about the boundary conditions like e. g. power supply, the used power capacity, the loudspeakers and the load impedance.
My experience is by soundcheck of power amplifiers, that the cause in the case of a bad sound are mostly due to hugh errors regarded power supplies, GND management and similar thinks.
For the perception of differences that are clearly due to the circuit topology I need perfect conditions and identical conditions concerning the environmental stuff (not very easy to realize).

.............this thing does not sound very good, flat as a pancake sound stage etc etc..........

This sounds like bad or too small supply capacitors, too small transformer, too low load impedance and/or bad GND/mass management.

in the no longer available German magazine "ELRAD" (from the same publisher of the popular computer magazine "Ct") this topology has been described in detail (February/March 1984 and an update to the old schematic in September 1988 and the ELRAD special edition number 2)
you can order copies there:
heise online - Home
here an ebay auction about the "ELRAD 2" special edition
http://cgi.ebay.com.au/ws/eBayISAPI....fvi%3D1&_rdc=1
Elrad,Band:2*Magazin für Elektronik*Bauanleitungen*TOP1 bei eBay.de: Computer Technik (endet 05.02.10 11:10:49 MEZ)
here some additional advices about NDFL (go to page 51 post #505 and #506)
Spice simulation
and
What is nested feedback, how it realy works and some examples...
Attached Files
File Type: pdf ELRAD 2 special Edition.pdf (64.3 KB, 374 views)

Last edited by tiefbassuebertr; 1st March 2010 at 09:55 AM.
  Reply With Quote
Old 1st March 2010, 10:12 AM   #9
AKSA is offline AKSA  Australia
diyAudio Member
 
Join Date: Sep 2001
Location: Melbourne, Australia
Welcome to the forum, Jon - long may you be here!

Not sure about tief's 'hugh errors'. Is this something I should be concerned about in my designs? Must check my ground regime....... (or speak to my Mother)

Hugh
__________________
Aspen Amplifiers P/L (Australia)
www.aksaonline.com
  Reply With Quote
Old 1st March 2010, 10:57 AM   #10
godfrey is offline godfrey  South Africa
diyAudio Member
 
Join Date: Nov 2009
Location: Cape Town
Quote:
Originally Posted by audiopip View Post
What does it sound like...well I think pretty good, no transistor 'edge' and very good resolution. It has powerful but uncoloured bass too.
Quote:
Originally Posted by audiopip View Post
...IMO, this thing does not sound very good, flat as a pancake sound stage etc etc.

Has something changed or are you just being more critical now?

btw I seem to remember that Dr Cherry's amp (published in ETI in May 1983) had a rather diabolical protection circuit that looked like it might cripple performance into real speaker loads.

The low frequency group delay correction was pretty neat, though.

Regards - Godfrey
  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
Cherry Wood Loudspeaker? Dennis5587 Multi-Way 12 5th February 2008 02:28 AM
NDFL amplifiers kenlow Solid State 4 25th April 2007 03:15 AM
Staining Oak into a dark red / cherry color? akunec Multi-Way 10 1st November 2005 06:34 PM
Adire Kit 281's cherry veneer aaird Swap Meet 0 15th December 2003 08:54 PM
Black Cherry Duck-Twacy Pass Labs 5 26th November 2003 08:06 PM


New To Site? Need Help?

All times are GMT. The time now is 08:37 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