Virtual ground in power amp applications - Page 4 - diyAudio
Go Back   Home > Forums > Amplifiers > Power Supplies

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 9th October 2010, 07:02 AM   #31
godfrey is offline godfrey  South Africa
diyAudio Member
 
Join Date: Nov 2009
Location: Cape Town
Quote:
Originally Posted by aspringv View Post
what do you think of the quad schematic? I'm curious to get feedback on efficacy as it'll be cheaper to do, vs the use of a LM3886?
It's an active circuit as Andrew said, but a really limp-wristed one, with performance even worse than the simple passive circuit shown earlier.

The goal of any of these virtual ground circuits is to keep the DC voltage on the virtual ground as close as possible to half way between the positive and negative supply voltages, even if there is some DC offset current.

Two weaknesses of the Quad circuit are:
A) It allows the voltage to drift about +-0.7V from where it should be before the active devices (the transistors) even turn on.
B) The transistors are current-limited to about 25mA. If the DC offset current is larger than that, the voltage error will quickly get large.

Below is a comparison of roughly how much voltage error to expect on the virtual ground for each of the three circuits, for different amounts of DC current:

Active circuit (post 8):
1mA DC => 1mV error
10mA DC => 10mV error
20mA DC => 20mV error
40mA DC => 40mV error

Passive circuit (post 3):
1mA DC => 50mV error
10mA DC => 500mV error
20mA DC => 1V error
40mA DC => 2V error

Quad 606:
1mA DC => 700mV error
10mA DC => 1V error
20mA DC => 1.3V error
40mA DC => 23V error !!!

The Quad circuit is still quite rational, though.
If the DC offset current is less than about 25mA (which seems to be their assumption), then the voltage on the virtual ground will be within about 1.5V of where it should be.

Points in favor of the Quad circuit are that it's fairly simple and inexpensive, and it should run cool without wasting much power.

If required, it could always be improved at the expense of some added complexity.

e.g. In the version below:
A) The diodes (and R5 and R6) have been added to give the transistors some forward bias to get rid of the +-0.7V "dead zone".
B) R3 and R4 have been reduced to increase the current handling to about 100mA.

Hmm, this is almost identical to one of the circuits you linked to in post 6, except that one didn't have current limiting resistors.
Attached Images
File Type: gif psu606.GIF (2.9 KB, 456 views)
  Reply With Quote
Old 9th October 2010, 07:47 AM   #32
diyAudio Member
 
bobodioulasso's Avatar
 
Join Date: May 2008
I have built an F5 with the Quad psu schematic. (posted in the F5 thread.)
It works well but needs a big cap across plus and minus. (or doubling the caps value)
  Reply With Quote
Old 10th October 2010, 06:49 AM   #33
diyAudio Member
 
Join Date: Sep 2010
Location: Perth, Western Australia
Default that's looking like a good solution

Cheers for that analysis of the voltage error in the various circuits - very helpful!

I reckon I'll give your solution a shot then. It seems to have the benefits of accuracy with a reasonable part count, and I've most of those bits on hand anyway (with the exception of those transistors).

Again, thanks for the help; I'll try and keep this thread updated with my progress as I go and whether the end result works out well.
  Reply With Quote
Old 10th October 2010, 07:26 AM   #34
diyAudio Member
 
Join Date: Sep 2010
Location: Perth, Western Australia
Actually, I do have a pile of small signal transistors - BC557, 559, 558, 549, 547, 548. Could I use these (or several of them in parallel?) in this circuit? I'm about to read all of their datasheets to get some understanding as to their suitability, but it's all new to me so i'd certainly be happy to hear any advice!
  Reply With Quote
Old 10th October 2010, 07:37 AM   #35
diyAudio Member
 
bobodioulasso's Avatar
 
Join Date: May 2008
Quad uses 1A transistors.
I used bd139/140 without heatsinks.
  Reply With Quote
Old 10th October 2010, 08:21 AM   #36
godfrey is offline godfrey  South Africa
diyAudio Member
 
Join Date: Nov 2009
Location: Cape Town
Quote:
Originally Posted by aspringv View Post
(with the exception of those transistors)
Nothing special about those - BD139 and BD140 just came to mind first as I've known them "since I were a lad", which isn't much of a recommendation. There's bound to be better modern parts that I'm not familiar with.

The ZTX ones used by Quad are better in terms of gain, but can't dissipate as much power.

To get some ideas, you could look at other solid-state designs and see what they're using for VAS or drivers.

According to these lists, the following look like likely candidates:
2SC3421 and 2SA1358
2SC2911 and 2SA1209
2SD669 D/AC/C and 2SB649 D/AC/C
2SC4793 and 2SA1837

I had a look at the datasheets - they all look OK, but might appreciate heatsinks.

Edit: oops, I missed a few posts there while I was scratching around.

Those small-signal transistors might be OK, if you use 2 or 3 in parallel instead of one bigger transistor BUT make sure their voltage rating is high enough. According to some datasheets I found:

BC546, BC547, BC550=good
BC548, BC549 =bad

BC556, BC557, BC560=good
BC558, BC559 =bad

If you do that, each transistor must have it's own emitter resistor though.

Last edited by godfrey; 10th October 2010 at 08:45 AM.
  Reply With Quote
Old 10th October 2010, 04:07 PM   #37
diyAudio Member
 
Join Date: Aug 2009
Location: Sacramento
Quote:
Originally Posted by aspringv View Post
OK, so I was running around getting a quote on some transformers for a F5 amp build, and I came across a transfo (625VA, 38-0, 38-0 secondaries) which was for sale for a song... so I bought it figuring it'd be good for a class AB build. The I got to thinking (always dangerous!) about my f5 build and started wondering about whether I could use the transfo I have somehow...

So the F5 wants rails of +/- 24V and with 38V secondaries I'd end up with an unloaded rail voltages of around +/- 53V.... which is almost double what I actually need... So I started wondering about voltage doubling rectification, and so arrived at the following questions.

Is there a way to (exactly) halve the voltage output after rectification?

Any suggestions?
Hello,
This is not half but is much closer. Consider a choke input power supply following your transformer. Check out this Hammond pdf http://www.hammondmfg.com/pdf/5c007.pdf . A little CRCR filtering and you are there.
DT
  Reply With Quote
Old 11th October 2010, 10:36 AM   #38
diyAudio Member
 
Join Date: Sep 2010
Location: Perth, Western Australia
Hi Godfrey,

Cheers for the answer - just as a check of the concept, I might try the BC5?? transistors first. Should I be happy about the end result, I'll give the appropriate single chip solution the go. I know DIY audio always tends to the extreme's (because we can), but I always tend to find myself at the 'bang for the buck' end of the performance curve, so I'll probably go with the BD transistors seeing as I've used them before quite happily.

Cheers for the emmittor resistor comment - I wouldn't have realized myself
  Reply With Quote
Old 15th October 2010, 06:19 AM   #39
diyAudio Member
 
Join Date: Sep 2010
Location: Perth, Western Australia
Well, I got myself all the requisite parts to go ahead and build this powersupply, so off we go. I'll try and build something over the weekend, but as it stands I have no amp to provide a load to test the end unit. Any suggestions as to how I can test what I build with a highly asymmetric, variable load to see how it goes? ...Actually, is this the right way to test it anyway? In lieu of an amp and a set of ears what is the best way to test performance here?
  Reply With Quote
Old 15th October 2010, 08:15 AM   #40
godfrey is offline godfrey  South Africa
diyAudio Member
 
Join Date: Nov 2009
Location: Cape Town
Hmm, Dunno, good question.
I suppose a good start would be to just hook up a load between the positive and negative rails to draw an amp or so of current and see what happens.

No need for anything exotic here; just scratch around the house for something that has about 50 ohms DC resistance - maybe an electric toaster or some lightbulbs wired up in parallel or .... ?
  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
virtual ground codacola Parts 6 7th March 2013 01:24 AM
Creating the 'virtual ground' in a single supply preamp/amp circuit??? DoiNk! Solid State 13 15th December 2012 12:57 AM
Virtual Ground ? Bernhard Solid State 13 4th March 2009 01:57 PM
Decoupling cap placement - virtual ground dual op amp? Triophile Solid State 1 27th March 2006 06:30 AM
virtual ground...help jsteigs Chip Amps 6 3rd February 2005 10:49 AM


New To Site? Need Help?

All times are GMT. The time now is 08:46 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