Op Amp Rail Splitter Question - Page 2 - diyAudio
Go Back   Home > Forums > Source & Line > Analog Line Level

Analog Line Level Preamplifiers , Passive Pre-amps, Crossovers, etc.

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 15th March 2013, 04:27 PM   #11
diyAudio Member
 
Join Date: May 2011
Location: Silicon Valley
I put the "Push-Pull Shunt Regulator" (Audio Amateur, 1/88, pp. 26-29) circuit schematic into LTSPICE, and ran some simulations. To my surprise, the circuit appears not to regulate at all; output voltage was displaced from the midrail point by approximately eleven volts (!). I've studied the schematic for quite a long time, and I confess that I cannot find any trace of a negative feedback mechanism which (a) detects when the output is non-midrail, and then (b) pulls it in the other direction (towards the exact midpoint between the rails). Indeed there isn't a single node anywhere in the circuit, whose voltage is at the exact midpoint between the rails.

Maybe this is merely a drafting error; perhaps R2 and R3 were intended to connect to the upper rail and lower rail, respectively. If so then their junction would have been a midrail reference point, and the opamp's feedback loop would have stabilized the output at the midrail point. However the accompanying PCB design matches the schematic shown here, with R2 and R3 not connected to the rails. The previous design from 5 years before, shown in Figure 2 of the TAA article, did connect these resistors to the rails. Hmmm.

In simulation I loaded the regulator with a 0-to-100mA sinewave at 2 kHz, from Top to Midrail; and with a 0-to-80mA sinewave at 7 kHz, from Midrail to Bot. I've attached a 2-panel plot of these currents, below. The "regulated" voltage at the output is shown in a second plot. Its amplitude is ~ 60 millivolts peak to peak, implying an output impedance of about 0.6 ohms. There is also a bit of oscillatory ringing at approx 72 kHz, which dies away after two or 3 cycles. Most importantly of all, the output voltage is displaced about 10.9 volts above the midrail point.
Attached Images
File Type: png Schematic.png (21.8 KB, 110 views)
File Type: png Load_Currents.png (31.4 KB, 107 views)
File Type: png Output_Voltage.png (20.1 KB, 103 views)
  Reply With Quote
Old 15th March 2013, 04:44 PM   #12
jcx is offline jcx  United States
diyAudio Member
 
Join Date: Feb 2003
Location: ..
not much chance if you bootstrap everything - really have to have a "external" reference V somewhere in the system - like a R divider from the the rails you're trying to split

bootstrapping an op amp's supply pins has unobvious stability consequence - and common Spice models don't get either gain reference or ps pin currents right

nor is 5534 unity gain stable

Last edited by jcx; 15th March 2013 at 04:52 PM.
  Reply With Quote
Old 15th March 2013, 04:48 PM   #13
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
Another point besides moving location of R's is your input is done wrong. Input must be floating -- not grounded. Like a 50v power source for +/- 24vdc output... ungrounded as explained in the article.

[Afterwards, try adding some series R or a series regulated input-floating... sim what happens.]

Thx-RNmarsh

Last edited by RNMarsh; 15th March 2013 at 05:16 PM.
  Reply With Quote
Old 15th March 2013, 05:27 PM   #14
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
As shown it was a rail splitter but with the R's changed to rails it regulates as well. That didnt get shown and was an oversight in writing/editing.
Thanks for doing a SIM on it and when it is working, it can be optimized. But I still want to get up here the article. There were no sim software at the time of developing concepts and it was very much a DIY thing. No computers either (except a Commodore and other toy compuetrs in their infancy). The derivation of the concepts is useful to others for expansion - though a lot of that has occured already by others who have taken the ball and run with it since then.

Thx-RNMarsh

Last edited by RNMarsh; 15th March 2013 at 05:49 PM.
  Reply With Quote
Old 15th March 2013, 06:10 PM   #15
diyAudio Member
 
Join Date: Jun 2011
wakibaki's virtual ground
  Reply With Quote
Old 15th March 2013, 06:26 PM   #16
diyAudio Member
 
Join Date: May 2011
Location: Silicon Valley
The original article had R2 and R3 incorrectly connected, in the same way I've shown; a snippet of the original schematic is below. The PCB layout faithfully duplicated the error, too.

I just found out that a guy named John Adelsbach noticed this same problem in 1988, and published a corrected circuit in TAA 4/88 (pp. 46-47). He concluded the input resistors were wrong in the original article. At the end of his article, an italicized passage appears: Mr. Marsh comments: The 5k resistors should go to rails as J. Adelsbach has indicated. ...

It doesn't take long to put together a simulation of the (incorrect) original rail splitter circuit, or to make modifications & test them out. Try it yourself, you'll enjoy it! The only slightly unusual element in this circuit is the 4.7 mA constant current diode 1N5314. I modeled it the easy way, using the primitive elements already present in LTSPICE: a JFET and a resistor. Voila, a 4.7 mA constant current diode with its own schematic symbol. Simulated Zout at 4.7 mA was half a megohm; plenty good enough.
Attached Images
File Type: png In_LTSPICE.png (21.8 KB, 93 views)
File Type: jpg In_TAA_Magazine.jpg (72.0 KB, 91 views)
File Type: png subcircuit.png (5.6 KB, 21 views)
  Reply With Quote
Old 15th March 2013, 06:40 PM   #17
diyAudio Member
 
Join Date: Jun 2011
4.7 mA?

There are sims in the link I posted. Read 'em and weep...
  Reply With Quote
Old 15th March 2013, 07:12 PM   #18
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
It works even better with a 3-term pre-reg in series as a source to the push-pull shunt reg. As I indicated in the article, I had also tried the idea on 1/2 of a stereo amp to power the other half. Worked great for power amps as well!

Thx-RNMarsh
  Reply With Quote
Old 18th March 2013, 06:06 PM   #19
RNMarsh is offline RNMarsh  United States
diyAudio Member
 
RNMarsh's Avatar
 
Join Date: May 2012
Location: 2457 Cascade Trail; Cool, CA. 95614
Quote:
Originally Posted by RNMarsh View Post
It works even better with a 3-term pre-reg in series as a source to the push-pull shunt reg. As I indicated in the article, I had also tried the idea on 1/2 of a stereo amp to power the other half. Worked great for power amps as well!

Thx-RNMarsh
Every 25 years or so it's best to go back over the basics and sometimes the origins of what becomes mainstream.... for newbees and refreshers for the old geezers:

Note Fig2 is the correct form to use re. R1,2 for R2,3..

Enjoy!
Attached Images
File Type: jpg P-P PS 1.jpg (924.2 KB, 41 views)
File Type: jpg P-P PS 2.jpg (865.3 KB, 44 views)
File Type: jpg P-P PS 3.jpg (720.3 KB, 41 views)

Last edited by RNMarsh; 18th March 2013 at 06:08 PM.
  Reply With Quote
Old 19th March 2013, 12:10 AM   #20
diyAudio Member
 
Sonic Wonder's Avatar
 
Join Date: Mar 2013
Location: Sunnyvale, California
I've solved the problem. Please stop emailing me. Thank you very much!
  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 / Rail Splitter using only LM317 JediRemi Power Supplies 4 6th April 2011 03:52 PM
Wanted: Op-amp with single supply, 32V or more, and rail-rail capability akis Power Supplies 3 13th November 2010 01:53 PM
cmoy sijosae rail splitter cachupito Headphone Systems 10 20th May 2010 03:42 PM
simple HV rail splitter/floating ground - would this work? Rodeodave Power Supplies 3 20th July 2008 04:12 PM
amplifier and rail-splitter forming a filter? Dominique Solid State 12 2nd May 2006 06:26 AM


New To Site? Need Help?

All times are GMT. The time now is 04:52 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