Low Drop Out Power Rail Filter - 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 20th January 2013, 02:58 AM   #1
diyAudio Member
 
Join Date: Oct 2006
Default Low Drop Out Power Rail Filter

Hey Guys,

I've been working on an active power filter for an amp I'm designing: Think Bybee Music rail.

This one is pretty simple, but offers exceptional ripple rejection with a low drop.

Using low Rds P-ch MOSFET and a pair each of 2N5551 and 2N5401, the results have me quite excited to build it up. The model I'm running has given me around 2mV total ripple with an output current draw of up to 12A peaks. The load current sources are a fixed 6A, a 3A @ 20Hz, 3A @ 1000Hz. I don't have a spice model for the IRF5210 so I substituted the Siliconix Si7469DP which seems to offer similar characteristics. The negative rail compliment uses a low Rds N-ch.

Here's a few shots of the simulation showing both the total ripple with current as well as the unfiltered and filtered voltage for comparison. Ripple has been reduced by a factor of just over 1000, or a little more than 63dB. Total dissipation in the MOSFET is around 15W with average delivered power of nearly 400W.

I honestly can't say if this has been done, but it's probably not new. I'm very pleased with the results as you can probably tell.
Attached Images
File Type: gif LDO Line Filter Sim1.gif (40.6 KB, 208 views)
File Type: gif LDO line filter sim2.gif (35.9 KB, 204 views)
File Type: gif LDO Line Filter.gif (21.7 KB, 217 views)

Last edited by pete_schumacher; 20th January 2013 at 03:03 AM.
  Reply With Quote
Old 20th January 2013, 03:26 AM   #2
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Blog Entries: 22
2mv of ripple with 12A peaks is very impressive! and it looks like you are only sacrificing about 2V to do it too.

Tony.
__________________
Any intelligence I may appear to have is purely artificial!
Some of my photos
  Reply With Quote
Old 20th January 2013, 03:48 AM   #3
diyAudio Member
 
Join Date: Oct 2006
Quote:
Originally Posted by wintermute View Post
2mv of ripple with 12A peaks is very impressive! and it looks like you are only sacrificing about 2V to do it too.

Tony.
2V was the target. If you remove 1 of the diodes feeding R4 and R9, you can get even less overall drop. But that does result in an increase in ripple when a full 12A peak 20Hz modulation is employed. In reality, such a demand on an amp would be exceedingly rare, so a single diode would probably be more than adequate for music signal loads into 8 Ohm.

Here's a shot simulating full power 20Hz into an 8 Ohm load and a single diode feeding R4.
Attached Images
File Type: gif LDO Line Filter Sim3 8 Ohm 20Hz.gif (24.0 KB, 207 views)
File Type: gif LDO Line Filter Sim3 rail ripple.gif (20.9 KB, 190 views)

Last edited by pete_schumacher; 20th January 2013 at 03:58 AM.
  Reply With Quote
Old 20th January 2013, 06:41 AM   #4
diyAudio Member
 
Join Date: Sep 2010
Location: Melbourne
That could be quite a handy circuit if it performs as simulated. I have used the the LT1085 before for similar purposes but it only goes to 7A
  Reply With Quote
Old 20th January 2013, 08:30 AM   #5
diyAudio Member
 
Join Date: Oct 2006
Quote:
Originally Posted by metalsculptor View Post
That could be quite a handy circuit if it performs as simulated. I have used the the LT1085 before for similar purposes but it only goes to 7A
The IRF3210 and IRF3415 can handle 40A with a 200W power rating (given adequate heat sinking of course). This set up should be able to clean up virtually any power amp supply you could think of. Parts cost is very reasonable.
  Reply With Quote
Old 20th January 2013, 10:51 AM   #6
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
JLH Ripple Eater?
__________________
regards Andrew T.
  Reply With Quote
Old 20th January 2013, 06:16 PM   #7
diyAudio Member
 
Join Date: Oct 2006
Quote:
Originally Posted by AndrewT View Post
JLH Ripple Eater?
The ripple eater is a shunt type regulator, and from what I have read so far, is limited to about 5 Amp. I'm not sure about its suitability for something as drastic as taking care of the ripple voltage of an unregulated PSU like that of an amplifier, or the large variation in current that would need to be supplied to an amplifier under load.

Also, based on the values of the feedback components, I'm not so sure that the Ripple Eater would have as high a rejection for rail variations as this one.

Here's a shot of the transfer function of output ripple/input ripple. This does a serious job of attenuation over the frequency range where current draw is highest via music power (below 300Hz) which creates the largest ripple levels on the main filter caps.

Adding a shunt regulator to this one for high frequencies may be beneficial though. I may have to break out LTSpice and do just that.
Attached Images
File Type: gif LDO Line Filter Xfer Function.gif (6.3 KB, 38 views)
  Reply With Quote
Old 20th January 2013, 11:30 PM   #8
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Quote:
Originally Posted by pete_schumacher View Post
I honestly can't say if this has been done, but it's probably not new. I'm very pleased with the results as you can probably tell.
Interesting design! Somewhat similar to AMB's a22 The σ22 Regulated Power Supply but you have some new good stuff in there. You've eliminated the zener in the voltage reference (I'm assuming the 1N4148s are less noisy than the zener, but haven't checked specs), have the mosfets turned the other way for a LDO rather than source-follower, and your positive and negative halves are completely independent. I especially like that last point since that issue makes troubleshooting the a22 a little harder - when one rail goes in a certain way they both go. Probably no pressing need for the mosfets to be configured as source-follower since the reference voltage for the error amp is sampled after the mosfets in both designs.

But I do like AMB's idea of feeding the voltage reference parts with the separate cap multiplier to create clean(er) power from the main rail, at the expense of a couple of volts. Especially in the case of really large currents moving through the main rails (and associated IR rail ripple), which both designs can handle. I've suggested to AMB before of connecting the cap multiplier(s) to the main rail through an inductor to help nuke any any frequency noise that otherwise would zoom right through the cap mults. Also beefing up the cap mult transistors to BD139/140 and putting a reverse diode across the BE junction given the caps on either side. I also like how AMB has actual current sources feeding the LTPs, rather than the (16k) resistors, for higher impedance looking back in. Those current source diodes are getting scare and pricey though. Probably better at this point to use a depletion mosfet or BJT based ccs.

Last edited by agdr; 20th January 2013 at 11:56 PM.
  Reply With Quote
Old 21st January 2013, 12:22 AM   #9
diyAudio Member
 
Join Date: Oct 2006
agdr.

Thanks for pointing out the Sigma22. Very similar circuit topology for sure. But unlike the Sigma22, this unit is not regulated. The diodes act as a fixed voltage drop to allow the reference voltage to track the average voltage on the main filter caps. This allows the MOSFET to maintain a fairly constant 2V drop, minimizing power lost in the device. The way the circuit is designed, it is basically acting as a 2nd order lowpass filter with unity gain. As the input level changes slowly, the output will track. But higher frequency variation, like rectifier ripple, is filtered out. It's sort of an active version of a CRC supply with much higher rejection capability in the frequency range where most power will be drawn from the supply (below 400Hz).

I'm still evaluating its performance as load current frequency exceeds 1000Hz. Seems fine so far as long as the current levels are not too high, but I have noticed at high power levels above 2000Hz that output ripple is reaching into the 100mV range. And if there is too much capacitance on the output, it seems to make it more difficult to keep ripple levels very low. I'm looking to improve the output impedance over a broader frequency range.
  Reply With Quote
Old 21st January 2013, 02:09 PM   #10
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Ah, I see, hence the missing zener. Well that is good circuit idea.
  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
Low pass filter for simple sub on single rail? tubee Chip Amps 9 21st February 2008 03:28 AM
Filter Cap Directly from Rail to Rail, excluding Center Tap Workhorse Power Supplies 8 29th August 2007 01:33 PM
Need low power high beta transistors - please drop me a few part numbers... peufeu Solid State 37 3rd November 2006 01:22 AM
Rail voltage drop in preamp mmerig Solid State 4 25th May 2006 05:12 AM
Rail drop test darkfenriz Power Supplies 8 24th February 2006 02:38 PM


New To Site? Need Help?

All times are GMT. The time now is 09:17 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