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 7th June 2011, 09:52 PM   #11
! is offline !  United States
diyAudio Member
 
Join Date: Jun 2005
Location: Midwest
^ It's not just a matter of proper voltage, regulated power is quite a bit cleaner.

It seems that the configuration that uses all these parts mentioned the best with minimal extra parts and least risk to the regulator or AC mains input fuse would be,

Transformer - > 0.7R -> Bridge Rectifier -> 2 x 100K caps -> Regulator -> (roughly) 10uF film or tantalum cap -OR- 150uF+ low ESR electrolytic -> Amp Board

... but frankly, I don't feel there is any need for one, let alone two 100K uF caps in this circuit. Given the high 1KVA transformer rating relative to the 2.5A load current, one 3,000 uF cap by itself to replace those two should provide sufficient filtering because it is regulated afterwards, but if that is done you might set an additional 1V drop across the regulator instead of the bare minimum.
  Reply With Quote
Old 8th June 2011, 07:25 PM   #12
diyAudio Member
 
Jason Hubbard's Avatar
 
Join Date: Apr 2001
Location: Poole, England
Tomchr - There is a soft start already on the primary side of the transformer (basically a timer relay that shorts out a 50 ohm power resistor after 1 to 10 seconds (depending on the delay that is set)).

agdr - Thanks for the RCRC idea - i will give it a try.

This regulator circuit is already up and running (though not for long so i don't know what the reliability will be like).

I was interested in any improvements that could be made sonically (i.e. i understand that having a few more volts across a regulator than the minimum amount can have benefits). I have read that power supply regulation in an amplifier is a good thing but you need as many microfarads after the regulator as before it, hence the big cap on the output that everyone seems to advocate moving/removing. I have followed the LM317 thread with some interest and wondered what improvements could be had at higher current levels. And if it is actually worth it for a power amplifier anyway.
  Reply With Quote
Old 8th June 2011, 08:07 PM   #13
diyAudio Member
 
bogdan_borko's Avatar
 
Join Date: Nov 2005
Location: serbia, zajecar
I tried LT1083 on my Aleph-like class A amplifier few years ago (regulator had 4 volts voltage drop) and CRC and CLC sounded better...but it`s very simple so you could easily try both and see for yourself what sounds better.
__________________
Svakog dana u svakom pogledu sve manje nazadujem.
  Reply With Quote
Old 8th June 2011, 08:52 PM   #14
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Since I did the schematic in LT Spice I went ahead and ran a simulation for fun.

The node labels are at the top of the plot and the scale for the green current trace is at the right.

The blue is the capacitor charging voltage spike with the green showing a charging pulse of around 7 amps through the rectifiers. The rectifiers in the model are 25A units.

I picked an input VAC of 28V peak (I've added 0.5R series resistance to the voltage source, plus the two rectifier voltage drops give the 23.5V peak on the blue trace), which results in the output voltage of the second RC stage, presented to the LT1083, of about 16.5VDC. With the R3-R4 resistors setting the regulator at 15VDC out, that gives (16.5V - 15.0V) = 1.5V across the regulator here, near the 1.2V minimum the datasheet graphs show for 2.5A. The test load is 6R to draw the 2.5A.

You can see the ripple difference between the output of the first RC stage (red) at about 18.3VDC and the output of the second (auqa) at 16.5VDC. If I were to remove that second RC stage I would expect to see practically no difference in the regulated (magenta) output, with the regulator taking care of that slight amount of incoming ripple. By hey, this is diy audio and you already have the second capacitor, so might as well.

Messing around with the values a bit, the resultant 15V output still has some minor fluctuations in the sim if you zoom in a ways. Increasing the input voltage source slightly to 30V peak to allow more drop across the regulator fixes that, as I would expect, as does upping the output capacitor to 3300uF from 150uF. But this is just a sim of course - with the real part you might not even measure an output ripple difference at 1.5V across the part vs. the 3.5V.
Attached Images
File Type: png rcrc LT1083 2.png (27.5 KB, 210 views)
File Type: png rcrc LT1083 2 plot.png (12.3 KB, 204 views)

Last edited by agdr; 8th June 2011 at 09:04 PM.
  Reply With Quote
Old 8th June 2011, 08:54 PM   #15
! is offline !  United States
diyAudio Member
 
Join Date: Jun 2005
Location: Midwest
People tend to think adding more parts always sounds better, but if you want to know for sure get out the scope and read power line ripple at the base of the output transistors, that is the only place it matters. Let me take that back for a moment, other areas matter too, but you'd have those isolated from the large current swings on the rail to the output transistors (or integrated amp IC), not just a star ground since the +- rails are effected.

What you will find is that decoupling and filtering on the amp board itself swamps any other factor after the regulator subcircuit. There is a big difference between putting lots of capacitance on a regulated PSU versus an unregulated.

You most definitely do not need as many uF after a regulator as before it. Before it, you have a large time interval between each AC mains cycle, ie 50/60Hz bridge rectified to 100/120Hz. The capacitance needs be sufficient enough that the rail voltage stays above the minimum forward voltage drop across the regulator. The higher your input voltage is compared to the output voltage, the less capacitance you need to make that happen.

A regulator is far faster than 120Hz cycle time. For practical purposes, and especially considering trace and wire impedance from it to a large capacitor bank then on to the amp board, it is well documented and specified by the regulator manufacturers that you really don't need more than a very low impedance ~ 10-20uF capacitor after it. What then matters is impedance to the output transistor base, keeping your ESR between that regulator and the output transistor base as low as reasonably possible.

If by adding larger capacitors, you can get them closer to the output transistor base pin than the regulation circuit is, this is good, an improvement. If you still have an equivalent length/width/gauge trace or wire from the capacitor bank to the output transistor pin as you would if you left the capacitor bank out, it defeats the purpose.

What having a larger capacitance bank after a regulator can do, is spread out the duty cycle. Instead of dealing with shorter peak currents it deals with longer duration lower currents.

Now a word about what sounds better. I have been describing electrical function, not how something sounds. Maybe someone prefers less stable voltage, a slower more linear drop in voltage with momentary peaks... or maybe they don't. While people usually claim they *want* the opposite, I cannot argue with doing something and preferring the change in sound because that's what it is all about, the sound.

Last edited by !; 8th June 2011 at 09:03 PM.
  Reply With Quote
Old 8th June 2011, 09:46 PM   #16
! is offline !  United States
diyAudio Member
 
Join Date: Jun 2005
Location: Midwest
Quote:
Originally Posted by ! View Post
What having a larger capacitance bank after a regulator can do, is spread out the duty cycle. Instead of dealing with shorter peak currents it deals with longer duration lower currents.
I forgot to mention the larger inrush current upon turn on with the larger capacitance bank.
  Reply With Quote
Old 8th June 2011, 11:20 PM   #17
BZed is offline BZed  United States
diyAudio Member
 
Join Date: Jun 2009
Have you also considered the AC line voltage and how much it changes over time? You need to make sure that the voltage headroom for the regulator is still there at low line and high load or you can loose regulation.

As for the difference in sound of reg v CRY I think that is mostly dodo with clipping onset and how the amp/ps deal with clipping,
  Reply With Quote
Old 9th June 2011, 12:59 AM   #18
agdr is offline agdr  United States
diyAudio Member
 
agdr's Avatar
 
Join Date: Sep 2010
Here is the same thing with just a CRC filter. The lead resistor in that previous RCRC is really pretty useless other than reducing initial peak current through the rectifiers. Just results in a higher input voltage requirement with no real ripple improvement after the regulator. Here the voltage source peak is (just) 24VAC resulting in the same 16.5VDC going into the regulator. So there you go - you can use just the 2 big caps the the 0.7R you already have on hand.

The previous plot was steady state, of course. I've included the initial part of the plot here showing the 40A or so through the diodes while the (initially) uncharged capacitors charge up. In the previous RCRC circuit R1 reduces the initial current to about 20A.

Green is the output of the rectifiers going into the filter at about 18VDC. Blue is the output of the CRC filter going into the regulator. Red is the regulated output and aqua is the current through D3 - current scale is on the right side. The current plot in the previous post was through R1.
Attached Images
File Type: png crc LT1083.png (28.7 KB, 184 views)
File Type: png crc LT1083 plot initial.png (13.4 KB, 177 views)
File Type: png crc LT1083 steady state.png (7.8 KB, 149 views)

Last edited by agdr; 9th June 2011 at 01:25 AM.
  Reply With Quote
Old 10th June 2011, 01:17 PM   #19
diyAudio Member
 
Jason Hubbard's Avatar
 
Join Date: Apr 2001
Location: Poole, England
BZed - I had thought about line voltage variation - I would not want to have an LDO regulator dropping only 1.5 volts if 1.2 is the minimum it can work with - I live in the countryside on a lightly loaded local transformer - the line voltage is generally high, but I can't count on it always being that way, especially if a local farmer plugs in something that pulls a lot of current and causes some violtage drop. I would prefer to see about 3 volts across the regulator (at 2.5 amps the regulator will only be disipating 7.5 watts, so no big deal on power losses especially bearing in mind it's a class A power amp anyhow).

! - Thanks for your interest in the thread - I concurr about getting some caps as close as possible to the output devices, but for me this isn't something I can easily do with the PCB layout i'm stuck with, though i'll see what i can do to upsize the 220uf's with something like 1000uf instead.

agdr - The rectifier diodes i'm using are rated 40 amps continuous when mounted on a suitable heatsink, mine are simply bolted to the chassis with some thermal grease, but given the low current draw i don't think the turnon current with only one resistor is going to hurt much, especially bearing in mind that there is a soft start on the primary sode fo the toroid. And thankyou very much for the sims - very informative!

Last edited by Jason Hubbard; 10th June 2011 at 01:21 PM.
  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
Externally programmed and adjustable regulated outboard amp power supply exeric Power Supplies 6 7th June 2010 06:14 PM
help me i need a linear regulated power supply for my amp udhay_cit Power Supplies 1 13th March 2010 01:10 PM
Choke regulated power supply for S/S amp Matt Rowland Solid State 31 22nd March 2009 01:58 PM
Regulated Power Supply for power amp blu_line Solid State 13 30th January 2006 06:42 AM
regulated power supply adason Swap Meet 1 11th December 2005 06:32 PM


New To Site? Need Help?

All times are GMT. The time now is 04:28 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