LM317 cleanup - 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 February 2011, 10:52 PM   #1
vaioble is offline vaioble  United Kingdom
diyAudio Member
 
Join Date: Feb 2011
Default LM317 cleanup

Hello, I'm currently building a psu from recycled components ( except for electrolytic caps- been there, done that, never again ! ) and I'm trying to tweak the best out of an LM317 and LM337 pair. Noise being an issue with these 3 term regs I found a reference in a post "Squeezebox power supply " august 2006 .


Better yet, if you can replace R2 with somethig like a zener diode*, you'll get lower noise all the way down to DC - because the dynamic impedance of the zener will much lower than the required value of R2 (just 1-5ohms insteead of 1-2Kohms, say). So it mimics the action of a truly enormous cap at C4 without the potential for damage, or cost. Use a zener in series with a normal diode (anode to 0v), and you get nearly-perfect temperature compensation for free, too.


Found more on acoustica.org at the following link Using 3-pin regulators off-piste: part 1

Attached is my intended application of these mods to standard 3 term reg config. The second resistor in the usual config is replaced in this case with 2 zeners in paralell thus lowering the dynamic impedance further. The diodes in series with the zeners should aid thermal stability and my reason for using pairs is that Vout= 1.25 + Vz + 2*Vf where Vf is the forward voltage of the 1n4148 and zener Vz is 12 volts. Ok it's not exactly 15V (14.65 volt -ish) but I'm glad to hit the ballpark if I can do so quietly and I could twiddle the choice of series diodes for a more convenient Vf.

Now assuming all the previous is correct, I come to my second question- what of the second resistor, usually labeled R1 between the output and adjustment terminals? What value to use ?

My interpretation assumes Iadj << Iz Iadj the adjustment current out of the LM317 and Iz the zener current with 1.25V as the voltage drop across this resistor so R=1.25/(2*Iz -Iadj) becomes R=1.25/2*Iz that's about 125ohms

All the above goes likewise for the LM337 of course.

Suggestions or comments appreciated, a tested schematic with these mods deeply so.
  Reply With Quote
Old 9th February 2011, 11:06 PM   #2
SY is offline SY  United States
diyAudio Moderator
 
SY's Avatar
 
Join Date: Oct 2002
Location: Chicagoland
Blog Entries: 1
Use a big cap on the output (plus protection diode). Use a big cap on the adjust pin (ditto). Sink a lot of current (100mA or more). Bob Pease and Errol Dietz did some very detailed analysis on the effects of these caps and regulator current on noise and source impedance- worth hunting those papers up.
__________________
You might be screaming "No, no, no" and all they hear is "Who wants cake?" Let me tell you something: They all do. They all want cake.- Wilford Brimley
  Reply With Quote
Old 9th February 2011, 11:48 PM   #3
diyAudio Member
 
abraxalito's Avatar
 
Join Date: Sep 2007
Location: Hangzhou - Marco Polo's 'most beautiful city'. 700yrs is a long time though...
Blog Entries: 101
Send a message via MSN to abraxalito Send a message via Yahoo to abraxalito Send a message via Skype™ to abraxalito
If you're concerned about noise you'll also note that the supply rejection drops off rather repidly at HF. An input RC filter helps fix up this weakness. Running at high current as SY suggests gets the output impedance down - I think the output impedance is quoted at 0.5A in the datasheet (from memory).

In general, noise isn't just a function of having clean regulators, layout and decoupling can make or break the design.
__________________
No matter if we meanwhile surrender every value for which we stand, we must strive to cajole the majority into imagining itself on our side - Everett Dean Martin
  Reply With Quote
Old 10th February 2011, 08:07 AM   #4
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Blog Entries: 22
You could think about implementing Fred's suggested addition in this post Improving the LM3x7 regulator circuit

I've been getting back to my sim to finalize a bom. Things I noticed that made a BIG difference were the RC filter pre reg as abrax suggests (I've gone a bit overkill and have CRCRC), and running a decent margin on the voltage (5V difference in to out seems to be good and in fact is the voltage difference that all tests in the data sheet have been done). This is of course only based on a sim and the LM317 model I have may be not so good. The early sims I did showed a 3 fold reduction in ripple with the transistor compared to a "standard" circuit. The latest sims I've done I've been having difficulty believing as they are too good. 1uV ripple with around 7V in-out margin... and 210mA draw.

Values (on Fred's schematic) for R1 and R2 1K and 100 ohms respectively, R3 10K and R4 will require some experimentation to get your desired voltage, I couldn't seem to get a formula that worked (although at the time I didn't realise my LM317 model's ref voltage was out)... a 50K pot at R4 is what I am planning. R5 Fred suggests 0.5 ohms, I'm going to try getting away with a bit less, between 0.2 and 0.3 ohms. The transistor in my sim is a BC560C the C is important as you need the high Hfe.

Tony.
__________________
Any intelligence I may appear to have is purely artificial!
Some of my photos
  Reply With Quote
Old 10th February 2011, 08:23 AM   #5
Salas is online now Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
Can you post the .asc?
  Reply With Quote
Old 10th February 2011, 09:45 AM   #6
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Blog Entries: 22
Hi Salas, I was planning to start a thread, but was holding off until I'd actually built the thing This is the latest incarnation. I will probably remove the 0.1uF bypass across the output capacitor after some output impedance tests I ran last night. I now have what I think might be reasonable inductance values in the caps, The output caps are Elna Silmic II's which I can't get any info on the self inductance for so taking a stab in the dark for 30nH The original impedance sims I did I think I had unrealistically low something stupid like 3nH...

The README.txt has instructions on what extras need to be put where to get the sim to run

The LM317D.sub and asy were the original LM317 model I had used, it is limited to 700mA and I have hacked it to get the adjust voltage right... The LM317.sub is a parametric model I got recently so perhaps is the reason why the figures seem too good to be true.. I haven't tried the LM317D (D for discrete) since I moved it to the new name...

The 50 ohm load resistors were just to simulate the load that I expect to have. I was planning on having 50mA on the reg itself (so the final circuit would have 200 ohms there) and the load is around 160MA per rail (which will actually be split a few ways with 1 ohms and 1000uF again between each part of the circuit it is driving. (so the output is planned to be CRC as well).

The 50K pots should be ok for the voltage I'm planning (around 10V) I haven't checked what range they will give...

I know it doesn't have an LM337 but the principal should be the same

Tony.
__________________
Any intelligence I may appear to have is purely artificial!
Some of my photos

Last edited by wintermute; 10th February 2011 at 09:50 AM. Reason: forgot the attachment!
  Reply With Quote
Old 10th February 2011, 10:01 AM   #7
diyAudio Member
 
abraxalito's Avatar
 
Join Date: Sep 2007
Location: Hangzhou - Marco Polo's 'most beautiful city'. 700yrs is a long time though...
Blog Entries: 101
Send a message via MSN to abraxalito Send a message via Yahoo to abraxalito Send a message via Skype™ to abraxalito
Quote:
Originally Posted by wintermute View Post
The output caps are Elna Silmic II's which I can't get any info on the self inductance for so taking a stab in the dark for 30nH The original impedance sims I did I think I had unrealistically low something stupid like 3nH...
3nH isn't outrageously optimistic for a 2.5mm lead spacing electrolytic. That's just for the cap though, as soon as you add any length of track to it then your 30nH figure becomes much more realistic.

Quote:
The LM317D.sub and asy were the original LM317 model I had used, it is limited to 700mA and I have hacked it to get the adjust voltage right...
I also found when building a discrete model of the TL431 the bandgap reference didn't come out at the right voltage.
__________________
No matter if we meanwhile surrender every value for which we stand, we must strive to cajole the majority into imagining itself on our side - Everett Dean Martin
  Reply With Quote
Old 10th February 2011, 10:06 AM   #8
Salas is online now Salas  Greece
diyAudio Chief Moderator
 
Salas's Avatar
 
Join Date: Oct 2002
Location: Athens-Greece
Hey Tony, thanks for the info. Let us know when prototyped.
  Reply With Quote
Old 10th February 2011, 10:10 AM   #9
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Winter,
why have you series connected the caps in the smoothing?
__________________
regards Andrew T.
  Reply With Quote
Old 10th February 2011, 10:23 AM   #10
just another
diyAudio Moderator
 
wintermute's Avatar
 
Join Date: Aug 2003
Location: Sydney
Blog Entries: 22
That is a really good question Andrew and something I don't have any reasonable answer for! I think stupidity is the answer, and until you mentioned it I just hadn't realised!

DOH I was effectively halving the value of the caps... glad I posted before I built! I will revise the schematic, the ripple pre reg go's from 2.4mV to 0.4mV only using one lot of caps. I can't believe I made such a silly mistake... and none of the 125 downloads from my blog have commented, I hope I haven't made a whole lot of people think it is a good idea sigh...

Tony.

edit: here's a fixed version Salas, would you mind deleting the other one so that it doesn't propogate any more than necessary, in the interests of not spreading dumb circuit
Attached Files
File Type: zip YARPS.zip (6.0 KB, 258 views)
__________________
Any intelligence I may appear to have is purely artificial!
Some of my photos

Last edited by wintermute; 10th February 2011 at 10:42 AM.
  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
Accuphase C-270 - Cleanup and Capacitor Ident xerocomm Solid State 1 15th April 2010 07:14 PM
Condo cleanup Cheap/Free stuff Issaquah, WA velmeran42 Swap Meet 5 3rd February 2008 11:07 PM
Lm317 jlm Power Supplies 9 2nd February 2006 07:31 PM
Free OScope! Spring cleanup jackinnj Swap Meet 0 26th February 2004 02:39 PM


New To Site? Need Help?

All times are GMT. The time now is 03:45 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