Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Chip Amps Amplifiers based on integrated circuits

Critique my PCB! Yet another parallel LM3886
Critique my PCB! Yet another parallel LM3886
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 2018, 10:57 AM   #1
abza is offline abza  South Africa
diyAudio Member
 
abza's Avatar
 
Join Date: May 2014
Question Critique my PCB! Yet another parallel LM3886

Howdy all! Right, after a couple of successful hard-wired, kit and breadboard LM3886 builds I'm starting to put together the LM3886 boards as detailed here. Essentially, 2 DRV134s driving 4 sets of paralleled LM3886es. So 8 LM3886es in total for the stereo amp.

Attached is my first stab at the parallel PCB. This is inspired by a bunch of different PA100 builds from in and around the internet, as well as @tomchr's Taming the LM3886 pages for component values.
  • The schematic shows only one of the parallel pairs - the second LM3886 is identical, replace the "A"s in the designators for "Bs".
  • Power (as per @tomchr's recommendation) will be +/-28V
  • I've thumbsucked the dimensions of my inductors - LA1 and LB1 - I'll be winding some today, but I doubt the dimensions will be far off.

Notes:
  • Top (red) pours: The two red planes are GND pours, separating the two chips and meeting at the GND spade terminal (X2).
  • Top (red) traces: These are the majority of the signal tracks. In addition, CA7 and CB7 get their own GND traces back to the X2 GND terminal.
  • Bottom (blue) pours: 3 planes, for V+ along the left, V- along the right and a cross-shaped GND plane for the input and output grounds, meeting at a small star point between C2 and R2
  • Bottom (red) traces: Only a few traces here for the output (X5).

At this point, I think the bottom layer is working pretty well, but the top layer seems needlessly complex. My temptation is to just flood it with a ground pour, minus the cross-shaped input ground.

Any comments/feedback/suggestions/advice for a fellow traveler?
Attached Images
File Type: png 1-Schematic.png (59.4 KB, 447 views)
File Type: jpg 2-PCB No Pours.JPG (159.0 KB, 445 views)
File Type: jpg 3-PCB Top.JPG (162.5 KB, 437 views)
File Type: jpg 4-PCB Bottom.JPG (160.1 KB, 432 views)
File Type: jpg 5-PCB Both.JPG (182.0 KB, 420 views)
File Type: jpg 6-Render 1.JPG (68.0 KB, 208 views)
File Type: jpg 7-Render 2.JPG (44.1 KB, 187 views)
  Reply With Quote
Old 7th June 2018, 11:12 AM   #2
Nisbeth is offline Nisbeth  Denmark
diyAudio Member
 
Nisbeth's Avatar
 
Join Date: Jan 2001
Location: Denmark
Critique my PCB! Yet another parallel LM3886
Many layouts wind the output inductors around the output resistors - that might give you a little more space here (especially if you position the resistors at an angle)?

Just looking at the layout I don't understand why the CAx and CBx caps aren't oriented the same way between the two halves of the board. Admittedly it probably doesn't matter much, but it triggers my OCD a bit

Also you might be able to shorten a couple of traces here and there by:
- turning CA7 and CB7 90 degrees (or even CA7 180 degrees)
- turning R1, R3 and C1 90 degrees
- swapping Rx4 and Rx6 around (possibly)
__________________
DIY projects blog
  Reply With Quote
Old 7th June 2018, 11:21 AM   #3
abza is offline abza  South Africa
diyAudio Member
 
abza's Avatar
 
Join Date: May 2014
Quote:
Originally Posted by Nisbeth View Post
Many layouts wind the output inductors around the output resistors - that might give you a little more space here (especially if you position the resistors at an angle)?
Ooh, interesting - will see if this can be done!

Quote:
Originally Posted by Nisbeth View Post
Just looking at the layout I don't understand why the CAx and CBx caps aren't oriented the same way between the two halves of the board. Admittedly it probably doesn't matter much, but it triggers my OCD a bit
Ha ha, fair enough - will see what I can do

Quote:
Originally Posted by Nisbeth View Post
Also you might be able to shorten a couple of traces here and there by:
- turning CA7 and CB7 90 degrees (or even CA7 180 degrees)
- turning R1, R3 and C1 90 degrees
- swapping Rx4 and Rx6 around (possibly)
Well spotted. Done!

Thanks very much!
  Reply With Quote
Old 7th June 2018, 12:05 PM   #4
asuslover is online now asuslover  Romania
diyAudio Member
 
Join Date: Aug 2016
Location: USA
Critique my PCB! Yet another parallel LM3886
I would not use a pair of R/L for each chip, just one for both, or for all of the chips.
I would add 1% 0.1-0.3ohms/3-5 watts output resistor for each chip .
Servo is optional.
  Reply With Quote
Old 7th June 2018, 06:58 PM   #5
Nisbeth is offline Nisbeth  Denmark
diyAudio Member
 
Nisbeth's Avatar
 
Join Date: Jan 2001
Location: Denmark
Critique my PCB! Yet another parallel LM3886
If your "out" trace is just the thin blue one, then it should probably be made thicker (looks like there is space for that).
__________________
DIY projects blog
  Reply With Quote
Old 7th June 2018, 07:24 PM   #6
Mark Whitney is offline Mark Whitney  Netherlands
diyAudio Member
 
Mark Whitney's Avatar
 
Join Date: Feb 2013
Location: Netherlands
A good way of evaluating a layout is to look at the current loops.

These two loops have the current return path through X2.
CP1-CB1 return CB1-X2-CP1
CP2-CA5 return CA5-X2-CP2
__________________
Regards Mark.
  Reply With Quote
Old 7th June 2018, 08:55 PM   #7
tomchr is offline tomchr  Canada
diyAudio Member
 
tomchr's Avatar
 
Join Date: Feb 2009
Location: Calgary
I suggest reading Douglas Self's amplifier design book - or his article series from Wireless World published in the mid 1990ies, which formed the basis for the book. Your supply routing is about the worst you can do for rail-induced distortion.
Your VEE connection is super skinny. You'll be pushing 7A+ through that...

The curved traces do nothing for you and in one case, you've created an acid trap (T intersection under R1).

You'll need ballast resistors as well.

Tom
__________________
MOD686: Up to 240W/8Ω, <0.00025% THD. MOD86: 40W/8Ω, 0.000054% THD. HP-1: 3W/20Ω, -130dBc THD, 128dB DNR.
Neurochrome : : Audio - www.neurochrome.com - Engineering : : Done : : Right
  Reply With Quote
Old 8th June 2018, 07:31 AM   #8
abza is offline abza  South Africa
diyAudio Member
 
abza's Avatar
 
Join Date: May 2014
Thanks for the feedback everyone!

Quote:
Originally Posted by asuslover View Post
I would not use a pair of R/L for each chip, just one for both, or for all of the chips.
I would add 1% 0.1-0.3ohms/3-5 watts output resistor for each chip .
Servo is optional.
Yeah, in the interests of saving space on the board I'm tempted to do the @AndrewT approach of putting the R/L's at the speaker terminals, off-board.

...which means I'll also have room for the 1% output resistors! Thanks!

Quote:
Originally Posted by Nisbeth View Post
If your "out" trace is just the thin blue one, then it should probably be made thicker (looks like there is space for that).
Agreed - the fattening shall begin! @tomchr raises serious concerns about my grounding layout (as suspected) so I'll redo that and then see how best to fatten up the output.

Quote:
Originally Posted by tomchr View Post
I suggest reading Douglas Self's amplifier design book - or his article series from Wireless World published in the mid 1990ies, which formed the basis for the book.
Thanks for the suggestion! I read the one on Small Signals, will try grab a copy of the Power Amplifier Design book.

Quote:
Originally Posted by tomchr View Post
Your supply routing is about the worst you can do for rail-induced distortion.
Ha ha, I appreciate the brutal honesty. I have a lot to learn!

Quote:
Originally Posted by tomchr View Post
Your VEE connection is super skinny. You'll be pushing 7A+ through that...
Would it be worthwhile running say, V+ on the top and V- on the bottom? That way, they won't be competing for space to the LM3886 ICs, and can be as fat as needed? The centre of the PCB could then be used on both top and bottom for various grounds.

Quote:
Originally Posted by tomchr View Post
You'll need ballast resistors as well.
Gotcha!
  Reply With Quote
Old 8th June 2018, 07:39 AM   #9
abza is offline abza  South Africa
diyAudio Member
 
abza's Avatar
 
Join Date: May 2014
...so perhaps something like the attached, where both power rails get the same chunky PCB area, the V+ pour on the top layer and the V- pour on the bottom?
Attached Images
File Type: png Power pours.png (19.7 KB, 98 views)
  Reply With Quote
Old 8th June 2018, 11:18 AM   #10
00940 is offline 00940  Belgium
diyAudio Member
 
00940's Avatar
 
Join Date: Sep 2003
Location: Liège
To give you an idea: PCB for paralleled inverting lm3886, with buffer and servo

Not perfect by a long shot but that's the overall direction wrt PS and ground. Read Tom's comments at the end of the thread.
__________________
All that is excessive is insignificant. Talleyrand
  Reply With Quote

Reply


Critique my PCB! Yet another parallel LM3886Hide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
HELP with LM3886 X3 in parallel. Aguilabrava Chip Amps 45 7th November 2016 07:54 AM
Parallel TDA2052 vs Parallel LM3886 for subwoofer goldyrathore Chip Amps 24 21st October 2013 05:50 PM
LM3886 x 2 in parallel jmar Chip Amps 46 24th October 2009 03:53 PM
LM3886 parallel PCB's jmillerdoc Chip Amps 0 27th May 2009 08:26 PM
Are two separate lm3886 = 2*lm3886 in parallel or bridge mode rs1026 Chip Amps 11 21st October 2004 10:24 AM


New To Site? Need Help?

All times are GMT. The time now is 11:49 AM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki