One more LM3886 PCB - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

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 8th June 2012, 01:28 PM   #1
diyAudio Member
 
Join Date: Sep 2005
Default One more LM3886 PCB

Hello everybody,

I've read through a lot of threads regarding the LM3886 and wanted to do a layout of my own.

Here it is:
Click the image to open in full size.

Rm (10k) and Rf (22k) will be soldered directly to the pins
Rz = 2R7, Cz = 100nF
Rg = 22k
Ri = 1k, Ci = 22F
Cr = 220pF
Rs = 10R

At the moment all resistors are 1/4W, should any of them be bigger (e.g. Zobel)?

Any comments, mistakes I've made, parts I forgot, things that could be improved?
  Reply With Quote
Old 8th June 2012, 06:30 PM   #2
diyAudio Member
 
Join Date: Sep 2005
Maybe it's better like this with the picture...
Attached Images
File Type: jpg LM3886_PCB.jpg (105.6 KB, 501 views)
  Reply With Quote
Old 9th June 2012, 12:32 AM   #3
hurtz is offline hurtz  Germany
diyAudio Member
 
Join Date: Jun 2010
Ok, a few things I noticed:

- Why is Cr physically so big?
- You might want to remove the vlaues of the components from your design so it will be easier to read.
- Have you read through the LM3886 datasheet? Also the LM4780 is a good read and has more detailed information than the lm3886 datasheet. (The lm4780 is two lm3886 in one housing)
- Combining a classic PCB with p2p (point to point) components is not considered to be "nice". Although there is no technical problem with that, you will have a lot of work whith those p2p components, and the question arises why not to make it a complete p2p amp.
- Why are C3/C4 the size they are?
- You can use a protection diode across Ci
- Do you already have a PSU?

cheers!
hurtz
__________________
- For there is nothing either good or bad but thinking makes it so
  Reply With Quote
Old 9th June 2012, 10:23 AM   #4
diyAudio Member
 
Join Date: Sep 2005
Hello,

thanks for the reply:

- Cr is so big because I didn't find the right model in Eagle
- will do
- Did read through the LM3886, will read the lm4780
- I already made a complete P2P amp and didn't like it to much (handling) that's why I want to make a hybrid version
- Values for C3/C4 are like most of the other layouts
- What kind of diode would be suitable and what exactly shall it protect from?
- Yes, PSU is available (either single rectifier and 2x3300F per rail (monoblock use) or dual rectifier and 3x4700F+10000F per Rail for stereo use). All Caps bridged with 0,1F film, all diodes bridged with 0,033F.
Attached Images
File Type: png LM3886_PCB_novalues.png (30.1 KB, 435 views)
  Reply With Quote
Old 9th June 2012, 12:18 PM   #5
hurtz is offline hurtz  Germany
diyAudio Member
 
Join Date: Jun 2010
Hi,

- Sorry I meant C1/C2 with 1000uF. Although you will achieve a slightly lower ESR of 0r04 as opposed to a 220uF with 0r09 (reichelt 105 elkos). But considering the physical size, a smaller capacitor might be placed closer to the lm and overall impedance might be reduced a bit further. So if you don't have a very long connection from the PSU to the amp a little less capacitance should work well.. since the datasheet suggests 10uF in any case.
- Cr should also be available as 2.5mm if you want to keep the size down.
- The diode protects Ci in case of power supply failure, in general i think also it is not needed but it adds a little extra layer of security. 1n4001 or something like that.
- Elliott Sound Products - Linear Power Supply Design this is a good source to get some info on PSU design. Or to make it short: 1VA for the transformer per 1W output power. Many small capacitors for the psu are better than one large capacitor. and 10.000uF per rail should be more than enough. I am using one PSU with 3 x 4700uF per rail for a stereo steup. A good test of the PSU is to connect one (two) analog voltmeters to each rail and then watch the voltage drop under heavy load. (I got a maximum of 3V with an 80VA 18V toroid).
- Also you might want to consider not using ground planes but individual traces to keep the grounds better separated (also helps readability ;-)). Also think of the traces as having their own very real resistances and electrons being rather slow. I'm thinking specifically of Cz which is placed after the speaker out. I have no idea what the impact of this is, maybe one of the experts around here might elaborate on this?
- Consider using a 47pF cap across the RCA input jack, helps RF noise. Around 330pF or less across Rin, again for RF filtering. Those two gave me a good 3dB decrease in noise.
- Get familiar with Rightmark Audio Analyzer (rmaa, free tool) and your soundcard. This will help you a lot when debugging possible issues, and it is also nice for everyone else around here to see some measurements. Especially since a lot of borderline audible effects can be made visible. I'm not saying a graph will tell how good the listening experience is, but it will definitly help you find any issues you might have.

cheers!
hurtz
__________________
- For there is nothing either good or bad but thinking makes it so
  Reply With Quote
Old 9th June 2012, 01:31 PM   #6
diyAudio Member
 
Join Date: Sep 2005
Hello,

- Size can be reduced to 10 mm or 8 mm diameter (depending on capacity and voltage), since I will be using a 2x18V toroid, 35V should be fine. I will take a look at the PCB and choose between 220F (8 mm) or more (10 mm).
- Found lots of them with 2,5 mm, don't know why I didn't find them before.
- Will be included
- Will check this when the amp is done, guess I'll start with the smaller units.
- Will make a different version with modified ground traces (as proposed).
- Will consider that
- Will do.

Next step:
Work my way through your thread, modifiy everything, come back
Might take some time since I'm pretty busy the next few weeks.
  Reply With Quote
Old 9th June 2012, 08:10 PM   #7
hurtz is offline hurtz  Germany
diyAudio Member
 
Join Date: Jun 2010
Good luck and don't give up! It's worth the effort :-)

cheers,
hurtz
__________________
- For there is nothing either good or bad but thinking makes it so
  Reply With Quote
Old 13th June 2012, 12:12 PM   #8
diyAudio Member
 
Join Date: Sep 2005
Hello again,

after reading through the datasheets and Hurtzs thread I modified some things.
  • 100nF Cap across Ci (from datasheet)
  • Diode across Ci for protection
  • Values for C1/C2 reduced from 1000 F to 470F in order to gain place
  • Tried to build up two ground stars, one for low-level and one for high level signals

Rf will still be soldered directly to the Chip. Rm will either be soldered to the chip or placed under it.
I probably will add a small ceramic cap across the Input socket (22 - 47pF).

Ri and Ci give a frequency of ~7,3 Hz, Cin and Rin ~2,8 Hz, so this should be ok.
Since Ri = Rb and Rg = Rf I omitted Rf1 and Cf.

I am not sure about the ground stars, should I leave it like that or try to make it more star-like?
What about the rest (width of traces etc.)?
Other suggestions regarding layout or parts?

EDIT: OK, stupid me, forget the jumper, will be placed bit lower as a normal trace...

Parts list:
Cin = 1 F (MKP)
Ci = 56 F (electrolytic, 35V)
Cz = 100 nF (Wima MKS-02)
C1/2 = 470 F (Panasonic FC, 35V)
C3/4 = 100 nF (Wima MKS-02)
Cib = 100 nF (Wima MKS-02)
Cr = 22 pF (ceramic)

Rb = 1k
Ri = 1k
Rg = 22k
Rf = 22k
Rm = 10k
Rz = 2R7
Rs = 10R
all metal or carbon, 1/4W.

Rl = 10R + ~0,6 mH (MOX, 2W)

D1 = 1N4001
Attached Images
File Type: png LM3886_V4.png (26.4 KB, 218 views)

Last edited by Wilddieb; 13th June 2012 at 12:23 PM.
  Reply With Quote
Old 13th June 2012, 12:27 PM   #9
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally Posted by Wilddieb View Post
................
  • 100nF Cap across Ci (from datasheet)
  • Diode across Ci for protection
    ..........
I don't understand why you have added these.
__________________
regards Andrew T.
  Reply With Quote
Old 13th June 2012, 12:59 PM   #10
diyAudio Member
 
Join Date: Sep 2005
Quote:
Originally Posted by AndrewT View Post
I don't understand why you have added these.
Hello master

the 100nF Cap comes from the LM4780 application note (page 19 - high performance considerations), the diode was proposed by you in the thread of hurtz ("The yet tiniest single-sided LM3886?" - post #196)
  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
lm3886 PCB 13814044 Chip Amps 20 25th October 2008 04:09 PM
Lm3886 Pcb SkinnyBoy Chip Amps 11 2nd January 2003 11:32 PM


New To Site? Need Help?

All times are GMT. The time now is 06:47 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