Required heatsink for 3 LM3886 in parallel? - 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 2nd January 2009, 12:28 AM   #1
CJ900RR is offline CJ900RR  Sweden
diyAudio Member
 
Join Date: Jan 2006
Default Required heatsink for 3 LM3886 in parallel?

Hello and happy new year!

I found this heatsink at a Swedish web shop, measuring 200 x 40 x 100 mm (B x H x D). R(t) is 0.55 K/W. Website, Swedish only, datasheet, only in German.

They are pretty cheap, about 25 Euro so I started thinking about using them for my boards with 3 pcs of LM3886 paralleled. The problem is that I suck at calculating if they are big enough. What do you think? My speaker is 4 ohm and voltage is a bit high, +/- 32V, but the voltage could be changed if required.

Best regards...
  Reply With Quote
Old 2nd January 2009, 05:47 AM   #2
sangram is offline sangram  India
diyAudio Member
 
Join Date: Sep 2002
Location: India
.5 is probably good for two chips.

For three chips putting out 100 watts assuming equal current sharing, the dissipation should be in the range of 30 watts for each chip. A 4 ohm load and the supply voltage you indicate is not a problem for the 3886.

Each chip also has a thermal resistance of about 1 c/w, beyond that the thermal calculations are a little tough to put here (I work better with paper). But it doesn't look very good - each chip will be seeing close to 2-3C/w thermal resistance from case to ambient as they share one heatsink.

Essentially you'd be running around 80-90 degrees on the heatsink at that kind of power output if it were sine waves, which is high but still within the chip's limits.

Now for the good news: It'd probably work for music. I run 2c/w heatsinks with *two* LM3886TF (higher resistance) on a very demanding load, and the temperatures never hit very high levels not matter what, actually cooling down a bit at extreme levels. I've never seen them go above 60 degrees under any circumstances. I would probably not run a sine wave test on it, anyhow.
  Reply With Quote
Old 2nd January 2009, 11:06 AM   #3
diyAudio Member
 
Join Date: May 2008
Pdmax = Vcctot/((2*PI)*Rl) = 64V/(2*PI*4Ohm) = 51,9 W
That is 17,3 W maximum per IC.

LM3886T is given with 1 K/W, LM3886 TF with 2 K/W, the junction assumed with 0,2 K/W.

Your ICs and your heatsink will be running nice and cool. Those heatsinks will heat up to around 29 K above ambient at worst and the IC junction will be around 50 K above ambient for the unisolated package and 67 K above ambient for the isolated package. And as sangram said, real music signals are much less demanding.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 2nd January 2009, 01:33 PM   #4
diyAudio Member
 
Join Date: May 2008
Edit.

Actually all those temperatures will be higher. The heatsink you chose has 0,55 K/W at a Pd of 120 W with DeltaT above 80 K. At Pd 52 W it is more like 0,87 K/W, with Delta T somewhere between 40 and 45 K. The unisolated IC will then be 60..65 K above ambient, the isolated one 80..85 K.

SPiKe sets in at 150 C. This heatsink will allow your amp to work up to an ambient temperature of 65 C at full throttle even with the isolated package before you reach 150 C at the IC. The ambient temperature is measured around the heatsink, i. e. for heatsinks inside a case you have to reduce that number accordingly. But then you can always use the unisolated package and be safe up to 85 C.

You should also take into account that heatsinks outside of the case should be cooler than 60 C to avoid that people burn their fingers. That would limit the use of your amp to below 15..20 C ambient temperature on worst case conditions.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 2nd January 2009, 02:18 PM   #5
sangram is offline sangram  India
diyAudio Member
 
Join Date: Sep 2002
Location: India
I was not calculating it to a science like you, but I guess I was close enough. The load impedance seen by each chip is not 4 ohm, but 12 ohm in this particular case. But that is per single chip. These are connected (strictly speaking, thermally) in parallel to the same heatsink, via three different thermal resistances, not the same.

So each chip sees its own Tj-Tc plus the thermal resistance of the heatsink multiplied by three - essentially for the isolated package 1+1.5, and for the isolated package 2+1.5, as the heatsink is shared between three chips. So for the isolated package (worst case) the total Tj-Ta is 3.5C/W. Not disastrous, but will probably run fine. At 20 watts per chip, that's 70 degrees above ambient.

Also parallel amps run warmer, as the chips battle each other to source or sink current equally.
  Reply With Quote
Old 2nd January 2009, 02:30 PM   #6
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
look up the graph/table on page 14.
three chipamps on a 0.55C/W sink will run at approximately the same temperature a a single 3886 chipamp on a 1.6C/W heatsink.
find 1.6C/W & Ta=25degC on the table.
Go across horizontally to the 4r0 plot and then down to supply voltage. It shows 60Vdc i.e. +-30Vdc.
Now find 1.6C/W @ 40degC repeat the across and down. Max supply for 4r0 is +-28Vdc.

However, National have used all the data that puts the 3886 chip in the best light.
These recommended heatsinks are too small. I always recommend that you double the heatsink capacity compared to what National show in the table.
To see the effect of doubling the heatsink look for 3.2C/W @ 40degC and the supply voltage for 4r0 is +-22Vdc.
Change to 8ohm speaker loads and you will find that the supply can run at <=63V. Your +-32Vdc is just about perfect for 8ohm speakers with Ta=40degC and 3off 3886 on a 0.55C/W heatsink.
__________________
regards Andrew T.
  Reply With Quote
Old 2nd January 2009, 05:35 PM   #7
diyAudio Member
 
Join Date: May 2008
Quote:
Originally posted by AndrewT
three chipamps on a 0.55C/W sink will run at approximately the same temperature a a single 3886 chipamp on a 1.6C/W heatsink.
1 IC: 1,6 K/W heatsink + 0,2 K/W thermal grease and washer + 1 K/W IC = 2,8 K/W total. -> DeltaT at Pd 52 W = 145,6 K.

3 ICs: 0,55 K/W heatsink + 1/3 * (0,2 K/W thermal grease and washer + 1 K/W IC) = 0,95 K/W total. -> DeltaT at Pd 52 W = 49,4 K.

Not even close.

3 ICs with corrected heatsink dissipation figure: 0,87 K/W heatsink + 1/3 * (0,2 K/W thermal grease and washer + 1 K/W IC) = 1,27 K/W total. -> DeltaT at Pd 52 W = 66,04 K.

3 isolated ICs with corrected heatsink dissipation figure: 0,87 K/W heatsink + 1/3 * (0,2 K/W thermal grease and washer + 2 K/W IC) = 1,6 K/W total. -> DeltaT at Pd 52 W = 83,2 K.


Quote:
Originally posted by AndrewT
Your +-32Vdc is just about perfect for 8ohm speakers with Ta=40degC and 3off 3886 on a 0.55C/W heatsink.
Even a single IC would be fine with 32 V, 8 Ohm and a 0,55 K/W heatsink. In PA-150 configuration each IC sees a comfortable 12 Ohm load.
__________________
If you've always done it like that, then it's probably wrong. (Henry Ford)
  Reply With Quote
Old 2nd January 2009, 06:19 PM   #8
AndrewT is offline AndrewT  Scotland
diyAudio Member
 
Join Date: Jul 2004
Location: Scottish Borders
Quote:
Originally posted by pacificblue
1 IC: 1,6 K/W heatsink + 0,2 K/W thermal grease and washer + 1 K/W IC = 2,8 K/W total. -> DeltaT at Pd 52 W = 145,6 K.

3 ICs: 0,55 K/W heatsink + 1/3 * (0,2 K/W thermal grease and washer + 1 K/W IC) = 0,95 K/W total. -> DeltaT at Pd 52 W = 49,4 K.
That is not the comparison I used.
One single 3886 on a 1.6C/W sink will run at about the same temperature as 3off 3886 on a shared 0.55C/W sink.
This last statement shows
DeltaT at Pd 52 W = 49,4 K.D
for comparison it should be
DeltaT at Pd [3* 52 W] = ? K.
Is it closer now?

If +-32Vdc is an acceptable supply for a single 3886 driving an 8ohm load using Ta=40degC and Rth s-a=0.55C/W then three amplifiers each driving an 8ohm load for the same conditions will operate at almost the same temperature when sharing a 0.55C/W sink.

If CJR has used these 3 amps in parallel to drive a 4ohm load then yes, each amp sees the equivalent of 12ohms and that makes the sink more conservative.

One nice thing about a 3parallel setup that is designed for 8ohm and driving a real 4ohm load is that each chipamp will use less of it's maximum capability and as a result the SPIKE and temperature protection should be triggered less often.
__________________
regards Andrew T.
  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 Heatsink size? henkel Chip Amps 33 30th July 2012 04:45 PM
LM3886 in parallel Denis.BR Chip Amps 11 12th August 2009 08:16 PM
LM3886 Heatsink ormo Chip Amps 27 27th July 2009 05:15 AM
Parallel TDA1541A ... some clarifications required skje Digital Source 3 25th March 2008 02:11 PM
Are two separate lm3886 = 2*lm3886 in parallel or bridge mode rs1026 Chip Amps 11 21st October 2004 11:24 AM


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