16 paralleled LM3886s or output transistors? - 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 24th November 2004, 04:28 PM   #1
diyAudio Member
 
xplod1236's Avatar
 
Join Date: Dec 2003
Location: On the couch
Send a message via AIM to xplod1236
Angry 16 paralleled LM3886s or output transistors?

I need to make a high-power amp, and I thought that the LM3886 would do the job. My power supply is +/- 42.5v at no load. The speaker that I want to drive has two 5.4 ohm voice coils. I was thinking about making 2 bridge/parallel amps, one for each coil. After doing some calculations I figured out that the amp would need to dissipate ~550-600 watts max. With 16 chips for the whole amp, each chip would need to dissipate ~35w.

Would it be better to use 8 chips per bridge/parallel setup (16 chips for the amp) or paralleled output transistors?

Would paralleling 4 chips work, or is it too many chips?
  Reply With Quote
Old 24th November 2004, 05:28 PM   #2
diyAudio Member
 
Join Date: Mar 2001
Location: Brazil
Default Re: 16 paralleled LM3886s or output transistors?

Quote:
Originally posted by xplod1236
I need to make a high-power amp, and I thought that the LM3886 would do the job. My power supply is +/- 42.5v at no load. The speaker that I want to drive has two 5.4 ohm voice coils. I was thinking about making 2 bridge/parallel amps, one for each coil. After doing some calculations I figured out that the amp would need to dissipate ~550-600 watts max. With 16 chips for the whole amp, each chip would need to dissipate ~35w.

Would it be better to use 8 chips per bridge/parallel setup (16 chips for the amp) or paralleled output transistors?

Would paralleling 4 chips work, or is it too many chips?
Power supply is to high, particularly for bridging.

IMO you have two choices:

1) Regulate the voltage to aorund +/-30v, then parallel/bridge.

2) Use another chip as first stage and boost the output.



Carlos
  Reply With Quote
Old 24th November 2004, 05:37 PM   #3
diyAudio Member
 
xplod1236's Avatar
 
Join Date: Dec 2003
Location: On the couch
Send a message via AIM to xplod1236
Angry Re: Re: 16 paralleled LM3886s or output transistors?

This is the only transformer that I have that is suitable for the chips. I know that the voltage is a little high, but wouldn't it decrease under load?

Quote:
Originally posted by carlmart


IMO you have two choices:

1) Regulate the voltage to aorund +/-30v, then parallel/bridge.

2) Use another chip as first stage and boost the output.


Carlos
1) I want to keep this amp simple. Wouldnt the chips handle 40v and 4 amps each? Each chip would see a load of 10.8 ohms. I have a lot of LM3886s, and I would like to use them.

2) See 1
  Reply With Quote
Old 24th November 2004, 06:01 PM   #4
macboy is offline macboy  Canada
diyAudio Member
 
Join Date: Oct 2003
Location: Ottawa, Canada
Do it man, forget the nay-sayers.

If you did a bridge-parallel arrangement with 4 paralleled chips bridged with another 4 paralleled (per voice coil), then each chip 'sees' an effective load of 10.4 Ohms (2x the real voice coil impedance). With a +-42 V supply, 10.4 ohms is OK; it is within the Safe Operating Area specified in the datasheet, and that's what matters. I also estimate about 35 W power dissipation (per chip) based on the data sheet, confirming your own observation.

I strongly recommend, no, absolutely insist, that you use the LM3886T, not the LM3886TF. The latter has too much die-to-case thermal resistance and will overheat internally when put under the kind of stresses to which you will be subjecting the device.

I don't know what kind of transformer your are planning to use, but it had better be huge. I think about 1500 VA sounds about right.

One other thing, carefully read and understand AN-1192: Application Note 1192 Overture™ Series High Power Solutions. You must make sure that the gain of all paralleled chip amps is matched very closely, and you must make sure that DC offset of each is minimal or, better yet, zero. You may need to use DC servos to achieve this. These steps will be of great help in reducing the power dissipation of the amplifier, both at idle and operating.
  Reply With Quote
Old 24th November 2004, 06:29 PM   #5
diyAudio Member
 
xplod1236's Avatar
 
Join Date: Dec 2003
Location: On the couch
Send a message via AIM to xplod1236
Quote:
I strongly recommend, no, absolutely insist, that you use the LM3886T, not the LM3886TF. The latter has too much die-to-case thermal resistance and will overheat internally when put under the kind of stresses to which you will be subjecting the device.
I would use the T version, but I already have 16 of the insulated ones, so I will use those. If they fry, then I'll switch to the non-insulated ones.

Quote:
I don't know what kind of transformer your are planning to use, but it had better be huge. I think about 1500 VA sounds about right.
The transformer is coming from a Fisher stereo receiver. I don't know the rating of it, but it looks pretty beefy. The fuses on the secondaries were 6A. I don't really care how much power I will get out of it. That's the only transformer that I have that will work with the chips. Another one I have is 35-0-35 and another is 52-0-55 vac, both of which would fry the chips.

Quote:
One other thing, carefully read and understand AN-1192: Application Note 1192 Overture™ Series High Power Solutions. You must make sure that the gain of all paralleled chip amps is matched very closely
I did read it a few times already. All the chips will be non-inverting, with a drv134 in front to split up and invert 1/2 of the signal.


Can I NOT use a cap between gnd and the resistor that goes to the inverting in?
  Reply With Quote
Old 24th November 2004, 07:02 PM   #6
Previously: Kuei Yang Wang
 
Join Date: Nov 2002
Default Re: 16 paralleled LM3886s or output transistors?

Konnichiwa,

Quote:
Originally posted by xplod1236
My power supply is +/- 42.5v at no load.
That is 85V between the PSU Pin's. The LM3886 is rated as 94V max, so you can JUST ABOUT get away with it, but I would recommend finding a way to reduce the Voltage slightly. If your mains transformer is a torroidal type you can wind by had a few doxend turns as "buck" winding for the primary, thus reducing the PSU Volate.

Alternatively using "amplified capacitor" circuits with a PSU "Clamp" at +/-40V would also work well.

Quote:
Originally posted by xplod1236
The speaker that I want to drive has two 5.4 ohm voice coils. I was thinking about making 2 bridge/parallel amps, one for each coil. After doing some calculations I figured out that the amp would need to dissipate ~550-600 watts max. With 16 chips for the whole amp, each chip would need to dissipate ~35w.
Sounds doable allright.

Quote:
Originally posted by xplod1236
Would it be better to use 8 chips per bridge/parallel setup (16 chips for the amp) or paralleled output transistors?
Pro LM3886 is that they are next to indestructable if applied with minimal care. The same cannot be said of normal output transistors in discrete amplifiers, at least at the kind of supply voltages and currents involved.

Quote:
Originally posted by xplod1236
Would paralleling 4 chips work, or is it too many chips?
Depends, for music only I suspect you could get away with 2-3 parallel chips. If you want to do sine wave testing near clipping you need to worry not just about peak currents but also dissipation, suggesting the need for more IC's. On the positive side, all that will happen if you do thermally overtax the system is that whole amp will rapidly shut down with the LM3886, with discrete output transistors thermal runaway tends to be a lot more catastrophic.

Sayonara
  Reply With Quote
Old 24th November 2004, 07:16 PM   #7
K-amps is offline K-amps  United States
diyAudio Member
 
K-amps's Avatar
 
Join Date: Nov 2003
Location: Indiana
Quote:
Originally posted by macboy
Do it man, forget the nay-sayers.

Words I live by!

  Reply With Quote
Old 24th November 2004, 07:30 PM   #8
macboy is offline macboy  Canada
diyAudio Member
 
Join Date: Oct 2003
Location: Ottawa, Canada
Quote:
I suspect you could get away with 2-3 parallel chips. If you want to do sine wave testing near clipping you need to worry not just about peak currents but also dissipation, suggesting the need for more IC's.
It is certainly true that real music is very different than sinewave testbench signals. Keep in mind that average power dissipation isn't the only important thing, instantaneous current is also important. The SPiKe protection may kick in if the instantaneous current shoots too high, even if the chips are kept cool.

He could get away with just 3 chips per parallel arrangement, which would still keep the effective load for each chip at 8.1 ohms or so. That would be acceptable, but is approaching the edge of the safe area. Since xplod wants to use TF devices, I would still recommend using 4 chips in parallel. This will help distribute the power dissipation amoung more devices, reducing the chances of overheating (which can be a real concern with the TFs).
  Reply With Quote
Old 24th November 2004, 07:36 PM   #9
macboy is offline macboy  Canada
diyAudio Member
 
Join Date: Oct 2003
Location: Ottawa, Canada
Quote:
The transformer is coming from a Fisher stereo receiver. I don't know the rating of it, but it looks pretty beefy. The fuses on the secondaries were 6A
Well, with 42 V after rectification your secondaries are 30 V each. 30 V(rms) * 6 A(rms) is 180 VA, per secondary. So, it is an approximately 360 VA transformer. You can also check out the specs on the back panel of the receiver it came out of to see its line rating (usually in Watts in the USA; but VA in Canada). Luckily, real-world music does not demand high continuous power. Most transformers can deliver a lot more than their rated power for short bursts, as long as the average power is low enough that the transformer doesn't overheat.
  Reply With Quote
Old 24th November 2004, 07:43 PM   #10
diyAudio Member
 
xplod1236's Avatar
 
Join Date: Dec 2003
Location: On the couch
Send a message via AIM to xplod1236
Quote:
Originally posted by macboy

Well, with 42 V after rectification your secondaries are 30 V each. 30 V(rms) * 6 A(rms) is 180 VA, per secondary. So, it is an approximately 360 VA transformer. You can also check out the specs on the back panel of the receiver it came out of to see its line rating (usually in Watts in the USA; but VA in Canada).

It says 230W max, but the fuse on the power line to the transformer is 5A.
  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
Using Many Small Signal Transistors Paralleled in place of one output Transistor John Biles Solid State 39 8th February 2014 06:58 PM
Thoughts on parallel LM3886s davidsrsb Chip Amps 4 6th March 2006 03:55 PM
output resistor matching for LM4780 paralleled Audiophilenoob Chip Amps 8 7th August 2005 10:03 AM
Long output lead or long output transistors leads? which one to pick? chris ma Everything Else 3 6th June 2003 12:06 PM
Output Transistors weinstro Parts 1 9th May 2003 11:51 AM


New To Site? Need Help?

All times are GMT. The time now is 11:15 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