Any chip-amps more powerful than LM3875?

[DeonC]

Hi guys

Are there any chips than give more power than the LM3875. I'm sure there are, but I don't really know where to go. These will be for subwoofer duty.

Thanks,
Deon

[xplod1236]

Yes. There are the LM3886s, which deliver 68w into a 4 ohm load, or 38w into an 8 ohm load. I have bridged 2 LM3886s, but I'm still not satisfied. I want more power. I have some samples of the LM4651/LM4652 170w class d amp, but I don't have time to get the parts for it and/or put it together. Does anybody know of a simple, high-power amp? I like free samples.

[theChris]

read the app notes and all. you have after the lm3875:
lm3886 - single IC, adds mute function DO NOT CONNECT MUTE DIRECTLY TO Vee!
LM4651/LM4652 - classD driver and FETs. 170W @ 4ohm, not full range.

then:
dual lm38xx bridge - 2 ICs to give double the voltage output, you can get about 120W @ 8ohm. 4ohm is no longer a good idea.
dual lm38xx parallel - 2 ICs to give extra current drive @ 4ohms, more difficult to build as well

quad lm38xx bridge-parallel - 4 ICs, 2 sets of paralleled amps that are then bridged. gives most power into 4ohm bridged.

and of course you can get a 50x2 or 60x2 IC in the lm4732 and lm4780 27 pin packages, the 27 pins are pretty close, though it is possible to bend them to fit in 100mil spacings like a typical board has.

I would like to see if the lm4651 could be used with alternative FETs to get more power.

[Steve Eddy]

Originally posted by DeonC
Are there any chips than give more power than the LM3875. I'm sure there are, but I don't really know where to go. These will be for subwoofer duty.

Want a chip with some REAL balls? Try this guy:



It'll take up to a +/- 75 volt supply and can deliver +/-30 amps of current.

Only $492.19 a pop!

se

[geolemon]

I've got the same question...

I'm wondering a couple things:
1) how much power CAN you squeeze out of a chip amp - using the commonly discussed chips on this forum (that one above looks like a beast! - no fair! ) - can you get high power?
2) what's the minimum impedance load that you can run one of these at - and I don't even mean that in the sense of "power keeps doubling as you cut the impedance"... I mean, could you even design your 50 watt amp to make it's 50 watts at 1 ohm, for example.

In reviewing the documentation for the AN-1192 chip, it seems like the answer to both of these is "yes", although I've gathered mostly "no" impressions around here so far...

So let me quote the document, and I'm anxious to hear what people think:

quoted from AN-1192:
The parallel amplifier is another configuration that can be used to obtain higher output power levels...
...the parallel topology provides a great way of achieving higher power levels while keeping within IC power dissipation limits by driving low impedance loads...
In fact, any number of IC's can be used in a parallell design, and when configured the same will share the power dissipation equally.
For example, using four IC's to drive a 1 ohm load means that each IC dissipates 1/4 of the total power dissipation. In other words, the load to each IC looks like a 4 ohm load (Number of IC's in parallel * load impedance = load impedance seen by each individual IC).

If the bridged and parallel configurations are combined, the outcome is a very high-power amplifier solution that far exceeds the capabilities of one IC alone...

This application note seems to indicate that this parallelling technique can be used to increase the power level seemingly without limit, and/or to create chip amps that are low-impedance compatable.

Opinions?

[hieroglyphics]

Whut up Geo!

1) how much power CAN you squeeze out of a chip amp

Well, that description you quoted is a good one. You can keep paralleling chips for higher output current but there are maximum supply voltages that shouldn't be exceeded. So the supply voltage is the limiting factor with chips. If you still want to get more power out of your amp, you'll need to parallel more chips. Double them if you want to cut the impedence in half.

2) what's the minimum impedance load

I couldn't tell ya. If I were to build one, I don't think I'd strive for anything less than 2 ohms. I think you may have stability issues as a result dunno for sure. I'd make sure your voltage rails are as high as possible to maximize the output power.

[matjans]

i just finished building a lm4780 based chip amp on veroboard (yes, it is doable) and it should pump out about 100-120W as per the datasheet (parallel config).

Man i went through hell trying to get one in NL. I had an american friend of mine send one via mail.

[JOE DIRT®]

Steve....I bet you never seen one like this..LMAO

[Steve Eddy]

Originally posted by JOE DIRT®
Steve....I bet you never seen one like this..LMAO

Hehehe. Nope. Appears to be about the same size as the PA-03 (which is 2.3" x 1.63").

Only found one reference to it on the web which says it's a linear regulator.

se

[tlmadsen]

To all you power hungry peolpe

Have a look at

200 watts Gäjnklon (Gainclone) - BPA200

It will give you more than 200W in 8 ohm, 350 W in 4 ohm and if you use (paralleling) enought af "Das Modul" you get (teoritical)close to 1000 W in 1 ohm............

Have fun

Thomas