What is a Gain Clone exactly?

[Peter Daniel]

I don't really care for his intentions here. He might know or might not. It's not for me to judge.

But I remember Phred mentioning once that he puts GC on a very short least of the best amplifiers he ever heard

[theChris]

well, to answer lars, gainclones is just a word now. there are trends in them. some people try many different variations because the parts are cheap enough and few enough, well, in general.

there are non-inverted and inverted gainclones, there are battery powered ones popular now. there are are "bridge-clones" which are used for subwoofers.

in any-case for PCB, there are few as most people will do a point-to-point wireing on the chip. the new 4780 and 4766 (i think) are 27 pins and are more difficult to p2p, though not impossible.

[peranders]

http://www.sakurasystems.com
http://www.sakurasystems.com/products/47amp.html
http://www.diyaudio.com/forums/showt...172#post222172
http://sound.westhost.com/madashell8.htm#gaincard

I have also made a GainClone but I'm waiting for Mr. Dieckmann's design review so I can start my multi million SEK business.... just kidding.
My approach isn't exactly minimalistic, it's a monster.

[Lars Clausen]

My intent of starting this thread is:

It could be fun to take this discipline of DIY audio even further. I mean:

Construction of Ultra Low Cost Amplifiers.

It shouldn't be too hard to construct something even cheaper and better than Gainclone .. I have been calculating last night, an amplifier with similar specs like the Gainclone (or better) can be built for around 10$ all parts and a nice PCB included.
Maybe less.......?

Could be fun.

Another example of Ultra Low Cost Amplifiers would be this Romanian built amplifier - yes based on old technology, but i am sure it works fine, and will be on level with other low cost solutions like the GC. And wait till you hear the price.

Take a look:

http://www.elmatrd.ro/amplificatoare.html

The 1 x 100 W RMS module cost 290.000 Romanian Lei that is equal to: ~ 9 US$ (Ready assembled and tested module).
How fun would it be to tweak a low cost module like this?

[peranders]

Originally posted by Lars Clausen
My intent of starting this thread is:

It could be fun to take this discipline of DIY audio even further. I mean:

Construction of Ultra Low Cost Amplifiers.

It shouldn't be too hard to construct something even cheaper and better than Gainclone .. I have been calculating last night, an amplifier with similar specs like the Gainclone (or better) can be built for around 10$ all parts and a nice PCB included.
Maybe less.......?

Could be fun.

You could also take a lot of money for it like Sakura Systems....

What do I get for 10 USD, heatsinks, PSU etc? A pcb + LM3886 + plus a few parts don't cost much. I guess you can get a pcb for 1 (not more than 2) USD even a double sided with plated through holes?

[Lars Clausen]

I see no reason to use a LM3886 for a project like that.

A discrete solution would be much better. And can be made at the same price. And is more fun to make. And has many more tweaking options than a chip amp.

For example: Try other output transistors in a LM3886 amplifier ..

The PCB's i have seen in here all cost 5$ plus shipping, and are mostly single sided low cost boards.
I would suggest a nice Dual Sided board (like the ones you make Per Anders).

A simple 10$ amp (parts and PCB) can be made to accept supply voltages up to +/- 80V Almost twice that of LM3886. Also 10-15 Ampere output current is possible, almost 5 times that of LM3886, so outputs of:

50 W in 8 Ohms and
100 W in 4 Ohms

is possible without stacking chips.

I would use IRF540 / IRF9540, a set of drivers, about 10 SMD mounted transistors, and some fair quality caps. All can be easily tweaked and scaled (unlike LM3886 solutions). Replace the output devices with ones that cost a couple of Dollars more (IRFP140 / IRFP9140) and you have 100 W in 8 Ohms / 200 W in 4 Ohms. Or with IRFP240 / IRFP9240 you can even make something like 200W in 8 Ohms. Still around 12$ price tag.

Who knows maybe it can be made even cheaper than my suggestion??

[peranders]

Lars, it's not very complicated to make a good MOSFET amp with few parts. Maybe you shall start a project called ClausenClone

It's a good thing because there aren't that many good high quality power amp projects with good pcb's. As you know I'm also into SMD at the moment. It would have been very fun with a cute little driver board.

I have talked to Fred and he feels that you should go for 0604 parts or smaller....

BTW: have you checked the swedish project XTF at www.hififorum.nu All SMD with error correction(!), very interesting so far, not a minimalistic design though. It's a monster in fact
http://www.hififorum.nu/forum/topic....archTerms=,xtf
http://www.faktiskt.se/modules.php?n...iewtopic&t=566

[ronc]

I once built a SE class A mosfet amp for about the cost of the GC, but IMHO didnt sound near as good as the GC.Also had to use the biggest H/S i had in my parts box.
ron

[Peter Daniel]

Originally posted by Lars Clausen

A discrete solution would be much better.

I very much doubt it.

Coming back to your initial question (as I didn't know what you really meant), the cost of building GC is about $10 per channel

[Lars Clausen]

Thanks for this info, Peter.

I guess i will have to prove my point