Multi purpose multi channel gainclone PCB: Team project - group buy

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Multi purpose multi channel gainclone PCB: Team project - group buy

Who’s interested in a flexible PCB design and group buy, useable for either:
- a cost and time effective active 3 way (or 4 way) power amp, or
- a home theatre power amp, with say two (or more) channels bridged to drive a single sub.
Even entry level eg Marantz integrated home theatre receivers have pre-outs on all channels. It wouldn’t take much of an amp to better most built in power amps (or get similar or a little better quality to many HT dedicated power amps, for fewer $ even than second hand).
IMO it’s not worth spending too much on HT power amps. Better to put the money into speakers or other things.


The opamp based stereo power amps – the commercial Gaincard and Monarchy Audio SM-70; Linkwitz’s simple diy design based on the National Semiconductor LM3886, the “Gainclone” and variations, etc:
Better than eg many Rotel etc amps, certainly the Rotel home theatre power amp.
The Monarchy has been rated by some better than Aragon!

Simple, cheap, not "SOTA” but very good, especially for the cost and ease of construction.
Big attractions are:
- Depending on design, components total only 9-12!
- Dependent on components, *total* cost of a gainclone 2 channel power amp, is around $US 40-60.

The other day at Madisound Doug posted about the 8 amps needed to drive the active 3 way Linkwitz Orion.
To drive this, the mid and tweet each get an amplifier, and each of the two woofers get a channel– an 8 channel amp.
Other DIYers are doing other active 3 ways (I’m considering the Linkwitz Phoenix, and my own active 3 way).

Doug’s suggestion was combining 8 channels of power amplifier based on the LM3886, with SL’s Orion filters with the eight 3886s all onto ONE board.
SL doesn't think there would be much demand for a specific unified board.

(I’m not an amp designer, and not (just yet) an amp builder).
Some opamps (including the LM3886) are bridgeable.

There are newer opamps (eg the TDA 7293) that sound better.

Although the trigger for my idea was a post on Madisound, diyAudio seems to be the place for this. Need to agree:

(A) Number of channels (maybe 10 is better) and how much flexibility about what is bridgeable, to make either:
1. Stereo 3 way with more grunt in the bass,
2. For the more ambitious, a stereo 4 way
3. Stereo 3 way with dual subs (much the same)
4. 5 channel with all channels bridgeable
5. 6 channel with double power to the L/ R/ C and sub.
6. Etc

(B) The chip & design
Exact design, the power supply, how many power supplies, etc.
Multiple power supplies, maybe 4? 5? would be good.

(C ) Who to do the group buy - mailout. I would be prepared to do this.
I’m in Australia, so freight would be a little higher (though weight will be low).
Aksa boards are locally made, so this should be competitive. And eg for our US participants, the USD is worth about $A 1.70!

Dependent in interest or opinion, it could be posted at the DIY section of AVS:
http://ww.avsforum.com/avs-vb/forumdisplay.php?s=&forumid=28
(would boost economies of scale)

As we sometimes say here, whatya reckon?

NB: I only get the chance to check the site 2-3 times a day.

Richard
 
Hi Rick,

Isn't the idea behind the gainclone to do p2p wiring rather than PCB?? Also everyone interested in this group buy would need to decide on what spec, chips, topology, parts etc. to use.

I have just received my LM3875 samples from NS (Thanks NS - I will be paying for more NS parts on another project) and keen to get my gainclone going.

I'm looking for a home theatre amp with the following specs:
- As high as possible 2 channel quality
- 3 x surround sound using similar but lower quality parts to keep the costs down.
- Based on the Peter Daniels tweaked up recipe.

A group buy on the components as well as the PCB would be useful. I am also based in Australia and having a tough time sourcing components like Riken resistors, Alps pots & Black Gate caps without paying ridiculous prices in US$$$$. There is a group buy on Black Gate on this site's WIKI but it seems to be dead.
 
p2p. Peter Daniels tweaked starting point

Although maybe_ all gainclones are p2p, I dont think this is critical to sound quality.
Anyone feel free to correct me.

It is sensitive to quality of components. Short distances are beneficial so circuit design should be done carefully.

I expect this project to take a while to get going. If youre in a hurry, component selection will realistically take a while, preceded by circuit design.

Peter Daniels tweaked recipe would be a good starting point. I hacvent followed recent gainclone threads. Do you have a quick or direct link


Richard
 
Re: p2p. Peter Daniels tweaked starting point

Koinichiwa,

Although maybe_ all gainclones are p2p, I dont think this is critical to sound quality.

Think again. Or even better, build one stereo Gainclone P2P and another on PCB. The results will surprise you quite a lot. In low level circuits it is even greater of a difference.

Plus, P2P allows a lot tigther layout.

Ask Jonathan Carr.

Sayonara
 
I never liked the idea of pcb for gainclones until recently. I realized that pcb's make it easier to connect both channels in the exactly the same way (length of connections) and ensure good connections. Also, the components can be mounted very rigid to the board. As long as the feedback resistor is soldered directly to the chip and the pcb is as SMALL as possible I do not see a problem. The original gaincard also uses a very small pcb and sounds great! :nod:

Though I wouldn't use pcb's larger than say 5x5 cm...

Fedde
 
integrated stereo amplifier

Thanks Nisbeth, though this is an integrated stereo amplifier . .

I was thinking that much of the very recent learning/ development that has been done by gaincloners (eg Peter fedde and others) could be translated without a huge effort into a multi-channel.

If someone here doesn't do it, I'll bet London to a brick that someone will. If I had the experience and time to do the design.

Doug thinks that the audio community regards projects such as this with distain. He's proably right. But I think they will be proved wrong, when it's done and listened to.

Whoever does it first (and well) would win a lot of kudos, and if they did it commercially, I suspect a lot of moola. I wish it could be me!
 
Rick this is an interesting idea. It's been a while since I've built anything, but what brings me here is an interest in building a nice amp for my PC speakers. In this thread I was told about this amp using the LM3876. That thread also has the most difficult design that I ever built, I am by far not an engineer. I repaired consumer electronics for quite a few years, so I have a very good understanding in theory, but I have no formal training in design. I haven't looked into building anything in years, and to be honest I'm suprised in the amount of info found on the net. I've only been on the net since around 1996. Newbie I am :p

I was thinking about using the LM3886 for the sub woofer, the LM3876 to drive the five inch mid/woofers and possibly the LM1875 to drive the tweeters. The tweeters might need the LM3876 in order to "keep up" with mids sort of speak. Also I was thinking about using one of the crossovers from this page.

Maybe if the design was done is a modular fashion people could choose the amount of channels that they need, such as 2.1, 4.1, 5.1 and so on. With a choice of different power supplies. It would also need to have the ability to change crossover frequencies easily, since everybody isn't going to be using the same drivers. The link that Nisbeth posted has a nice, compact circuit board already designed for the IC which can be downloaded HERE.

I would like to build one if it was built with fairly standard components. I don't have the money or desire to build an amp with $20 resistors or $200 capacitors, unless needed, to be honest. I've just started looking in to building this amp this week, I still have a lot more research to do, but I will be building one soon. National Semiconductor has tons of info on these IC's and design parameters, you can even get free samples.
 
DA 7294 may be better

I think this will be a while before it gets wings. Most are interested in just two channels.

When I get time, I'm going to do a couple of versions with two channels, then move on.

My reading of the threads suggest that the TDA 7294 may be a better and/ or more powerful chip.

Regards

Richard
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.