Page 1 - Idea and circuit choices
Page 2 - P3A Amplifiers
Page 3 - Crossover and Surround Channel Amps
Page 4 - Power Supply details
Page 5 - Testing
Page 6 - Improvements
Well, this project was meant to come after my planned pre-amp which would be part of a computer science uni project.
However the pre-amp is falling behind this massive power amp although construction is ongoing in parallel to building my 5 channel power amp.
These few pages are written during the construction, although they started towards the completion of all the PCB's.
So onto the initial idea. Well, as I said I was planning on building a pre-amp which would be microchip controlled. But I would also need a good power amplifier to work with the pre-amp and I decided I was not too badly off and started spending out on parts and accessories for the project to take shape.
I didn't realise exactly how much I would end up spending in the end... as although (as you will find during this article) I got some great bargains off ebay, I still had to do long orders for parts to several different stockists to get all I needed. At this moment I will still need more.
It took me a while to decide on what power amplifiers to use. I had set my heart on building what is referred to as the P3A (or project 3A) from ESP (Elliott Sound Products). Rod Elliott is the designer of this amplifier and a very great designer and person to have details available on his site.
The ESP site is fantastic and many of my projects are work from the site, however the sites existence depends on the sale of PCB's and if that is reason enough, the PCB's are good quality (i.e. better than anything I could manage) and you get the benefit of saving time and effort and the reasurance that you have a good working design to build your project on.
I must confess though, the amplifiers in this project are built on my own PCB's. I did buy one P3A board and this is most soldered up now and will reside until I decide to bride it and produce a subwoofer amp.
My PCB's are mostly based around the artwork of the ESP ones, but with modifications for my components and having the (essential) fuses external to the amp in order to fit them to my heatsinks. Each PCB contains two amplifiers, and there are two PCB's.. giving a total of 4. For more explanation, see bi-amplifying below...
Do not ask for my PCB design as I will not release it. It is unfair to the selling of PCB's on the ESP site.
I can say they were a lot of work. But (hopefully) it will all be worth it. For more details, follow this link to see Rod Elliotts' article on the P3A amplifier.
I grabbed the LM3886's off ebay at a good price and I thought these would be ideal for the rear channels, certainly high enough fidelity for them anyway!
At this moment in time, the PCB is built (and is very small for two amplifiers) but I am still searching for a decent heatsink.
But anyway, a little description...
The LM3886 is a 68W integrated single chip amplifier. This basically means not many components are needed to build a 68W amplifier. 68W is no small power. Unlike most of my projects with single chip amplifiers... this one is not based entirely off the typical application in the datasheet.
Instead my work has been based on the work Rod Elliott has performed with his LM3876 design.
Again it is a very simple and good design. Rod has based his article on the LM3876, which is similar to the LM3886's I have but less power (50W). The article is here and provides a very good amount of detail to allow someone to easily build this power amp. I have had no trouble myself (so far).
There is a slight difference, which is described in his project and that is an extra connection is needed for the LM3886. My PCB design includes this so all should be ok but it is a useful fact to note. Again, do not ask for my PCB design.. it would be unfair for me to release them when PCB's can be brought from the site.
For common use into rear speakers the LM3876 will be absolutely fine. I only got 3886's due to them appearing on ebay.
Apparently a cheap and common amp. Another ebay bargain (I hope to get 5 at this time). These chips claim to give 100W (!) from 100V. Whilst I will be using +/-35V I'm sure these will give around 60W which is certainly enough for a centre channel amplifier and the quality should be more than adequate.
Construction details for this amp will appear on my site once I finish with the board and its design. Look out for TDA7293 on my main page. It will all be based on the typical application and therefore PCB details shall be given.
Certainly an exciting concept, but if your confused right now, let me give a quick explanation...
Normally the high and low/mid is split by passive crossovers inside the loudspeaker cabinets themselves. Passive crossovers have problems though which prevent them to perform as an ideal crossover. Power handling, impedance problems are all common that manufactures spend lots of money on getting around for a decent pair of hi-fi speakers.
Bi-amplification has several benefits:
These are quotes from the Bi-amplification article on the ESP website (linked below):
That was actually more than I remember. If your thinking about bi-amplification, Rod Elliotts' bi-amplifiction aritcle is a must read :)
So that is the lazy way to quote why I decided to go for bi-amplification. But why did I actually decide to buy a bigger power supply, and more amplifiers when there are so many systems out there which do very well for standard amplification.
This project was never meant to be small time when I decided to do it and there were cases where I thought about going for standard amplification, but I held on to the bi-amplification idea simply because I want to treat my main speakers with the best quality I can. It didn't really cost me that much more anyway.
However one thing you must do to benefit from bi-amplification is to disconnect the passive crossover in the speakers. Otherwise any benefits will simply be undone. At this moment in time, it seems the crossover in my mission m71i's will easily be disconnected and replaced at a later date with ease, but I have not checked my Moduant Short Pearl's yet.. hopefully it will be similar.
Where the magic of bi-amplification begins. This is quite an advanced crossover that is presented once again on the ESP site.. Project 9.
I decided to design my own PCB. I think I managed to pull off a pretty decent design. It is certainly small and should give good results with hope. I decided to put the output buffers and controls separate to the main board (which also contains voltage regulators for the op-amps). Again I won't be giving out any layouts for this.
The final board inside the power amplifier that I would need is a simple fan controller. The idea is to have two 80mm fans, one for each P3A board. These would provide effective cooling for when the amps are pushed hard, but should not need to operate under normal listening levels so the noise of the fans would not cause a problem.
The work is based on another ESP project. Project 42 is a thermo fan project that is able to run off the power amps power supply and happily operate two 12V computer fans whenever a set temperature is reached. Rather then use expensive temperature sensing components.. this works straight from normal rectifier diodes... and quite well too!
Thats about all that can be said about the initial ideas. The next page gives more information on the construction of the power amplifier.