Sorry about these advertisements too. They come with the territory.

LM3886 Amplifier

The plan

After the research phase, as outlined elsewhere, I set about building the amplifiers for the Orions.  Using Linkwitz's suggestion that the LM3886 provided adequate amplification for each of the separate drivers, I went to the DIYAudio site and chose CarlosFM's design (Make sure you read the whole thread to find the most recent version. It has 0.3 ohm values for the resistor snubbers.) The fundamentals are:

1. the National LM3886 chip, five obtained as samples direct from the company,
2. my own circuit board design using Circad98,

1. Short feedback loops (note the fb resistor is essentially on the pins of the LM3886 chip

2. The "last" filter cap VERY close to the chip itself (the large electrolytic black "can". This was confusing, but it is what CarlosFM call these capacitors on his schematic diagram.)
3. V+ to V- capacitor, (the white rectangular one) even closer!

4. The big blue rectangular block is the 2.2uF input capacitor. There is lots of room for "bypass" capacitor (on the TOP side) or other capacitor options;
5. Having large enough pads at the feedback/ground resistor network so that "hifi" resistors from Texas components can be put in if needs be.
6. Dual layer board, with lots of "through wires", because I cannot do "through vias" on a home made board.

Here are the internals of the stereo amplifier up and going:

This is a little close-up of my homemade bridge, all attached:

I think in the next versions though, I will not make the board double layer, because of the extra soldering by greater than 2 times, and because someone has suggested to me that I might be setting up the conditions for radio frequency feedback to cause oscillations ... the first anyway is a wonderful reason, and I do not think I need tracks as large as I have for 50-60 Watts B(?)-class amplification.

The tracks and pads are quite large, so the soldering is not as beautiful as it might be.  Certainly it takes a definite but slightly longer time to heat the tracks before I roll the soldering iron onto the component pin.  So far, though, all the components work!

The power supply is CarlosFM's unregulated final version.  Make sure you read the whole thread to find the most recent version.

Once I had put together the capacitor bank, my home-made ultrafast bridge rectifier, and worked out how to wire it up, I was busting to test it, and I came up with the following test setup:

Note the method I have used to mount the two LM3886 onto the heatsinks ... it has the advantage that there are no holes in the flat heatsink surface so that it can be easily re-used over and over again.

To make the PCB's, I used the home-brew Toner Transfer system as I have said elsewhere.  I found that using the commercial Pulsar Toner Transfer system WITHOUT the green foils, that the etchant (aluminium persulphate) often ate through the plastic from the laser printer a bit too easily, producing a pitted (but still functional) trace.  I also purchased the laminator which the Pulsar bloke suggests, but the one in Australia does not go hot enough to make the toner stick to the circuit boards. I still use an iron (with a slowly rotating broomstick) to get the toner to adhere to the PCB copper

Web Hosting · Blog · Guestbooks · Message Forums · Mailing Lists
Allwebco Web Templates · Build your own toolbar · Site Building Articles · Audio, Fonts, Clipart
powered by bravenet.com