LM3886 Servo Design

Status
Not open for further replies.
There is less than 2 mV DC on one channel (my equipment can't read it), but 19 mV DC on the other. I'm wondering if I may have made a bad solder joint, because I feel that's a little high.

After making a change and re-checking everything on the board, I noticed the servo on the channel with more DC offset wasn't working. After a lot of comparison with the working board, I discovered I had inserted the clamping transistors backwards on one channel, and they were effectively shorting the input of the op-amp.

Not too much of a surprise, since the prototype boards I built mine with were designed for 4 pin transistors, but I ended up using 3 pin. Once I corrected this problem, both channels have <1mV of offset and it nulls each time I change the volume pot, as it should.

The mods I made were to put R5 directly on the LM3886 pins and to change C3 and C4 for 47uF 63V caps. The sounds is more full, smoother on the top end and with a bit more punch in the mid-bass.
 
Last edited:
Thank you very much!

Last, but certainly not least 😉

Could you post a few pictures of your build? I'm keen to seen the construction and how the mods were made.

Truth be told, I'm a bit of a beginner, so a visual reference will be great.

Kind regards,
Chris
 
I assembled the "newer" black and gold 1.1 boards shown in this thread last weekend. I had previously been using the green 1.0 boards for the last few years.

I put R5 directly on the pins of the LM3886, the 220pF cap across R2, and used 56uF for C3,C4 instead of 10uF. The 56uF and 1000uF caps are now Panasonic FR.

The difference is stunning! Less noise (I had some 60Hz hum due to a ground loop from using different power circuits), clearer highs, more prominent mid-range.

Interestingly, I had noticed that the sound improved quite noticeably over a few days as the FC caps I used in the original build were broken in, but this time I'm not sure there was any difference with the FR caps.
 
Status
Not open for further replies.