Search This Blog

Wednesday, August 1, 2018

Audio sw - project logs - part 2

Projects / Audio Switching Original post date: 06/24/2016

Looks like I bugger up the stereo jack footprint. I fixed the connections. (See bottom of page)
Other than that I haven't find any major problems yet. The program runs on this right away which means most of the stuff are working. That row of low pass filter caps on the top side of the ARM chip seems to help with the noise issues I was having before.

I have decided to build the prototype using toner transfer as it would have cost me $20 for this small PCB. I have been mostly ordering PCB for the last few years, so not at the top of my game.

I am using a 12/12 design rule until I get a few more rounds making PCB again. The TSSOP has a pin pitch of 0.65mm (25.6mils), so a 12/12 rule is just good enough to break out the part. I used 16 mils track for the audio signals and 40-mils for some of the power. For the area fill, I use 12 mils isolation. The problem with tone transfer is that sometime when I used too much heat, it swells up the tracks, so the extra space helps to prevent shorts.

I print out the artwork (flip horizontal) on the Ikea Flyer I got in the mail the other day. The paper has a special glossy coating and the paper just separate from the PCB after a bit of soaking. Unfortunately they have a bit of dark spots that got transferred to the fine track near the ARM chip.

I find it easier to fix stuff after the etch as the toner is pretty flaky and I cause more damages in fixing the shorts. I patched up some of the fill area with permanent markers as the replacement toner for my laser printer isn't that great.

Toner is transferred onto bare PCB to protect the tracks from etching

Here is the PCB almost done after soaking on a heated bath of Ammonium persulphate. The blue solution is caused by the dissolved copper ions.

The exposed copper is etched by chemical reaction
After cleaning off the toner with a scouring pad, I leave it in the water for a while for washing off the chemicals. Not a perfect PCB, but something that I can patch up later tonight.

After etching, toner is removed and the copper tracks are now exposed.
The caps are a bit too close together and I got the stereo connectors flipped. :(

I have decided to tin the whole PCB. I soldered in the rest of the parts.
Now to tackle the wrong footprints. I drill a hole where the pin should go and made a pad by cutting the off thin strips of copper isolating it from ground. Notice that one of the pads rings was half detached. I am going to fix it up.


I use some wire wrapping wire (30AWG) to wrap around the pin. This helps to anchor the pin to the PCB. This trick also helps a lot for to get a nice fill on a home made PCB with these flat pins as there are no plate through hole nor anything for the solder to bridge all the way to the pad.

Similarly, the misplaced pin is connected to the signal via a wire.


Here it is the finished prototype.



There is a front end analog circuit daughter card with the balanced/unbalanced converters and analog MUX that fits between the headers.

This is how it would fit together eventually. I haven't soldered in the bottom connector to a tall header pin.


The rest of the parts seems to checks out:: +5V output, negative voltage charge pump, +/-5V regulators for the analog module.

I guess I might need to make more PCB at home as Canada Post might go on strike. The backlog of packages from China is going to get much worse than what I am seeing right now.

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.