Page 1 of 1

Don't Want to Use the MIDI Interrupter/Fiber Optic Cable?

PostPosted: Sun Mar 24, 2013 3:19 pm
by nick@acata-tech.com
Here's a quick and dirty alternative to the somewhat expensive/fragile fiberoptic components. This allows you to send a signal from an Arduino, useful for testing with a fixed waveform 50 uS on time - 950 uS off time (or something of the sort). Or you've come up with your own MIDI device that you're not worried about getting zapped.

Image
Image

First you're going to need a right angle male break away header.
Image

Then you'll need to clip off three positions. The three positions will breakout the through holes that would normally be occupied by the fiber optic receiver.
Image

As you can see the spacing is wrong of the headers is wrong, too wide for the through holes of the fiber optic receiver.
Image

This is easily overcome by squishing the leads together (the shorter bent end) with a pair of pliers. This is most certainly just a guess and check procedure.
Image

Then squish the header into the holes with the pliers, being careful not to damage the any traces underneath, and solder. Voilà!
Image

The red is 5V+ (not necessary to connect if your microcontroller has it's own power supply), the blue is the signal (be careful of what pulse width/duration you feed it!), the black is ground.
Image

Re: Don't Want to Use the MIDI Interrupter/Fiber Optic Cable

PostPosted: Sun Mar 24, 2013 9:37 pm
by Bayley
Nice documentation of your mod! However, the fiber optic isolation is essential! The sparks are a huge source of noise, and a copper cable forms an excellent antenna. Logic-level pulses from the micro can easily be drowned out by RF noise, causing the coil to be on continuously (which we sometimes call CW mode) rather than carefully pulsed at low duty cycle, as is necessary to keep the IGBTs from overheating in a matter of seconds.

While your trick will work fine for testing at low powers, please use optical isolation when running at full power! Also, use caution when modifying pulse widths, as the improper settings can destroy the coil.