Arduino standalone MIDI interupter - Interference issues. / Tesla Coils / Forums

Forums

4hv.org :: Forums :: Tesla Coils

« Previous topic | Next topic »

Arduino standalone MIDI interupter - Interference issues.

1 2 next

Move Thread

LAN_403

Doug86

Tue Nov 10 2009, 06:11PM

Registered Member #2424 Joined: Tue Oct 06 2009, 08:02AM
Location:
Posts: 17

Hi all,

Ive writen an arduino sketch that reads MIDI data on the serial input and uses it to generate squarewaves of the apropriate frequency (it also uses the velocity byte to set the pulse width) and i'm using it as an interrupter on Audio:Deviant's coil.

This simple monosynth works perfectly on the bench but as soon as I power up the coil the midi line picks up so much interfearance that it disrupts the communication.

The midi data is being sent from my interface, over a short butchered midi lead, through a 4n25 optocoupler, to the arduino. I found good direction on midi>arduino here

The arduino's signal is then inverted twice (as a kind of buffer) before being sent down a coax cable to the coil.

Ive tried grounding the shielding of the midi lead to AC ground with maybe slight improvement but still unusable. If I hard code a tune into the arduino (no midi control) then it plays fine and drives the coil nicely.

Ive seen on these foums that a few of you have run coils from midi before, how did you get round this issue?

Many thanks in advance.

Doug

TESSiER

Tue Nov 10 2009, 07:28PM

Registered Member #2453 Joined: Fri Oct 30 2009, 09:10PM
Location: Surahammar, Sweden
Posts: 3

I would go for fiber optics :)

Goodchild

Tue Nov 10 2009, 07:46PM

Registered Member #2292 Joined: Fri Aug 14 2009, 05:33PM
Location: The Wild West AKA Arizona
Posts: 795

Try moving the arduino far away from the tesla coil and also put it in a metal box and run it from an isolated power source such as a battery, and one last thing replace the coax with fiber.

also take a look at steves old setup

Hope this helps.

Also if it's not to much to ask may I see your arduino code I have been working on almost the exact same thing for my DRSSTC and I am not the best at writing software.

Thanks

Doug86

Tue Nov 10 2009, 08:34PM

Registered Member #2424 Joined: Tue Oct 06 2009, 08:02AM
Location:
Posts: 17

Thanks guys,

The link between the Arduino and the coil is ok and rejects enough inteference to work (for the moment). We will eventually replace it with fiber.

If I hard-code a melody into the arduino it plays through the coil perfectly.

The problem I'm having at the moment is with the midi data into the arduino. The midi line is opto isolated from the arduino and coil so im not sure the fiber will have a massive advantage. We need to be able to connect an unmodified midi device to the system.

where should I be grounding the shield for the midi line?

I'll try running a longer coax so that the midi controller can be further away from the coil...

Goodchild, I'll post up the source code for you when I get home.

plazmatron

Tue Nov 10 2009, 08:57PM

Registered Member #1134 Joined: Tue Nov 20 2007, 04:39PM
Location: Bonnie Scotland
Posts: 351

Without a doubt, fiber optics is the way to go with these. Any amount of wire will act like an antenna, and cause all manner of badness with ÂµC's, the worst being outputs that go high and stay high, killing FET's

My MIDI interrupters run on Arduinos too, but the signal is sent down Toslink to the DRSSTC's

I just showed a group of friends these musical coils the other day, so I'm still basking in the warm glow of extra geek points

GeordieBoy

Tue Nov 10 2009, 11:16PM

Registered Member #1232 Joined: Wed Jan 16 2008, 10:53PM
Location: Doon tha Toon!
Posts: 881

First of all, put the micro board in a metal box to shield it from radiated EMI. Once this is done, all you need to worry about then is conducted EMI...

From a conducted EMI point of view the micro box has at least two cables going into it: The MIDI cable from the MIDI master, and the cable coming from the TC. Both of these will pick up interference and carry it inside the micro's shielded box. If you are using an external power supply for the micro then this is a third cable that will also pick up EMI and carry it towards the micro. All potentially leading to either a software crash or corrupted MIDI data.

The MIDI cable already uses an opto-isolator so unless you are breaking down the isolation barrier with several kV, I would suspect that this cable is not the problem

That leaves the interrupter cable from the TC and the power connection. You can easily investigate the potential of conducted EMI down the power cable, by powering the micro from a battery inside its shielded box. If this works, then either stick with the battery power or put a filter, ferrite chokes etc on the power feed cable.

As for the cable to the TC, this is by far the most likely offender because it goes close to the operating TC's electric and magnetic fields. As others have suggested an optical fibre connection will almost definitely fix this problem for sure.

If you replace the cable from the micro to the TC with several metres of optical fibre, you can move the micro, it's power supply and all of the MIDI cabling well out of the interfering fields from the TC.

If you are set on sticking with a wired connection, then look at filtering. In particular use a tightly twisted pair, and wrap this wround some honking great big high-permeability ferrite cores at each end of the cable. These common-mode chokes at each end won't effect the wanted differential-mode signal from the micro to the TC, but they are just about as good as you can get for blocking common-mode interference going in the opposite direction!

Just some thoughts from someone who used to work in EMC approvals. Hope this helps,

-Richie,

Avi

Wed Nov 11 2009, 01:39AM

Registered Member #580 Joined: Mon Mar 12 2007, 03:17PM
Location: Melbourne, Australia
Posts: 410

TESSiER wrote ...

I would go for fiber optics :)

thats what i had to do, i used a toslink cable. However the receiver module may be difficult to work with. Also this needs to go in a metal box too.

Goodchild

Wed Nov 11 2009, 01:52AM

Registered Member #2292 Joined: Fri Aug 14 2009, 05:33PM
Location: The Wild West AKA Arizona
Posts: 795

Doug86 wrote ...

Goodchild, I'll post up the source code for you when I get home.

Thanks very much!

Wolfram

Wed Nov 11 2009, 07:24AM

Registered Member #33 Joined: Sat Feb 04 2006, 01:31PM
Location: Norway
Posts: 971

In addition to what the other posters here said, you might want to change your optocoupler for something more suitable.

I'll quote myself from an other thread:

I have also used common optoisolators for MIDI, but this has given me a lot of trouble. It seemed to be working well together with some MIDI controllers, but would not work, or only work intermittently with others. The recommended PC900/6N138 are darlington optocouplers, so they have a much higher current transfer ratio than common optocouplers. Common optocouplers will work fine together with stronger midi outputs, but with weaker ones it will work intermittently or not at all. I tried adding an external transistor to the optocoupler, to make it into a darlington, and while this improved the situation, it still only worked with about half of the devices I tested it with.

I suppose the low gain of the 4N25 could make it more noise-sensitive too. The typical current transfer ratio of the 6N138, which is the usual choice for MIDI, is 1100%. Compare this to 50% for the 4N25.

Anders M.

Doug86

Wed Nov 11 2009, 01:05PM

Registered Member #2424 Joined: Tue Oct 06 2009, 08:02AM
Location:
Posts: 17

Thanks to all, I've ordered the fiber optic and some 6n138s and will report back.

Here's the source code to my arduino although:

THIS VERSION HAS STILL GOT A BUG!!!!

Ive realised that if the pulse width is longer than the interupt timer i'm using then the interupt is missed and the frequency drops a little. this version works but produces notes that are out of tune at longer pulsewidths. I'll repost with a better version when I work out how to stop it.

Apologies for the messyness of my code but it's still work in progress :)

You need to install the MIDI library from here

for this to work

I also used code from the "melody" example that comes with the arduino software.

EDIT : CODE REMOVED, PLEASE USE IMPROVED VERSION BELOW.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob