XBee wireless bi-directional MIDI link

[Sasha]

Very cool XBee project from Ladyada

http://www.ladyada.net/make/xbee/midibee.html

[Smithy]

Awesome!

I was worried about the range you could get with this thing,

but jesus christ!

Look at the max ranges for the different modules!  :o

http://www.ladyada.net/make/xbee/modules.html

[frailn]

15 Miles!? Good Lord.

Hmm. After hours, get up in the drop ceiling at work, wire up the audio of my monosid to the PA speaker in my office. Next day, come in with my wireless XBee and start sending SID sounds throughout the office randomly...

[Sasha]

Awesome!

I was worried about the range you could get with this thing...

Only thing I was worried is latency, but it is just 5.5-6ms. :)

It is not very cheap to make, but it is much cheaper than commercial products.

[SLP]

Are there any handshake-routines implemented?

A corrupted SyX-transfer could be a bit bad actually...  :P

It would be cool to get it work without PC. like the M-Audio wireless MIDI-stuff

[Smithy]

A corrupted SyX-transfer could be a bit bad actually...  :P

Yeah, youd prob end up taking down a plane or something!  :P

[Sasha]

It would be cool to get it work without PC. like the M-Audio wireless MIDI-stuff

It can work without PC. I`m not interested into using it with PC myself, if I would I can always connect it to the GM5. ;)

[ultra]

connecting to the gm5 was my first thought too.  we had to write a paper at work on various wireless communications, and this was one of them.  but i never thought of using it for midi.

don't you think the latency might be a problem?

[Sasha]

don't you think the latency might be a problem?

5.5-6ms?

[Smithy]

5.5-6ms?

Not if you dont mind the latency of a decent audio interface, which would be about 5-6ms id imagine, without artifacts in the sound of course.

[johnh]

I just finished putting one together:

I haven't done a lot of testing yet but it seems to work. 

[Sasha]

Cool! Let us know the results of testing.

[stryd_one]

I'm really curious about jitter on this (latency timings are given but jitter is not - jitter will be a big issue with clock signals for example) ... look forward to the test results!

[johnh]

I haven't done a lot of testing yet but it seems to work. 

Well, I've done some testing now and have since upgraded the original XBee's to the Pro version in an attempt to get better range.  I'm using the ones with 'chip antennas' and I understand you can get much better range if you use proper external antennas.  The Pro radios work around the lower level of my house, around/through walls, etc.

I've pumped a lot of MIDI sequences through without any obvious issues, and would be happy to do some real testing if y'all let me know what you want to see and how to do it.  I'll give it a shot at least.

As an aside, for MIDI I'm using a Roland serial driver and not the Korg driver mentioned on LadyAda's site.  I couldn't get the Korg driver to work properly.

[Sasha]

Glad so hear it works good!  :) I have no idea how to test it more intensively.

[stryd_one]

Fun! As for testing.... do you have a scope?

[johnh]

do you have a scope?

I could borrow one...

[stryd_one]

Good news :) The reason I ask is, that you can't really use a PC to test timing, because the timers on them that would be used as the time reference...well... suck. Likewise they can't be trusted to generate the signals you're testing... fortunately we have midiboxes for that ;)

Probably the easiest way to start a test, would be to push midi clock from a midibox, across the xbee link, and use your scope to measure the serial output pin on the receiving-end xbee. Because midi clocks are 0xf8, you should get a fairly decent square-ish pulse out of it every time a clock message is rec'd. Then I guess take a walk around with them and see how it fares...