AIUI, joerg_rw meant, that IrDA and CIR can cross-talk, albeit at reduced sensitivity, + as we know, IrDA have much weaker beam power than CIR (ywt, works at higher speed).

Thus, having CIR TX diode *should* ensure no-problems with RX of other devices, as beam power will be orders of magnitude greater than normal sending power for IrDA - compensating for loss of sensitivity due to color offset.

Now, it *could* be a problem to *receive* signals from other devices, if we would use receiver tuned to colors - that's why using IrDA receiver is mentioned. Now, if two Neo900 are able to talk with each other, they should be able to talk with other IrDA's too - as both Neo900 uses standard IrDA receiver, which implies TX CIR diode is capable of IrDA'ing (as other's side standard IrDA receiver decipher it correctly).

By the way, it shouldn't cripple capability of using Neo900 as "learning" remote, too - IrDA receiver will be less sensitive to CIR, but for learning remote purposes, we can put "to be copied" remote at point blank, so it is OK.
---

Now, not that I did any scientific tests to reinforce all of the above - I tried to use pure logic here. If anyone knowledgeable about IrDA/CIR notice some catch, feel free to correct me.

From earlier entries in this thread - moving around will be necessary, but not they way you're thinking about it. It's exactly the other way around - they're actually moving parts around PCB, to make it fit with N900 body

/Estel