short range wireless data transfer at 3 Mbps schematic ?

[maniana]

I am looking for a schematic or just an idea of how to build a short range (5cm) wireless data transfer system with a speed of at least 2.5 Mbps. It has to be cheap. Data transfer over coupled coils.
Can anyone suggest some solution ?

maybe there is ie. X10 interface with that speed ?

 

[FvM]

You can refer to baseband pulsé transmission, using an encoding method to get a DC free signal, e.g. manchester encoding or 8b/10b. Or basic ASK or FSK modulation.

 

[maniana]

ok, but I think ASK or FSK at 2.5Mbps with a large coil (has to be at least 20cm in diameter) may not work.
I was hoping to get some help on how the analog TX driver and RX front end should look like in this case.
I am building the TX coil driver with DSL line driver chip but I am not sure whether this will work. And I have no Idea at the moment of what the RX aplifier should look like.

 

[jiripolivka]

Induction coil has to be built so that it only resonates with circuit capacitances above ~ 5-10 MHz.
Why not use optical coupling? There are optical IR links that carry > 10 Mbps over several feet. They use LEDs and PIN photodiodes. Try to google available commercial models.

Over 5 cm you can possibly use a capacitive transfer, by trying fast opamps to pick up several mV,either directly your baseband or modulated on a suitable carrier, from 1 Hz up to 900 MHz.

 

[FvM]

ok, but I think ASK or FSK at 2.5Mbps with a large coil (has to be at least 20cm in diameter) may not work.

Something you should find out. I assume it can work for the time being. Of course I don't know the constraints you'll reveal next.

 

[maniana]

@jiripolivka Optical link is not an option due to the fact, that the air gap can be dusty - reliability problems.
@FvM you are right, it works for the time being. I only have a problem that I gat lots of transmission errors as there are high power electrical motors working inside the same box. So simply ASK works but is not reliable, and the air gap has to be <1mm. I need at least few milimmeters more. I was driving the TX coil with a mosfet driver MC34152. I know that the coil design is very importand here. I plan to use some ferromagnetic core for the receive coil to improve coupling (cannot be done for TX coil). RX coil has to be outside the perimiter of the TX coil.
And I should add that everything is made of metal on which the coil has to be mounted.

 

[FvM]

Every post brings new constraints. It's effectively useless to make detail suggestion without a complete application specification.

Of course I don't know if the suggested encoding and modulation methods will work for you. This depends both on application conditions unknown to me and "art of the electronics".