High Power IR remote.

[mrpackethead]

Just musing on a design idea at the moment for running a very high power IR remote system.

I have some Infrared leds that I'll run in series, and will need about 1A current.     Driving that is not really a problem and i have a suitable Buck Switcher that runs at about 500kHz or so.  If i modulated that with a 38khz control signal, i am thinking that the filtering on the receivers probably will filter out the 500Kz and they will only see the 38khz..

Before i go and try this on the bench, anyone want to hazard a guess at what might happen?

[NiHaoMike]

It should work if you set the bandwidth of the low pass filter to somewhere between 500kHz and 38kHz. 50kHz would be a good start.

[SeanB]

Depends on how the converter handles the modulation. You might get slow switching that effectively removes the modulation, due to built in soft start circuitry. You probably will want to convert to a fixed high voltage and have a big capacitor, and then use a switched constant current sink for the LED's instead. That way you get a good 38kHz signal that the receive modules ( I assume you are using the 3 pin transistor shaped IR modules) will work well with.

[paulie]

Why would anyone want to filter output from a buck converter in this application? A 500khz switcher outputs DC not 500khz.

[mrpackethead]

Why would anyone want to filter output from a buck converter in this application? A 500khz switcher outputs DC not 500khz.

Doh, brick hits me in the head.  Not sure where i was going with this at all, Somehow i simply forgot the fundamentals of what i was working on.  Thank you for pointing that out to me.   I ran this through the sim and what i get is a a small amount of ripple overlaid on a DC signal.  That unlikely will be a big problem, and will certainly be rejected by the receiver.

Now to see if we can get the controller to Dim Fast enough directly, without getting aliasing..

[ajb]

There are IR hearing assistance systems (used in theaters and such) that modulate massive arrays of IR LEDs with enough bandwidth for stereo audio, so it's definitely possible.

This system apparently uses a 2.3-3.8MHz carrier 

[NiHaoMike]

IRDA has specs for up to 1Gbps, but it never caught on due to its requirement for line of sight. What would work nicely is gigabit IR in overhead APs as a way to offload traffic from a more conventional RF wireless network.

One good control strategy is to AND the 500kHz switching frequency, 38kHz carrier frequency, and data signal together. Use that to set a positive edge triggered flip flop, with a current comparator resetting it as soon as the set current is reached.

[MixedSignalSystems]

IRDA has specs for up to 1Gbps, but it never caught on due to its requirement for line of sight. What would work nicely is gigabit IR in overhead APs as a way to offload traffic from a more conventional RF wireless network.

One good control strategy is to AND the 500kHz switching frequency, 38kHz carrier frequency, and data signal together. Use that to set a positive edge triggered flip flop, with a current comparator resetting it as soon as the set current is reached.

This idea (or at least a very close relative) is being researched and commercialised as "LiFi" or "Visible Light Communications (VLC)".

A Scottish company called pureLiFi, among others, is working on the concept:
http://www.purelifi.com

They have some pretty cool demos streaming HD video and other such novelties.

[Yansi]

One good control strategy is to AND the 500kHz switching frequency, 38kHz carrier frequency, and data signal together. Use that to set a positive edge triggered flip flop, with a current comparator resetting it as soon as the set current is reached.

As it was said, 500kHz buck outputs DC (with small ripple under a few % when designed properly), not AC 500kHz. The only issue is, if the buck can switch on or off fast enough, which almost without a doubt cannot at 38kHz. It is only a few cycles more than 10. That will not work properly.
Some switchers are designed for PWM modulation, but not at tenth of their switching frequency.

[NiHaoMike]

You're really building a specialized class D amplifier.