[mythtv-users] USB dual channel IR blaster?

[Mark Knecht]

On Mon, Feb 22, 2010 at 12:18 PM, Travis Tabbal <travis at tabbal.net> wrote:
>>
>> +2 for a dual channel device. Possibly only +0.5 for a single channel
>> device.
>>
>> What might a device like this look like on the transmitting side? Is
>> it an external USB device? Plugs into any USB slot? Has two ports on
>> it to hook up to two IR transmitter devices? What IR transmitters
>> would you support so I can check out the costs on those? Can I use
>> multiple devices if I want to control more than 2 STBs?
>>
>> I'd certainly pony up a little money (say $25-$50) to help with your
>> initial PCB costs, etc. I assume that if and when I works I get at
>> least a 2-channel device to own, etc.
>>
>> As an alternative a 2 or 4 channel 1x PCI-Express card for internal
>> use might be of interest if you already know of an appropriate
>> microcontroller and want to go that way.
>
> It could be made to meet all your requirements. Internal USB headers are a
> possibility, but they don't seem to be 100% standardized like the external
> plugs are. As for transmitter devices, they are nothing more than an IR LED
> on a stick with a mono 1/8" jack on the end for the most part. Just about
> anything would work in that case. I could easily drive about 8 of them
> without issue, possibly with different signals for all 8. Once one of them
> works, it's really not a big deal to add more. You could even build your own
> if you wanted to, though I know I've seen these things pretty cheaply out
> there on the net. irblaster.info has them for about $6, they look to be the
> same as the ones Smarthome sells.
> PCIe is a real pain to talk to. USB is far easier and significantly cheaper
> to interface to for this level of project. We don't need the bandwidth, so
> there's no reason to deal with the problems, IMO.
> I'm willing to do the hardware and firmware as an OSS project if others want
> to contribute either funds or code. The biggest code help that would be nice
> is someone to write the LIRC driver. It's going to take me some time to get
> that part working if I have to do it myself. Or is there an easier way to
> get Myth talking to it? The simplest method to do the firmware I can think
> of right now is to make it look like a serial port to the computer, so you
> just open /dev/ttyS0 or whatever the device name is and send it commands
> like "SEND CH1 <IR timing string>". We could even toss a receiver in while
> we're at it. May as well cover all the bases. It shouldn't be hard to be
> very competitive with the existing products price wise with the parts I have
> in mind, and it would be OSS, so no need to worry about binary drivers and
> such.
> If there are a few interested people out there, please contact me directly
> so we don't clutter the list. I can probably knock up a prototype on the
> bench pretty quick. PCBs will take a little time to be made after I can
> verify correct operation.

I suspect that USB would serve a wider community and even in my case
our current server is a PPC-based Mac Mini which wouldn't accept any
internal cards so it's probably better for me anyway.

If you went with USB then I wouldn't personally focus on internal USB
headers or anything like that. I'd probably go for some little
external box, all female connectors and the user supplies a USB cable
and two IR transmitter thingy's that you specify ahead of time.

Cheers,
Mark