I am having great trouble learning ir codes from my ir keyboard. Looking at the hex it seems that the carrier frequency changes. Is this likely? Or am I just getting bad data?

Any tips would be great, thanks.

Regards
Jamie

What keyboard is it? Keyboards generally have two or three commands per key. A make command when the button is pressed, a hold command if you keep holding the button down, and a break command when you release the key.

To learn the key initiate the learning process in the Pronto, then give the key a sharp tap and release -- that way you will get the entire sequence. Sometimes you learn just the make and break command which is usually OK. If you learn like you normally learn then often only the hold key gets learned and on some keyboards that's the same for every key.

If you are trying to do compound key strokes, like Alt-Enter, then the problem gets harder. But generally we can analyze the command structure and create separate commands and then you can do the sequence in a macro. To understand for alt enter to work you need Alt(make) Enter(make) Enter(break) Alt Break.

Usually you only just need Alt(make) Enter(make) since the computer executes when the Enter is pressed.

I have decoded a number of IR keyboards and I may be able to help if it's something I've seen before.

Hi there,

Well I will have to find a link to the kb I am using, as it doen'st have a name on it! I bought it about 4 years ago and don't have the box either.

A quick search on the net didn't reveal anything. Chances are you probably haven't come accross it, by its lack of popularity!

To explain how it looks, just in case 8-) It is a laptop style, with the thumb operated mouse about the "B", and another on the top right of the keyboard. It has two mouse buttons at the base and on each side. The board protrudes around these keys also. It also has multimedia function buttons accross the top.

Thinking about it, I should have a driver CD somewhere. I will check and get back.

Thanks for the help.

Oh, btw, I don't need to do multiple key presses. Just standard one button presses.

Cheers

Jamie

Most IR keyboards operate at 38KHz and/or 56KHz (presumably for the Mouse). If your keyboard doesn't operate @ 38KHz, you may have trouble learning it.

Hi, sorry I took so long getting back. I can't find anything on the keyboard. I have a digital pic though, but how can I show it here?

According to the carrier hex the main keys are on 72khz (0048). So do you think this will prove to be a problem?

Oddly enough my old RU890 learns the IR much better. In fact I have been using it to "catch" the signal then cut and paste the hex from prontoedit4 to PPENG.... Seems a bit strange to me.

Thanks guys

Regards
Jamie Wilson

If it were me... I'd learn all keys necessary on the 980, save a CCF, then import this CCF to PCF and merge with your current PCF.

Hi Lyndel,

I'm working on that now! Seems the best way to do it. Good thing I didn't sell my 980! Yet...

Anyway, I need the same thing on a couple buttons as you helped me with on the lexicon volume hex. That is, when I hold the button down (cursor in this case) I can keep scrolling down the page, instead of having to press the button seperately about 100 times.

I have read the hex article you suggested, and I'm not sure I am following all of it. Certainly not well enough to edit my hex code to fix my problem. Can you please help again!?

Here is the code for the two button presses.

0000 0049 0009 0000 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045

0000 0049 0009 0000 0076 0045 00A5 0075 0018 0045 0046 00A5 0018 0BB1 0076 0045 00A5 0075 0018 0045 0046 0045

Thanks again

Jamie

Jamie,

Burn me once, shame on you. Burn me twice, shame on me. Next time, it's gonna cost ya. :P


Start with the first code you need and apply the same pattern as follows...

0000 0049 0009 0000 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045

first 4 sets of 4 digits are as follows:

0000 - 0000 signifies Raw Code (usually Never Ever touch this)
0049 - 0049 is the timebase multiplier (typically Never touch this)
0009 - 0009 is the # of non-repeat IR pairs that follow
0000 - 0000 is the # of repeating IR pairs that follow

Here are the pairs...

0076 0046
00D3 0045
0018 0045
0046 00A5
0018 1333
0076 0045
00D3 0045
0018 0045
0046 0045

Get the picture?

Note that 00A5 above is hexadecimal representation for decimal 165. Use Calc.exe to convert in scientific mode if need be. Why is this important, all these entries are in hex. For example, a length of 0012 is not 12 decimal, but rather 12 hexadecimal which is 18.

So, if you had a code with a non-repeat length of 12, you would find 18, not 12 pairs following.

Now, what you need is a one-shot non-repeat plus repeat. What you do is copy the pair data first such that your code now looks like this.

0000 0049 0009 0000 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045

Then, adjust the repeat pair count for the # of pairs (in your case 9)

0000 0049 0009 0009 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045 0076 0046 00D3 0045 0018 0045 0046 00A5 0018 1333 0076 0045 00D3 0045 0018 0045 0046 0045

So your original code becomes both one-shot and repeatable.

There is no freaking magic here. I did not pull IR code data out of thin air. I simply used existing data, and duplicated it to add a repeat sequence. I did the same thing for your lexicon volume code.

Hi Lyndel,

What can I say!

Thanks for clearing that up. I feel I have a pretty good grasp of how the ir code works now. But, it doesn't work!

I tried holding the keyboard key down a little longer during capture, and I get a different code, which is understandable. But this code does the same thing. Even when I put a repeat in.

On a different note, is there a reccommended IR keyboard, that is known to be "easy" to capture? It might just be the easier solution.

PS Lyndel, your time is really appreciated!

Cheers
Jamie Wilson

Search the Pronto Files section for IR Keyboards. I know Dale Crawford learned one at one time or at least for sure, designed a Keyboard CCF.

Lyndel

I wouldn't normally think of decoding those learned commands this way, nor do I think two commands are a sufficient sample to leap to this conclusion, but based on having decoded a lot of keyboards, my best guess at this point is it uses an Async sort of protocol. John Fine figured this out a couple of years ago, and most IR keyboards that I have decoded use a variation on the theme. To date they have been all 1200 baud and this one looks like it's 2400. That means that a pulse of -417 uSec (Off) = One and 417 uSec (On) =Zero. This is different from the normal burst pair concept.

If so (and I don't have enough data or spent enough time to figure out the byte boundaries) here would be the binary:

First command
00000111000000000111011100011111110 Make
00000111000000000111011100011111111 Break
Second Command
00000111000000011111011100011111110 Make
00000111000000011111011100011111111 Break

These are separated at the big gap between the two segments.
Notice they are identical, with the exception of the last (bottom) bit, and both commands are consistent. As I mentioned earlier, keyboards generally have a make and break command. The first is when the key is depressed the second when released. Usually one bit changes to denote the change.

There is also a hold command for most keyboards. Sometimes hold is the same command for all keys and sometimes it's the make command that repeats. Jamie, if you would learn the same two commands only hold the key down longer, then I could possibly see this repeat command. My guess is that you need the Make command once and then a repeating hold command. Most keyboards are forgiving in that you can ignore the break command that the Pronto just can't do (if you want the repeating hold command).

FWIW the carrier frequency is 57KHz, and I haven't decoded a 57 KHz keyboard.

I could be wrong about all this but it's worth a shot.