Probably they don't exist.

But I think the device uses commands from a set of 256 signals. With the MakeHex program you can generate a clean set of all 256 signals. Then with the IrPanels program you can convert that into a CCF file with 256 numbered buttons. Then merge that CCF into your own CCF, load into your Pronto, and test all 256 buttons to see what they do. If discretes exist you would probably find them.

Read this first:
[Link: remotecentral.com]

Using MakeHex for this IR protocol is probably far from obvious. It would take me some investigation time to figure it out and I wrote MakeHex. I'll try to find some time soon to investigate and explain. (For most other IR protocols, the readme file in MakeHex.zip should tell you more than enough).

I'll probably want to see a few of your working learns. For MakeHex I need to know what really correct signals look like. As I explain below, signals in this IR protocol might work without being correct. I expect the CCF you found is correct, but to be more confident, I'd like to compare a few signals you learned that work to the same signals in that CCF that work.


Good luck. I wouldn't expect much from customer support.

I do know why most learning remotes fail to learn these signals. I'm less sure of why some learning remotes sometimes learn them well.

Almost all IR signals (including these) are based on a series approximate durations. The learning remote measures and stores the durations. But each duration tends to be slightly mismeasured.

Almost all other IR signals use just a few different durations and no two durations are close to each other. So if a signal includes several durations that are close to each other, you can conclude that they were originally all the same duration and were mismeasured. Then you can correct them all to some value in the middle of the grouping and thus store a more correct signal than you actually captured. Most learning remotes do something like that.

But this IR protocol uses many different durations, which don't differ from each other by much. When a learning remote sees a group of durations that are all close to each other, that likely contains two different original durations. So when it "corrects" them to some middle value some (maybe all) of them become wrong.

To further complicate things, this IR protocol has signals that are unusually long and complex and much of the information in each signal is ignored by the actual device. So you could learn a set of signals and have at least one error in every learn and yet several of those signals would work anyway because their errors happen to be in parts of the signal that the device ignores.


I wouldn't bury that question (as you have) in a thread on a specific different topic.