Posted: Sat Jan 03, 2009 2:15 pm
You're very close, so here's my report card.
First off, when you try converting the signal to binary, you need to make a best guess as to which pairs are the logical 1s and 0s, and whether the signal should be read as MSB or LSB. This is a subjective judgment, so there is no correct answer as such, but once you make your selection you do need to be consistent with it.
In my opinion, the 1 pair is "+550 -1600" and the 0 pair is "+550 -550" and the signal is MSB. You appear to agree as the OBCs and device codes that you used in KM match this, and you have MSB selected as the signal style in KM, so so far so good.
The problem is in the PB file because there you have "+550 -550" selected as the 1 pair (rather than the 0 pair). UEI typically gets this backward also, which is why we usually treat their executors as COMP'd. In order for the executor to work with the 1s and 0s backward, you would need to select MSB-COMP in KM.
Also, I notice that you have the leadin selected as "1st frame only" whereas in the learns I see the entire signal repeating while the button is held, so this should be "Same Every Frame". You have 'Repeat Value" set to 1 to force a minimum of 1 repeat, this is redundant as you will always get 1 repeat of the signal, this feature is needed if the device requires 2 or more repeats even with a short press. Entering 1 shouldn't do any harm, but it does waste code. Finally, you have the leadout time entered as 54,650 uS with the "OFF as total" set to YES. As the leadout time is consistent regardless of how many 1s are in the data, this should be set to NO and the correct time entered, which is -25,600.
Regarding the original learns, it looks like you accidentally re-learned the 0 button when you tried to learn the 1 button. The shift-right learn shows us that the 3rd device code is 0 in shifted mode and 16 in normal mode, so you would need a 2nd upgrade if you wanted to include shifted functions. Alternatively, you could treat the variable data as 2 bytes to make it a combined upgrade.
The GUIDE, MUTE and INFO buttons are all the same, so that's either how it is or it's a learning mistake. Likewise for SWAP and CC.
Looking at your codes.txt files shows me that you didn't use the rounding feature in IR, so converting the pairs to binary would have been harder for you. I rounded everything to the nearest 50, which only gave me a couple of 600 times, if I'd have used 550 it would have been even better.
The following zip file contains my attempt at a PB and KM file, plus it also includes my timings text file and the spreadsheet that I used to get the OBCs.
https://www.hifi-remote.com/forums/dload ... le_id=6071
First off, when you try converting the signal to binary, you need to make a best guess as to which pairs are the logical 1s and 0s, and whether the signal should be read as MSB or LSB. This is a subjective judgment, so there is no correct answer as such, but once you make your selection you do need to be consistent with it.
In my opinion, the 1 pair is "+550 -1600" and the 0 pair is "+550 -550" and the signal is MSB. You appear to agree as the OBCs and device codes that you used in KM match this, and you have MSB selected as the signal style in KM, so so far so good.
The problem is in the PB file because there you have "+550 -550" selected as the 1 pair (rather than the 0 pair). UEI typically gets this backward also, which is why we usually treat their executors as COMP'd. In order for the executor to work with the 1s and 0s backward, you would need to select MSB-COMP in KM.
Also, I notice that you have the leadin selected as "1st frame only" whereas in the learns I see the entire signal repeating while the button is held, so this should be "Same Every Frame". You have 'Repeat Value" set to 1 to force a minimum of 1 repeat, this is redundant as you will always get 1 repeat of the signal, this feature is needed if the device requires 2 or more repeats even with a short press. Entering 1 shouldn't do any harm, but it does waste code. Finally, you have the leadout time entered as 54,650 uS with the "OFF as total" set to YES. As the leadout time is consistent regardless of how many 1s are in the data, this should be set to NO and the correct time entered, which is -25,600.
Regarding the original learns, it looks like you accidentally re-learned the 0 button when you tried to learn the 1 button. The shift-right learn shows us that the 3rd device code is 0 in shifted mode and 16 in normal mode, so you would need a 2nd upgrade if you wanted to include shifted functions. Alternatively, you could treat the variable data as 2 bytes to make it a combined upgrade.
The GUIDE, MUTE and INFO buttons are all the same, so that's either how it is or it's a learning mistake. Likewise for SWAP and CC.
Looking at your codes.txt files shows me that you didn't use the rounding feature in IR, so converting the pairs to binary would have been harder for you. I rounded everything to the nearest 50, which only gave me a couple of 600 times, if I'd have used 550 it would have been even better.
The following zip file contains my attempt at a PB and KM file, plus it also includes my timings text file and the spreadsheet that I used to get the OBCs.
https://www.hifi-remote.com/forums/dload ... le_id=6071