Maybe a stupid post but I would appreciate if someone could enlighten me and explain why the below workaround solved my problem.

I have a very old Cable box (Macab) and the learned codes (directly to my Pronto TSU7000) worked very erratically. When changing channel (3 digit input) the box accepted one, two or three digits but there was no way telling when and how they would be accepted. I tried with various delays before, between and after each digit but that didn’t seem to matter or change the way the box interpreted what was sent.

The codes learned by the Pronto were recorded as follows:

900A 006D 0000 0001 875E 01FE (digit 1)
900A 006D 0000 0001 875E 02FD (digit 2).

After spending the better part of the weekend and the two last evenings trying to solve the problem I came across a post ([Link: remotecentral.com]) in which John Fine helped out another poster. Although I didn’t really understand what he was saying (still doesn’t) I gave it a shoot, followed his approach to the letter and succeeded in converting the 900A´s into hex, copied and pasted those to my IR-page and now everything works like charm. The thing that I don’t understand is why this worked since when copied in the long strings converted to 900A strings (the very same strings as I had before). Why is this?

Regards

Hans

What you report doesn't fit the way I understand these things to work. I can't explain why it solved your problem. I don't even believe it solved your problem. Truely controlled tests, in which you changed only what you meant to test, are quite rare. If a test gives an improbably result then you probably didn't test what you think you tested. (I've been wrong about that sort of conclusion several times before. But more often I'm right and It's the best place I know to start).

In that other thread, the issue seems to be the unusual frequency of the correct signal and the fact that the NG Pronto ends up ignoring that frequency and using the typical frequency instead.

That can't be your problem, because your original signals were already at the typical frequency, and there are LOTS of samples of the same signals in various CCF files (for both Macab devices and other devices that use the identical code set as Macab) and every single one is at the typical frequency.

In that other thread in post 10 I gave a very specific proceedure which results in signals which will be at that non typical frequency and which will NOT be converted back to 900A form by PENG.

You seem to say that you followed that, but that the result was converted back to 900A form.

Yes I followed that. I took the NEC1.irp and changed device number to 135.94 (that’s the only thing I did with that file). I then and dragged and dropped the new irp to MakeHex and that came up with the following hex for digits 1 and 2.

Function 1

0000 006D 0022 0002 0157 00AC 0015 0040 0015 0040 0015 0040 0015 0015 0015 0015 0015 0015 0015 0015 0015 0040 0015 0015 0015 0040 0015 0040 0015 0040 0015 0040 0015 0015 0015 0040 0015 0015 0015 0040 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0040 0015 0040 0015 0040 0015 0040 0015 0040 0015 0040 0015 0040 0015 0689 0157 0056 0015 0E94

Function 2

0000 006D 0022 0002 0157 00AC 0015 0040 0015 0040 0015 0040 0015 0015 0015 0015 0015 0015 0015 0015 0015 0040 0015 0015 0015 0040 0015 0040 0015 0040 0015 0040 0015 0015 0015 0040 0015 0015 0015 0015 0015 0040 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0015 0040 0015 0015 0015 0040 0015 0040 0015 0040 0015 0040 0015 0040 0015 0040 0015 0689 0157 0056 0015 0E94

I the copied and pasted the digit functions that I needed and it worked. I since have replaced all the other codes the same way and now also these works perfect (before doing this I sometimes had to press twice). As to my testing approach I think that I didn’t change anything since all my buttons are linked to an IR-page and the only things that changed was the IR-input. The thing I really hopped for (my last chance) was that the “long” signals would take longer to transmit and that that would make it work but as it turned out it converted back. The good thing, and the only thing that really matters, is that it worked but it’s beyond me why.