JP1 Remotes

Hi

I bought the URC8811 from Rob (thanks for the speedy delivery!) and have now the following problem:

I programmed it for the Sky digibox dowloading the codes from yahoo. Protocol was RC6-20-0. But all signals I sent to the digibox were sent twice, e.g. I pressed Power, it went on and off; I pressed channel up, it jumped two channels up, etc.....

So I decided to learn all the keys from the original remote but discovered that some keys use the Sony12 device 1, Sony 12 device 3 and RC6-20-0 protocol. I believe this combination is not supported in RM. Just programming the remote with Sony12 device 1 and 3 did not work at all - not a single button.

Can somebody please bring some light in this? I am a bit stuck here....

Many thanks
Amazing

Do the learned signals work properly (ie, they're only sent once)? If so, please post a copy of the IR.exe file that contains the learned signals so one of us can look at it. If this file also contains the upgrade that you loaded, even better.

The Robman wrote:Do the learned signals work properly (ie, they're only sent once)?

Not sure what you mean. I did the learning as per manual, i.e. press as long as the 8811 blinks twice (and somewhere here on the board I read that's better than just pressing once). But I didn't get the learned signals to work anyway as they had the mix of Sony and RC6-20 codes in them. The ones which are send twice are the ones I dowloaded from the yahoo group - they were for another remote but all RC6-20.

I can't send you the learned signals immediately as I've already deleted them. I'll try to learn them again.

Amazing

I learned again a few buttons now but interestingly, this time they are all RC6-6-20. Maybe I had the remotes too close together yesterday....

Here is now the IR.exe Summary. I'll try programming these codes and let you know......

Amazing


24 DVD 1 RC6-6-20 0 0 1 T=0
25 DVD 2 RC6-6-20 0 0 2 T=0
26 DVD 3 RC6-6-20 0 0 3 T=0
27 DVD 4 RC6-6-20 0 0 4 T=0
28 DVD 5 RC6-6-20 0 0 5 T=0
29 DVD 6 RC6-6-20 0 0 6 T=0
30 DVD Guide RC6-6-20 0 0 204 T=0
31 DVD Down RC6-6-20 0 0 89 T=0
32 DVD Right RC6-6-20 0 0 91 T=0

By "do they work", what i meant was, if you point the URC-8811 at your STB and press the buttons that you just learned, do they perform the function that they are supposed to do, and do they do it just once (ie, they don't keep doing it twice like the ones from the upgrade).

As for the IR.exe information, I was actually asking you to download the memory and save it as a file, then post the file in the Diagnosis Area where one of us could look at it. We would be looking at the raw signal itself, not just the summary info that you posted. We might even load it into a similar remote and test it out ourselves.

However, I think I know what's going on. All RC-5 and RC-6 signals have a toggle bit built into them, the purpose of which is to control situations just like this. When you press a given button twice in a row, the toggle bit will change with each press, this confirms to the device that you did indeed press the button twice rather than holding it down. It's because of this very feature that learned signals using these protocols generally only work once, requiring you to press a different button inbetween. For example, pressing 2-3-2 (using learned signals) would work, whereas pressing 2-2-2 would not (only the first 2 would register).

I think there may be a bug in the protocol code where it's toggling the wrong bit. To confirm this, could you learn the same button multiple times (to different buttons). For example, learn the '2' button to the 1, 2 and 3 buttons on the URC-8811, then post the IR.exe file.

I think John Fine is the most knowledgeable person here on the RC-5/6 protocols, so hopefully he can throw his 2 cents in here too.

Hi Rob

Sorry, I did misread your sentence actually. I think I figured out the changing protocols while learning: The sky remote was probably in TV mode and later by going thru the buttons changed it to the Sky mode.

But that still doesn't explain the duplicate signals. In the codes learned today (without changing protocols), the OBC matches the one I had downloaded but the learned codes don't specify the EFC and HEX( they are all empty ??) - see above the summary.

As for the toggle bit, here is the description in IR509 (probably makes more sense to you than to me.)

"This protocol supports any RC6-M-20 variant where the toggle bit doesn't actually toggle (in decodes the T value is always the same).

The only examples we have seen are for the "Sky" device.
In those examples M is always 6, Device is always 0, Sub Device is either 0 or 12, and T is always 0."

I did some more tests now with the new learned signals and the toggle bit doesn't seem to be an issue (even though one frequency always changes - see uploaded file). It just appears that I am not releasing the button quick enough - if I press the button ULTRA(!)-fast then the signal is being sent only once (I got to really concentrate on the speed though - so it's not really practical). It's also not the remote at fault because I can change channels on my tv quite normally.

Don't know what the solution is - delaying the signal somehow? And many thanks for your help so far, Rob!

Amazing

Did you post a file? If so, please post a link to it (and PLEASE read the sticky post that explains how to post links).

I would like to see the results of the experiment that I asked you to perform for me to see for myself whether the toggle bit is an issue.

I would also like you to confirm whether the learned signals perform properly.

The file was uploaded, sorry - here is the link:

http://groups.yahoo.com/group/jp1/files ... opfield.IR

I did what you said: I learned the buttons with a single keypress and tested it straight away - the keypress worked fine, no matter how longed I pressed a button the right signal was sent only once. (The learned keys are in the IR file as DVD buttons). I also put the number 4 key on buttons 4-5-6 and the channel down key on buttons 7-8-9. They all worked.

I then used the learned signals to create the keymap (1847 in IR file). Signals are sent twice again. In fact it's not only twice, it depends on the duration of the keypress, the longer I press a button, the more often the signal will be sent - repeatedly (this is actually where the fault is - the orignal remote sends the signal once).

The downloaded keymap which showed the same problem is put as 0847 in the IR file.

Amazing

Amazing, your link is no good. Please read Posting links to Yahoo files.

Hi

I did some more tests: I used the upgraded 8811 on my two other sky digiboxes (Panasonic and Grundig) and it showed the same problem even though with the Grundig everything was a bit slower. I then upgraded my other remote 8910 with the same keys and that remote also had the same problems.

Guys, my jp1 remote is almost finished and it really rocks but I just need to get this Sky sorted - then it would be perfect. Please have a look at my uploaded IR file and tell me what to do. Is then no one else out there who had the same problem? I tested it now with 3 sky digiboxes and two jp1 remotes.

Many thanks
Amazing

The Robman wrote: However, I think I know what's going on. All RC-5 and RC-6 signals have a toggle bit built into them, the purpose of which is to control situations just like this. When you press a given button twice in a row, the toggle bit will change with each press, this confirms to the device that you did indeed press the button twice rather than holding it down.

Despite what might be read into the statement above, Rob knows that some OEM remotes using versions of RC6 do not toggle any toggle bit. Some receivers of RC6 don't care about the toggle bit. Some even require that the toggle bit NOT toggle.

Clearly the original remote in this case doesn not toggle the toggle bit.

The chosen upgrade was RC6-M-20n, where the "n" in the protocol name was my way of say that it doesn't support a toggle bit.

So the toggle bit seems like it is not part of the problem

The Robman wrote: I think John Fine is the most knowledgeable person here on the RC-5/6 protocols, so hopefully he can throw his 2 cents in here too.

Sorry I failed to notice this question earlier.

Amazing has provided very convincing evidence that the original remote sends exactly one frame of the signal regardless of how long you press the button and that the receiving device relies on that behavior.

That contradicts what I thought I knew about Sky devices from a variety of other sources.

Our RC6-M-20n protocol sends frames as long as you keep the button pressed. That would be a trivial feature to change, though I don't have the documentation handy right now which says exactly what to change.

Do any other experts and/or Sky users have a guess about whether Amazing's devices are different than other RC6-6-20 devices, or whether we were just wrong about this protocol normally having the repeat while held feature?

johnsfine wrote:Amazing has provided very convincing evidence that the original remote sends exactly one frame of the signal regardless of how long you press the button and that the receiving device relies on that behavior.

That depends actually on the button. On my original remote, all Channel Up/Down, number and misc buttons (backup, red, green, etc) are sent once no matter how long you press it. The only buttons which send the signal repeatedly are the Left/Right and Up/Down Arrow buttons (which makes sense otherwise browsing through the EPG would be a pain).

I am surprised that this issue did not come up before as on yahoo there are "working" Sky files (one of which I also tested). Indeed, a second user comment would be useful to conclude on this.

Amazing

Just one more thing regarding the arrow keys. On the original remote, it appears that with a long keypress there is a delay between the first frame and the second frame of the signal, e.g. in the EPG (Electronic Program Guide) if I press the Down Arrow the next item is selected, it stops, and then browses quickly through all the other items. On the upgraded remote with RC6-6-20, as soon as I press the button, the items are browsed quickly through - it does not stop for a moment at the first item.

Amazing

Any news on this issue?

Amazing