XBOX - do I have 56khz problems?

[Mog]

As before, I am trying to put together an upgrade for my URC 8550 Topline8, and this time it's the XBOX

I have searched the forums for previous discussions, and it is clear that there are a number of variants (MS XBox, Noname XBOX etc)

I have experimented with the various configurations, without any success.

Most of the codes can be viewed here as I have learned them with my Kameleon
These work fine, and the Kameleon can control the XBOX - I have a few questions as a result...

Two frequencies are reported, 57142 and 56737 - does this mean that to correctly handle the XBOX, I would need two protocols,
and maybe combine them?

Should I expect the Topline8 to be able to control a device at these frequencies?

I seems that my attempts to upload the Topline8 produce IR data that the Kameleon can't learn at all.

And yes, I have tried DVD 522. I have to change to the P8/740 chip, and KM seems happy to make up the $AF protocol for the remote,
but the XBOX doesn't respond, and the Kameleon doesn't want to learn the output.

(Yes, I have new batteries, and the Kameleon will learn from other remotes without problems)

Any suggestions guys?

[johnsfine]

Two frequencies are reported, 57142 and 56737 - does this mean that to correctly handle the XBOX, I would need two protocols

No. The learning remote only approximately measures the frequency, so it could easily report those two different frequencies for two learns that are actually the same frequency.

Probably the protocol executor for that has the right frequency more exactly than you could tell from learned signals. But even if it didn't the actual device won't care if the frequency is wrong by a moderate amount.

[The Robman]

You obviously have the official Microsoft remote kit, because those signals use the RCA protocol and are an exact match to the DVD/0522 code. Unfortunately, the URC-8550 doesn't have this code installed. Furthermore, it doesn't have either version of the RCA protocol installed either, so you will need to use a protocol upgrade.

Try this upgrade, and be sure to set the "Device Button Restriction" to a specific device button (rather than "All Devices") due to the firmware bug present in this remote ...

Upgrade Code 0 = 92 0A (DVD/0522) XBOX (KM v8.24)
AF 00 B0 FA 79 5F 30 31 32 33 34 35 36 37 38 39
20 22 2A 2A 1D 1C 19 1F 08 59 58 56 57
End

Upgrade Protocol 0 = 00 AF (P8/740) RCA (KM v8.24)
05 0E 11 80 0E F4 72 1D 04 08 03 85 02 38 08 97
E4 06 2B 4C 00 FF
End

[Mog]

Rob, that was one of the combinations I have tried

With that upgrade the results are rather odd - the XBOX does respond to REW and FF, which seems to work as Up and Down
The Up and Down keys seem to operate as repeating versions, so a single press of Up moves the selection all the way to the very top,
and a single press of Down moves the selection all the way to the very bottom

None of the keys can be learned by the Kameleon...

[The Robman]

It probably won't make any difference, but try these upgrades instead...

Upgrade Code 0 = 92 0A (DVD/0522) XBOX (KM v8.24)
2D 00 B0 FA 79 5F 30 31 32 33 34 35 36 37 38 39
20 22 2A 2A 1D 1C 19 1F 08 59 58 56 57
End

Upgrade Protocol 0 = 00 2D (P8/740) RCA (Old) (KM v8.24)
05 0E 11 80 0F F4 18 1D 04 08 03 87 02 3A 09 01
77 06 27 05 22 00 A9 39 20 DB 00 A2 F6 A0 08 22
44 7F 7F 3C E4 73 22 06 90 EA 60
End

Also, go to the Service Menu and make sure that AP (Anti Piracy) is turned off. (See the URC-9800 manual on my site for directions to the Service menu)

[Mog]

No, REW and FF still work as before. The Up/Down keys no longer seem to have any effect

Not sure about the AP, I have checked it and set it to disabled.
Should I then reload the upgrade?

I have tried further to record these signals, and did persuade the Kameleon to record something. It doesn't seem useful though, with lots of red boxes and warnings. I only managed to get anything by holding the Topline TX about 1cm from the Kameleon, and holding the keys down for long periods

And I thought the XBOX would be easy...

[Mog]

I have finally hacked my WinLIRC stuff to do a decode of the 56Khz XBOX Remote Control, and recorded a few keys here

It seems that the decode shows the RCA nature of the command, and has the dev-cmd-!dev-!cmd format,
with 4 bit device and 8 bit command, used in, for example, protocol $AF

However, maybe the reason I am having problems, is due to the lead-in sequence, which is 4035-1264-101-2541

As far as I can see, the 101-2541 pulse is not supported by the protocols I have used thus far.

This leads me to the conclusion that I will have to hand-code my own burst sequence to achieve this?

( I see there is an option for the Lead-in of 'half-size after first', which is close but no cigar )

By the way, I don't think my remote is Microsoft, it looks just like the picture posted in this thread

[johnsfine]

However, maybe the reason I am having problems, is due to the lead-in sequence, which is 4035-1264-101-2541

I'm certain that is not the lead-in sequence your xbox remote is supposed to send. I don't know enough about LIRC to be sure of the cause. It could be an interaction between bad batteries in the xbox remote and some weakness in the IR receiver used by LIRC. Or maybe it's a software error in LIRC, or maybe an OS latency glitch.

Whatever it is, you should ignore it and trust that Rob knows how to generate the right xbox protocol.

[Mog]

I'm certain that is not the lead-in sequence your xbox remote is supposed to send. I don't know enough about LIRC to be sure of the cause.

Well I am familiar with LIRC, and the checks for signal repeatability are comprehensive.
Note that everything else in my decoding is exactly as expected

It could be an interaction between bad batteries in the xbox remote and some weakness in the IR receiver used by LIRC. Or maybe it's a software error in LIRC, or maybe an OS latency glitch.

The batteries in the XBOX remote are quite good enough to provide full control of my XBOX...

"Whatever it is, you should ignore it and trust that Rob knows how to generate the right xbox protocol.

Well, it's difficult to ignore when it is clearly there for each key recording.

And, unfortunately, the xbox protocols I have tried so far don't work for me

[Mark Pierson]

You did heed Rob's advice from earlier, correct?

be sure to set the "Device Button Restriction" to a specific device button (rather than "All Devices") due to the firmware bug present in this remote

[The Robman]

If he hadn't followed that advice, every button would send the command assigned to the ZERO button, which is clearly not happening here, so I suspect that it's something else.

If I get any free time over this weekend, I will try and load these upgrades into one of my 740 remotes and see if I can get it to send the right signal.

[johnsfine]

Well I am familiar with LIRC, and the checks for signal repeatability are comprehensive.

So whatever is going wrong is doing so consistently.

Note that everything else in my decoding is exactly as expected

That observation cuts both ways. It gives us confidence that the signals your xbox remote should generate are the same RCA signals we expect, so it gives us confidence that the extra glitch in the lead in shouldn't be there.

The batteries in the XBOX remote are quite good enough to provide full control of my XBOX...

1) IR receivers in ordinary devices are typically much less sensitive to problems in the batteries of the sending remote than are general purpose IR receivers in learning remotes and PCs. The xbox might be receiving exactly the right signal and the PC receiving the wrong signal, even with the sending identical.

2) The end of a long lead-in is the least reliable part of an IR signal, and is pretty much useless for information content. The xbox might well be programmed to ignore an error in that part of the signal even if it receives exactly what the PC received (and trusting that I'm right that what the PC received was an error).

And, unfortunately, the xbox protocols I have tried so far don't work for me

URC-8550 details are way outside my expertise (more so than LIRC is), so all I can say on that one is make sure you really did what Rob instructed.

The structure of IR signals and the behavior of IR learning systems when batteries are bad are topics very much in my expertise. So I remain certain that the lead in you posted is not a correct lead in for your xbox remote and is a plausible symptom of bad batteries in the sending remote.

[Mog]

You did heed Rob's advice from earlier, correct?

Indeed, I did.

The structure of IR signals and the behavior of IR learning systems when batteries are bad are topics very much in my expertise. So I remain certain that the lead in you posted is not a correct lead in for your xbox remote and is a plausible symptom of bad batteries in the sending remote.

Fair enough. One thing I am certain of is that you guys have way more knowledge about all these matters than I do.

However, equipped with new batteries, I still consistently get the extra pulse

[jon_armstrong]

Let me summarize:

1. The learned Kameleon commands work.
2. Any generated commands in your 8550 result in only a few commands that work at best.

Right now there are two possibilities:

a. The problem is caused by the xBox needing to see an extra lead-in pulse that isn't being generated by the protocols.
b. Something else.

It seems like it can't be a because of 1. There are no extra lead in pulses in the learned commands in the Kameleon. The Kameleon just transmits the pulses back that it has learned.

So it may well be that the OEM remote does transmit an extra pulse, but because of 1 it doesn't matter.

If you accept that it's b, then I would start looking at things like total frame length (the sum of the absolute value of all the pulses including lead in and lead out). You could use the LIRC IR receiver to compare the two signals. One of the reasons, I am suggesting that is the behavior of the down arrow that moves very fast. The second reason is that I can't come up with a better idea

BTW, does DVD/0522 work in the Kameleon?

[johnsfine]

Earlier I missed the detail that you (Mog) also have a Kameleon and it learns the xbox remote signals without the lead in glitch.

I assume Rob's intention, when he has time, is to learn the signals from the upgrade he gave you (using some remote compatible with an 8550) into some other JP1 remote such as the Kameleon. After downloading those to IR.EXE, and comparing to the real xbox signals, it should be easy to diagnose what is going wrong in your 8550, and hopefully create a better upgrade.

If you do the 8550 to Kameleon learning step before Rob gets to it, that might speed up the process, and allow the upgrade to be fixed by anyone with any 8550 expertise, rather than just Rob.