DecodeIR error with TV/0135

Post by **The Robman** » Mon Oct 06, 2003 5:30 pm

I just tried learning buttons from TV/0135, which is for old RCA TVs, and it causes DecodeIR to blow up. This code uses protocol $0056. I don't have any devices that use this code, so I have no idea if the learned signals would actually work.

johnsfine · Post by **johnsfine** » Mon Oct 06, 2003 6:12 pm

I would have expected YOU to give much better details than "blow up".

An eeprom image of those signals is probably needed. I can try TV/0135 in one remote learning to another, but I'd be surprised if I duplicated your result.

I assume you can tell what protocol 0056 is supposed to generate, and I assume you can tell if the learned signals match that, so I'm surprised that you wonder whether the learned signals would work.

Post by **The Robman** » Mon Oct 06, 2003 6:55 pm

I know, I was in a rush, dinner time. Actually, Idon't know what TV/0135 signals are supposed to look like, I was learning them in order to see what they look like. I only glanced at the protocol code, but it looks goofy. There's no data block. So, I would expect that you would be easily able to reproduce the same results.

i have loaded a file in the Diagnosis Area for you to look at.

But at any rate, when you first download using IR, you get an IR message like this...

Then if you click on one of the learned signals, this happens...

Does that help?

johnsfine · Post by **johnsfine** » Tue Oct 07, 2003 7:47 am

I looked at your file in the Diagnosis area.

I think the Access violation in the DLL is more a bug in the call to the DLL from IR.EXE than a bug in the DLL. The caller passes an invalid address. In general subroutines are expected to trust that sort of input from the caller. However, in this case IR also passes zero as the total length of the signal to be decoded.

I will change the DLL, so that when the length is passed in as zero, the decoder just returns a failure result (no decodes found) before touching any of the other data passed in. That will eliminate the access violation.

It would be good if one of the experts could examine the hex sequence of that learned signal in the raw eeprom image and determine whether it represents a form of learned signal that IR.EXE just doesn't understand or whether it represents the results of a bug in the RCU810's learning logic. I don't think I'll have the time to do that examination.

Can your Pronto learn this signal? If so then the examination of that learned signal in the RCU810 would be greatly simplified by learning both the orriginal and the RCU810 learned copy into a Pronto and comparing the Pronto Hex.

I will try to find some time to figure out what TV/0135 is supposed to send.

johnsfine · Post by **johnsfine** » Tue Oct 07, 2003 8:54 am

I'm FAR from sure, but my wild guess is that TV/0135 sends a single long burst of IR and the OBC encodes the frequency of that one burst rather than the structure of the signal.

Post by **The Robman** » Tue Oct 07, 2003 9:06 am

johnsfine wrote:Can your Pronto learn this signal?

Nope. That was the first thing I tried. My guess is that this one uses a strange (ie, too high or too low, or none at all) frequency.

johnsfine wrote:I will try to find some time to figure out what TV/0135 is supposed to send.

If I get a chance, I will look also, but don't make this a high priority, there's nothing hanging on knowing the answer, I was just putting it out there as something that wasn't working properly.

jamesgammel · Post by **jamesgammel** » Tue Oct 07, 2003 9:26 am

Just for the heck of it:

Mill4 has TV_0135 native. I learned a few signals from it to my 6800 and used IR to decode.

Results:
RCA protocol, Device number 15, frequency was 57,553.

Power efc is 099, Channel+ is 043, and ch- is 051.

Hope this helps.

Jim

johnsfine · Post by **johnsfine** » Tue Oct 07, 2003 9:40 am

I may try my mill4 (maybe both) later, but I find that hard to believe.

Are you sure the switch to TV/0135 really took and are you sure the 6800 didn't have those signals left over from a previous use?

You are reporting the standard RCA TV signals, matching TV/0047.

Post by **The Robman** » Tue Oct 07, 2003 11:03 am

I too would be surprised if that's really the case. The Mil-4 does indeed have TV/0135 installed and it does use the strange un-documented $0056 protocol, not the standard RCA protocol.

jamesgammel · Post by **jamesgammel** » Tue Oct 07, 2003 11:52 am

OK, I redid everything. Apparantly IR didn't set the TV button to 0135 as I had set it. It indeed was 0047, confirmed by both Ir and blink-back. I reset it AGAIN to TV_0135, and blink-back confirmed that's what it's set to.

I 981'd my 6800 and relearned the same signals. Sheesh, talk about messing a remote up!!!!!. All kinda weird messages, including the wrong rdf, overfilled learned area (3 signals), and about 5 other buttons that weren't learned to. Changed strategy.

Purged 6800, 981'd again for "virgin" 6800. Learned 5 buttons from Mill4, TV_0135. The remote said success pretty quickly for each button learned. Now download 6800 again.

WEIRD messages again, wrong learned area, rdf wrong, etc. Only one of the 5 learned buttons showed up, TV power. However, nothing there except a frequency of 242. Whatever TV_0135 sends sure screws up a 6800 learning process, evidently just as bad as it does a 810. I guess it's up to you brains now, I don't see where I can be of much help, at least not with what I have. Maybe Jon can use Tommy's analyzer (or do one of you guys have one?), and shed some insight? Sorry I can't be of any help, this really sucks and is weird.

Jim

Post by **The Robman** » Tue Oct 07, 2003 12:46 pm

I have just made a quick attempt at disassembling the protocol code behind this signal. It appears that the carrier frequency varies based on the EFC, take a look.

jon_armstrong · Post by **jon_armstrong** » Tue Dec 16, 2003 5:54 pm

I recently stumbled across the same thing trying to find an existing setup code/protocol for an old RCA TV. I guess I forgot this exchange. I actually just saw Rob's commented disassembly of the protocol in the diagnosis area today. Rob and/or John, did you ever figure this out?

Post by **The Robman** » Tue Dec 16, 2003 8:42 pm

I never looked into this any further myself.