problems getting new ir code to work

majik · Post by **majik** » Sat Aug 25, 2007 8:04 am

This is a repost from AVSForum. I'm hoping somebody here (who has not seen the original post at avsf) can help me out:

I am trying to load IR codes for the Motorola VIP1200 (ATT U-verse). I created the .rid file and uploaded the file to my 5040 RTV (network and date/time meter appear....woohoo! Thanks FlipFlop for your bootable CD). However, the IR code doesn't seem to work.

One thing that caught my attention was that in the step where RIDgen is supposed to fill out the "Initial bits", it didn't. It only filled out the "Repeated bits." See below:

I got the device file from Robman's post (Robman, you noted that it's untested...maybe that's why it isn't working for me? How can I create this file myself? I have JP1 serial cable but unfortunately the Motorola VIP1200 remote doesn't seem to have JP1 connection...at least not easily accessible).

What am I doing wrong that is causing the "Initial bits" not to generate? Or is it normal that there aren't any "Initial bits"? Here is what I see in RM:

johnsfine · Post by **johnsfine** » Sat Aug 25, 2007 8:08 am

Nokia32 has no "initial bits", so that isn't the problem.

I'm not sure enough about RIDgen to know what the problem is.

majik · Post by **majik** » Sat Aug 25, 2007 8:23 am

Thanks for clearing that up. I had downloaded random device files and loaded them on RIDgen and all happened to have initial bits.

perhaps the problem is with the device file The Robman provided? How is this file created? I could not find any info on this.

PS: johnsfine, I'm guessing you got my e-mall about registration problems....and if you didn't, then I guess the email just took long to be delivered. I checked for it last night and it wasn't there (checked spam folder too). I checked again this morning and finally saw it in the spam folder.

johnsfine · Post by **johnsfine** » Sat Aug 25, 2007 8:40 am

I'm trying to get RIDgen to work with MakeHex for Nokia32, but I haven't figured it out. The OBC/EFC translation in RIDgen is very confusing and probably wrong.

In Rob's file, a value named "X" is set to 38. RIDgen has no way to understand that value, so you would need to edit the Nokia32 file before running RIDgen to set X correctly. Even after doing that, I haven't figured out how to make RIDgen work right.

I don't know whether Rob's file is correct. But until you get Ridgen to generate output correct following Rob's file, you shouldn't assume Rob's file is wrong.

I don't know how to help with registration problems and I ignored an email from someone recently about that. Usually I reply and tell them to contact Rob instead.

johnsfine · Post by **johnsfine** » Sat Aug 25, 2007 8:52 am

Some basics about RIDgen that may help anyone else trying to figure this out:

The "Encoding" section defines the set of one-character symbols used in the "Reated bits" section. For Nokia32:

800f000a is the Start symbol
80060c?? is the End symbol (regardless of what ?? is). RIDgen defines multiple end symbols for Nokia32. It only needs one.

8006000a is the 0 digit
80060010 is the 1 digit
80060016 is the 2 digit
8006001c is the 3 digit

Based on Robs file, every function should be:

Start 0 2 1 1 0 0 0 0 0 2 1 2 ? ? ? ? End

(The four ?'s represent the OBC number in base 4).

Comparing that to what RIDgen generated for me, I saw RIDgen is not passing the SubDevice 0 to MakeHex correctly.

So I edited my Nokia32.irp file to add the line
Default S=0
which says that when RIDgen fails to pass the Subdevice, MakeHex should use subdevice 0. Now my whole Nokia32.irp file is:

Device=37.0
Function=0..255
Define X=38
Default S=0
Protocol=Nokia32
Frequency=36000
First Bit=MSB
Zero=164,-276
One=164,-445
TWO=164,-614
THREE=164,-783
Form=;412,-276,D:8,S:8,X:8,F:8,164,^100m

Now RIDgen is generating reasonable looking output, but I still can't be sure that output really matches Rob's file, nor can I be sure Rob's file is correct. But at least I'm no longer sure that RIDgen's output doesn't match Rob's file.

Guest · Post by **Guest** » Sat Aug 25, 2007 9:12 am

I'm fixing the Nokia32.irp file and about to recreate the .rid file. I'll let you know if it works.

johnsfine · Post by **johnsfine** » Sat Aug 25, 2007 9:14 am

I found one more problem. RIDgen runs out of symbols because it duplicates the trivial difference between nearly equal End symbols and then (depending on the sequence of functions in the set) it may discard the important difference between functions.

To fix that, I made one more change in the .irp file. I changed the "^100m" at the end to "-90m". That will make the repeat rate slightly wrong on some functions, but probably not by enough to ever be detectable, and it avoids the much more serious consequences of RIDgen running out of symbols.

majik · Post by **majik** » Sat Aug 25, 2007 9:33 am

didn't work. now i'm trying by making the "-90m" change.

majik · Post by **majik** » Sat Aug 25, 2007 9:50 am

that didn't work either.

Guest · Post by **Guest** » Wed Aug 29, 2007 5:55 am

does anybody have any other suggestions I can try?

majik · Post by **majik** » Fri Sep 14, 2007 6:16 pm

Well I imported the protocol to a JP1 remote and confirmed that the device file works (Robman, you can remove the "untested" notice...I tested it with the VIP1200 and VIP1216). I now suspect that it may be the IR blaster/cable. I'll need to dig out my 2nd one.