Can we make GI protocol in C7(7800) the same as 1067Bx3 S3F8

Post by **ElizabethD** » Sun Sep 06, 2009 6:51 pm

From Decode.html

G.I. Cable and G.I. Cable{1}
UEI protocol: 00C4
IRP notation: {38.7k,490}<1,-4.5|1,-9>(18,-9,F:8,D:4,C:4,1,-84,(18,-4.5,1,-178)*) C= -(D + F:4 + F:4:4)
EFC translation: LSB

When the {1} is shown as part of the protocol name for G.I. Cable, it just means that the repeat part of the signal is not present. That doesn't indicate any difference in the actual protocol nor even any unreliability in the decode. It may indicate that use of the learned signal will be less reliable, so you have more than usual reason to replace it with a KeyMove, Upgrade or cleaned up version.

Widget's IRscope signals display does not show {1}, the power button signal decodes just fine as GI cable, device 0, obc10, hex50 as recorded from Comcast's 1067Bx3 remote and used in a Motorola DCH70 box.
When I use it in C7 (7800) something is a tad flaky about that power signal and sometimes other buttons.
I recorded the 7800 signal for power, pressed twice, and I see a slight difference from what the Comcast remote sends.
I do not understand the significance of the slightly different signals.

The Comcast remote built-in version of the protocol is not accessible to normal humans, so I hope this evidence via the Widget and pics in this zip file might help someone help me make a better protocol for this Moto box
https://www.hifi-remote.com/forums/dload ... le_id=7199

I looked at this protocol in PB but don't know how to tweak
GI cable for 7800 in KM, (8910 will just have 0146 not 0133 on last line), and RM version is identical
Upgrade Protocol 0 = 00 C4 (S3C8) GI Cable (KM v9.17)
43 8B 01 8B 12 CC 4D 00 08 00 FA 08 B6 00 FA 04
51 C3 3C 11 94 08 B6 20 11 08 03 F6 80 36 08 04
F0 C0 56 C0 F0 04 04 C0 60 C0 56 C0 F0 06 C0 10
F6 80 36 8D 01 33 1C 08 C0 C0 10 04 1A FA AF
End

Post by **vickyg2003** » Sun Sep 06, 2009 9:49 pm

Liz, can you make this easy for me, and post the 7800 IR file, so that I can take a look at it in my widget? I'm not sure what you meant by "I recorded the 7800 signal for power, pressed twice. " When using our widgets to interpret a protocol we need to use the same strategy as we do with the when learning for protocol decodes. The press and hold is really important. What we are seeing here is a the comcast repeat frame is different, or the 7800 repeat is missing. I like to dump these into excell and compare them. When you do that you can see right away that there is a frame missing from the 7800 signal (In excel, lines 40,41,42,43 are missing from the 7800 signal), but again, are these press and holds or press-press signals.

Post by **vickyg2003** » Mon Sep 07, 2009 8:33 am

Liz, how did you get the 2 keypresses? Was it a macro? I'm a little more awake and sober this morning, and the macro scenario crossed my mind, seeing how you are the queen of macros! I learned so much about macros from your test files! I see that your keypresses are missing this repeat frame. (18,-4.5,1,-178)* If these were learned from a macro this would mean that the protocol needs to send at least one repeat frame.

Can you point me to the upgrade where this came from so I can make sure this works in all versions? Your IR file would be helpful too, since I'll be using the 7800 to test the s3c8 processor. I still need to perform my own learns to see how this file to see how its supposed to act.

Post by **ElizabethD** » Mon Sep 07, 2009 8:44 am

Vicky, this file is under construction again, but works ok. It's that new CBL device I have to have working well.
https://www.hifi-remote.com/forums/dload ... le_id=7202

I don't think you need it really. The Widget showed it all. Yes, the missing blips you quote is what I think if I put into the 7800 signal, the Moto box might work better.
I recorded 2 presses in each remote version. Not press and hold. And not from a macro.
But, it's the buttons within the macros and special keymoves that are at issue being somewhat flaky.

Edited: Vicky, 7800 is funny also in the jvc protocol. It sends just one signal, whereas the builtins in 8910 and 6131 send 3. Once modified into 0134 with 3+ repeats, jvc proto works great. Which is why I kinda think it might help here if we add those itty-bitty endings to the GI job.

Post by **ElizabethD** » Mon Sep 07, 2009 9:40 am

Vicky,
I just recorded 8910 native CBL/0476. It does have the little repeat section so newer S3C8 won't be at issue here, just 7800.

I also just recorded where CBL/Power is used in LKP and/or MACRO macro on 7800. No blips at the end of those GI signals.
So the longer hold would have no effect in 7800. You can just press MACRO key to see (or not) the missing tiny repeat. But since it's a toadTog, remember, on alternate presses you may get no power signal there, though 1,2 should go through.
Oh, and set the Widget for max duration, 10000. We barely fit it in that time.

Post by **vickyg2003** » Mon Sep 07, 2009 11:06 am

I'm sure you are running into the same issues that you ran into with the JVC protocol. We're not getting any repeat. According to protocol help,

[Repeat] Value
--------------
This is the number of times that you want the remote to repeat
the command.

The proper number of Repeat commands is important for reliable
recognition in some hardware. If a command is used in a macro,
then by default, it will be repeated 3 times, unless it is
specifically set to fewer or no Forced repeats, or is a non-
repeating key where repeats are specified only for certain keys.

The Repeat parameters may require some experimentation. Certain
IR receivers want to see a repeated value. Some only want it for
certain commands. Sometimes, for example, the Volume UP or DOWN
increment may be too big or too small, so these are settings to
tinker with to get the behavior you want.

Type
----
A Repeat Type of "Minimum" is the number of times the signal will
be repeated with each single button press. If you keep depressing
the button then you can get more repeats than the minimum.
Setting the [Repeat] Value less than "3" with "Minimum" does NOT
override the 3 repeats in a macro.

"Forced" repeats are the exact number of repeats that are in
[Repeat] Value regardless of how long you hold the button. When
selected, it will override the default of 3 when used in a macro.
For a manual key press, the "Forced" number of repeats will be
issued regardless of how long the button is held down.

Hold
----
The Repeat Hold setting has four choices:

1. "No" indicates no Repeat is used.

2. "Yes" indicates Repeat when button is held.

3. "Ch+/-, Vol+/-, FF, Rew" determines "Yes" for the specified
buttons, "No" for all others.

4. "No data bits in repeat" generates repeat frames that are very
different from normal frames and do NOT include the data portion.
These are like the NEC1 Protocol.

The current settings are
repeat value:
type:forced
Hold:yes

You'd think changing this to 1,minumum would do something, but instead you get no signal at all. According to KM both 8900 and 7800 would use the same protocol values.

I'm not sure if we can do this with just value settings. We could do this by turrning repeating off, calling the xmitter and then turning repeating on and then jumping to xmitter, but this is a lot more code than just setting bits. Since you have so little space, I'm hoping that a change similar to the JVC will be what you need.

Post by **vickyg2003** » Mon Sep 07, 2009 12:00 pm

While a value of 1 didn't work at all, a value of 3 will get you 3 repeat repeat frames which should make the keys work better in macros.

repeat value:3
type:minimum
Hold:yes

Upgrade protocol 0 = 00 C4 (S3C8) PB v4.01
43 8B 01 8B 14 CC 5D 00 08 00 FA 08 B6 00 FA 04
51 C3 3C 11 94 08 B6 FF 03 20 11 08 03 F6 80 38
08 04 F0 C0 56 C0 F0 04 04 C0 60 C0 56 C0 F0 06
C0 10 F6 80 38 8D 01 33 1C 08 C0 C0 10 04 1A FA
AF
End

This is a 65 byte upgrade.

Post by **ElizabethD** » Mon Sep 07, 2009 12:43 pm

Whoa!
For a person who doesn't know anything about protocols you're not doing too badly

Ok, while you did what you did, I tried 3 repeats, no hold, and NoDataBits in repeat setting and it got me totally nowhere (GAP). It changed 4D to 4F at 8006, that's all. Now I will take your thing and use and study and will report back.

Space is no issue. It can overflow or I can zap multiplexed devices or VCR.

Post by **vickyg2003** » Mon Sep 07, 2009 10:44 pm

Elizabeth if you have the space try this one instead. I could have sworn that last one worked, but when I checked again I wasn't getting a repeat. I'm not sure what is going on with the 7800, so I wrote a version that forces one repeat.

Upgrade protocol 0 = 00 C4 (S3C8) gi (PB v4.01)
43 8B 01 8B 12 CC 4C 00 08 00 FA 08 B6 00 FA 04
51 C3 3C 11 94 08 B6 20 11 08 03 F6 80 3C 08 04
F0 C0 56 C0 F0 04 04 C0 60 C0 56 C0 F0 06 C0 10
F6 80 3C F6 01 33 E6 29 4D 8D 01 33 1C 08 C0 C0
10 04 1A FA AF
End

Post by **ElizabethD** » Tue Sep 08, 2009 2:03 pm

Vicky,
I was having trouble with the site, not the protocol, hence the delayed response.

Upgrade protocol 0 = 00 C4 (S3C8) PB v4.01
43 8B 01 8B 14 CC 5D 00 08 00 FA 08 B6 00 FA 04
51 C3 3C 11 94 08 B6 FF 02 20 11 08 03 F6 80 38
08 04 F0 C0 56 C0 F0 04 04 C0 60 C0 56 C0 F0 06
C0 10 F6 80 38 8D 01 33 1C 08 C0 C0 10 04 1A FA
AF
End

works very, very well, and produces identical signal to what Comcast now issues. I changed your FF 03 to just 02 repeats, since only one of those short things were missing.

Upgrade protocol 0 = 00 C4 (S3C8) gi (PB v4.01)
43 8B 01 8B 12 CC 4C 00 08 00 FA 08 B6 00 FA 04
51 C3 3C 11 94 08 B6 20 11 08 03 F6 80 3C 08 04
F0 C0 56 C0 F0 04 04 C0 60 C0 56 C0 F0 06 C0 10
F6 80 3C F6 01 33 E6 29 4D 8D 01 33 1C 08 C0 C0
10 04 1A FA AF
End

works equally well, and produces identical signal to what Comcast now issues.

I also see that in S3C8(new) - my 8910/9910 and 6131 I will be making the change, because both of those remotes also omit those extra four signals. So I'll just change 0133 to 0146 and it should be ok.

Why that tiny difference made a difference, perhaps Rob can tell.
How you made the changes is beyond me, because I don't understand the bits in pf1.
I made a crossreference of sorts for anyone who might try those versions and it includes the Widget data
https://www.hifi-remote.com/forums/dload ... le_id=7204

Anyway, many, many thanks for your wonderful work.

Edited: misspelt 8910

Post by **vickyg2003** » Tue Sep 08, 2009 2:43 pm

How you made the changes is beyond me, because I don't understand the bits in pf1.

I don't understand the bits in pf1 either, that's what I use PB for.

I looked at your spreadsheet and I can shed light on why that JP1.3 version didn't decode. S3F8 is the same as S3C8+ as far as assembly language and offsets go. In order to get the protocol to decode in PB you need to change the s3f8 to s3c8+.

Post by **ElizabethD** » Tue Sep 08, 2009 6:35 pm

Thanks again, Vicky.
I didn't know i can use s3c8. Perhaps we can use it in KM as well then when configuring 1067Bx3 s3f8. What do you think?

BTW, I mentioned earlier that this might be similar to the jvc issue. I don't think so. JVC needed the 3 repeats for C7(7800). This protocol change needed a protocol change, i.e. what comes out, not how many times. Am I wrong?

Post by **vickyg2003** » Tue Sep 08, 2009 7:08 pm

I didn't know i can use s3c8. Perhaps we can use it in KM as well then when configuring 1067Bx3 s3f8. What do you think?

Not the s3c8 but the s3c8+. Remember that KM'ss job is to format the upgrade. If KM isn't set up for the 1067b you'll need to use RM.

BTW, I mentioned earlier that this might be similar to the jvc issue. I don't think so. JVC needed the 3 repeats for C7(7800). This protocol change needed a protocol change, i.e. what comes out, not how many times. Am I wrong?

No, we did exactlyyou said. We increased the repeats. The repeat frame is ,(18,-4.5,1,-178). What we needed was a repeat frame, just as you needed in the JVC. No matter what the JVC repeat looks like, you needed to repeat it. Go back and read my infrared protocol primer. It took me YEARS to understand what a repeat frame was, and hopefully I've conveyed that with pictures.

Post by **ElizabethD** » Sun Sep 13, 2009 12:46 pm

Thanks again.
Yes, I didn't know s2c8+ will disassemble. And writing s3c8 was my omission.

Regarding repeats - I actually got the drift with Tommy's IRSA a little while back. I think what your wonderful primer does is bring home that when you add repeats, it's not some willy-nilly fix, but the use of the built in protocols and the signal structure based on recording like we do onto a remote - press and hold till the learning is complete.

The bottom line for me is, now that 7800 is taken care of, that I think 6131 will also need repeats, but 8910 does not. 6131 and 7800 when running from a macro (extended or unextended) only send the first part of the GI protocol. Just to annoy you

6131 JVC signal does what it's supposed to. Go figure.

Post by **vickyg2003** » Sun Sep 13, 2009 1:15 pm

ElizabethD wrote:Thanks again.
Yes, I didn't know s2c8+ will disassemble. And writing s3c8 was my omission.

Regarding repeats - I actually got the drift with Tommy's IRSA a little while back. I think what your wonderful primer does is bring home that when you add repeats, it's not some willy-nilly fix, but the use of the built in protocols and the signal structure based on recording like we do onto a remote - press and hold till the learning is complete.

Thanks Elizabeth. I wrote that piece because for the longest time EVERYTHING about the numbers seemed willy-nilly. I'd follow Rob's excellent examples, and just could not get why some numbers were being tossed out.

BTW: I've sent you two emails. Did you get them or are they going into your trash bin?