JP1 Remotes

gruffnutz666 · Joined: 26 Mar 2004 Posts: 4

Can someone have a look at the learned signals in file '9910 Pace DC420NF download.txt' in the Diagnosis area of the JP1 group. I may be an idiot (no opinions please, I hardly know you...) but I can't figure the Nokia protocol out. Is this device using Nokia Quad or what? How do I replicate these three command codes?

Cheers! Twisted Evil

[Edit by Rob:]
PLEASE include links to file for diagnosis!!!
Here's the link to the file: 9910 Pace DC420NF download.txt

jon_armstrong · Posted: Sat Mar 27, 2004 12:45 am Post subject:

I haven't really analyzed it in detail, but it looks like Nokia (aka Nokia Quad) with four more burst pairs for an additional 8 data bits.

My guess is that that structure confused the decoder and it returned multiple decodes.
_________________
-Jon

jon_armstrong · Posted: Sat Mar 27, 2004 8:57 am Post subject:

Here is my take on the protocol:

{36k,msb}<164,-276|164,-445|164,-614|164,-783>(412,276,D:8,S:8,A:8,F:8,164,^-100m)+ As long as key is held

{36k,msb}<164,-276|164,-445|164,-614|164,-783>(412,276,D:8,S:8,(A+128):8,F:8,164,^-100m) when released

Where there is no label below that was in the "Sent Once On Completion of Other bursts" and reminds me of keyboards where the key when released is 128 decimal higher. So my theory is that A is 0x26 when held and 0xA6 when released.

21 A0 26 54 menu
21 A0 26 0F info
21 A0 26 CC guide
21 A0 A6 CC
21 A0 26 83 stop
21 A0 26 81 enter
21 A0 26 8E prev
21 A0 A6 8E
21 A0 26 F4 play
21 A0 A6 F4
21 A0 26 FD fwd
21 A0 A6 FD
21 A0 26 D5 rew
21 A0 A6 D5
21 A0 26 20 ch+
21 A0 26 21 ch-
21 A0 26 5C sel
21 A0 26 5B right

I would guess that we can create the repeating segment in PB by defining One to be 166,-270 and Zero to be 0,-170

And the conversion in bits from the Quad would be:

00 => 1
01 =>10
10 =>100
11 =>1000

I don't have time until the afternoon to create the protocol in PB. Is it possible there is already a UEIC protocol for this?
_________________
-Jon

The Robman · Posted: Sat Mar 27, 2004 4:36 pm Post subject:

jon_armstrong · Posted: Sat Mar 27, 2004 5:51 pm Post subject:

I just created it with my kludge method and in my limited testing it works. Rob, you are correct I just pad all the short segments with a 1 plus enough zeros to always give me 16 variable bits. Then set the leadout to 100200 and set it to "Off as a Total" and that gets you back a uniform frame.

I looked up the gear and it appears to be a new HD STB with a lot of features, so I suspect UEIC may develop an upgrade.

gruffnutz666, its time to test this for me. There is a file in the JP1 Files|Diagnosis area|Pace_DC420NF.zip that has three files. 9910 Pace DC420NF -- REV1 download.txt is your image of the 9910 but I deleted the learned commands and installed the device and protocol upgrades. All you need to do is upload to your 9910 and test. Everything should be back on the same keys. There is also a KM Master upgrade and the PB file that are both just FYI, for now.

If this works, you need to learn every command and post it so I can translate the commands to work with this Protocol. If you are up to testing all 256 possible commands to find hidden features, discrete commands, etc. then I can generate those as well, but as always you do that at your own risk.
_________________
-Jon

gruffnutz666 · Joined: 26 Mar 2004 Posts: 4

I will test the file ASAP. This may take me a few days (New daughter! Lots to do!) Absorbing the information kindly provided by you all in the meantime. Will post reply soon!

P.S. I think there is a UEIC protocol as I seem to remember reading of similar Pace boxes (Same chassis / series, box is also known as a Pace Mongoose) being supported by European OFA remotes (not much good to me though, US remote and living in Australia, I can't talk local support into finding an appropriate upgrade). Will check.

You may remember (messages still in the archives) many Aussies having the same problem with codes for older (analogue) Foxtel boxes (Pace 100 / 120 / 150, Amstrad STU200) which were only partially and sporadically supported by OFA in Australia.

P.P.S. Love your work...

gruffnutz666 · Joined: 26 Mar 2004 Posts: 4

The protocol is spot on so far.

I've captured the remaining keys and they're in a file below the original in the diagnosis area.

I have a basic understanding now of the info above but I'm not knowlegable enough about the process of building the new protocol to understand the hex translations or 'shifted bits'. I'll keep reading...

If you have time to redo the KM file with the extra keys I'll clean it up with the appropriate function names and publish.

The PB file aided my understanding, thanks.

Cheers!

P.S. I post all device codes I develop and aim to help more as my knowlege grows.

jon_armstrong · Posted: Tue Mar 30, 2004 4:55 pm Post subject:

I just added the additional learned commands and a spreadsheet that will give all possible commands to test for hidden/discrete commands. It also has the decoding method for the learned commands. I put the known commands along side. The decimal commands are the "true" OBC's the double hex bytes are the values to put in the OBC column in the device upgrade.
_________________
-Jon

gruffnutz666 · Joined: 26 Mar 2004 Posts: 4

Thanks John, and others! I've just had the Foxtel box officially installed (I've been working with a pre-rollout box borrowed from an installer friend) and fully tested the first batch of codes with complete success. The original remote is incredibly frustrating to use compared to the programmed 9910. Even with the RF/IR repeater disconnected the 9910 can be used from almost anywhere in the room whereas the original remote requires me to fire it from around 1ft above the floor and directly in front!

I'm sure a lot more people will want to replace the original remote.

I'm not convinced it's just a matter of IR strength either. The original remote seems to take forever to send each key press (as if it was actually executing three seperate Nokia commands for each key press as DecodeIR indicates). Weird!

I'll look into it further.

Will label and clean up the KM file and test all OBC's / EFC's from the spreadsheet for hidden features. (I'll proceed with caution having previously cocked, and fixed, up my Philips DVD 951 playing with codes.

Results posted when I'm done.

Cheers again!

Haggis · Joined: 14 Sep 2003 Posts: 27 Location: Brisbane

I have been using the Sky Digibox KM device code (from the Yahoo group JP1 files section under satellite) on the Pace420 (cable) (with only sight key re-arranging) and it works without a problem. I was wondering why the OBC codes seem to differ between this solution and the OBC's?

jon_armstrong · Posted: Sun Apr 04, 2004 7:40 pm Post subject:

I didn't realize we had decoded this protocol before. I think the Sky Digibox upgrades may use the official UEIC protocol. In this recent case I just kludged together a protocol using PB and it's pretty ugly. I'll test it to make sure tomorrow and kill the kludge version if so.

Thanks for bringing this up.

BTW, did you ever figure out your air conditioner?
_________________
-Jon

Haggis · Joined: 14 Sep 2003 Posts: 27 Location: Brisbane

I decided to test the two remotes (the original that came with the Pace420NF:- URC1633B00) and my 7562 (with Sky digibox device codes) against my GetIR.xls spreadsheet that uses DecodeIr.dll. Now both of these control the Pace 420 without a problem, yet the IR decodes are very different. The URC 7652 gives me a consistant decode of RC6-6-20 while I have trouble getting a consistant reading from the original (I have had various gap decodes as well as a Nokia decode). I have read that according to the manufacturer, the IR signal is supposed to be a RC5/6 type and so I am a bit confused to the cause of inconsistant readings.

BTW. I was eventually able to fully decode the air conditioner signals, but other than the off button, all the rest sent a complete status of the air conditioner settings to the unit. I.e., the temperature setting, fan speed etc and hence setting up the combinations on a URC would not be feasible even if we could create a 15 byte protocol.

jon_armstrong · Posted: Mon Apr 05, 2004 6:20 pm Post subject:

I took a look and the upgrade Haggis is using and it is definitely RC6-6-20. The IR Protocol that gruffnutz captured is a modification of what we call Nokia Quad with 32 data bits instad of 24. The coding of a "quad" is using four different burst pairs that each signify two bits.

These are very different IR Protocols. Since the model numbers are close if not identical, it is possible that it responds to two command sets. It is also possible that the same hardware is used by two different cable providers and they specified different IR command sets for system compatibility.

If one of you gets curious you might try the other upgrade to see if the gear responds.
_________________
-Jon

The Robman · Posted: Mon Apr 05, 2004 7:50 pm Post subject:

Just FYI Jon, I'm pretty sure that I've seen a 4-burst protocol that is either a match for this, or is very similar, but I can't remember which one it is. I do remember thinking that the 4-burst situation was handled very nicely though. If I find it, I'll post some details.
_________________
Rob
www.hifi-remote.com
Please don't PM me with remote questions, post them in the forums so all the experts can help!

Haggis · Joined: 14 Sep 2003 Posts: 27 Location: Brisbane

Hi Jon, I tried out your protocol / device upgrade, and it works on the Pace box. Given that both protocols work, I agree that the STB must currently respond to both command sets. Given that the cable company can upgrade the STB firmware at any stage and that your upgrade corresponds to the remote that came with the box, I believe that your upgrade is the correct way foward.
Regards, Haggis