Xiaomi Mi Box Streaming Android TV Device

binky123 · Post by **binky123** » Mon Apr 30, 2018 8:31 am

I finally am setup to do some testing/capturing of the signals. There are several errors in the code and the HCS08 timing is off from what the S3F80 generates.

I'll work on the timing and then work my through the code and do incremental testing of pieces of the code.

binky123 · Post by **binky123** » Tue May 01, 2018 1:34 pm

Alright. I think I have something that is close based on the timings I've captured.

[Xiaomi]
PID=02 3B
CmdParms=Dev:8, OBC:8
CmdTranslator=Translator(0,8,0) Translator(1,8,8)
DefaultCmd=00 00
Code.S3C80=43 8D 02 8B 18 C0 00 01 2F 01 12 01 2F 01 A3 01 2F 02 36 01 2F 02 C7 02 02 01 12 3B 10 E4 1C 0B E4 1D 0C 68 03 F0 C6 B4 03 C6 B4 04 C6 F0 C6 B4 04 C6 F6 FF 3E F6 FF 3E F6 01 0A 7B F8 AF E4 26 1C E4 27 1D F6 01 76 F6 FF 54 E4 26 1C E4 27 1D F6 01 82 AF E4 0B 1C E4 0C 1D 1C 22 F6 01 4C 38 03 F6 FF 73 38 04 F6 FF 73 38 C6 F6 FF 77 1C 12 8D 01 4C 5C 04 8B 02 5C 02 90 C3 90 C3 08 C3 56 C0 03 86 04 C0 06 C1 12 F6 01 4C 5A EC AF
Code.MAXQ610=33 69 02 96 00 05 17 00 16 00 17 00 21 00 17 00 2C 00 17 00 37 00 73 04 27 00 16 00 00 84 00 00 00 01 0C 07 D2 D0 D1 51 D3 D2 00 07 D2 D3 D2 63 54 03 01 70 71 32
Code.TI2541=01 07 01 21 02 96 00 05 17 00 16 00 17 00 21 00 17 00 2C 00 17 00 37 00 73 04 27 00 16 00 00 84 00 00 00 01 0C 07 04 02 03 51 05 04 00 07 04 05 04 63 54 03 01 70 71 32
Code.HCS08=20 1B 23 47 02 C0 00 01 2F 01 26 01 2F 01 B7 01 2F 02 4A 01 2F 02 DB 02 02 01 26 3B 10 4E 74 68 4E 75 69 B6 60 62 B8 60 B8 61 62 B8 61 B7 53 AD 08 AD 06 CD FF 92 25 F9 81 55 7E CD FF 74 AD 0A 4E 7E 74 4E 7F 75 CD FF 56 81 4E 68 74 4E 69 75 AE 7A CD FF 65 4E 60 55 AD 0F 4E 61 55 AD 0A 4E 53 55 AD 0A AE 6A CC FF 65 6E 04 54 20 03 6E 02 54 B6 55 49 49 B7 55 49 A4 03 AE 04 42 AB 6A 97 CD FF 65 3B 54 EB 81

alfi33 · Post by **alfi33** » Tue May 01, 2018 11:03 pm

I added your red text to my protocols.ini file and now the upgrade works pretty well on my URC-8820n.

All buttons work and do what they're supposed to with the exception of the Menu button. It looks like that one may mapped to the Mic button on the original remote. In most cases, the Menu button doesn't do anything when I press it...but in at least one app, it does the same thing as the Mic button on the original remote. Not sure what's going on here but I would guess that this may be an issue with the device upgrade and not with what you've been working on.

Also, it feels like I get some double-presses unless I'm being careful to press the buttons quickly. However, it's certainly very tolerable for me though!

Thank you very much for the work you did on this!

binky123 · Post by **binky123** » Wed May 02, 2018 12:12 am

Maybe the gap between the frames is too long and resulting in the device seeing it as a new press.

Code.HCS08=20 1B 23 47 02 C0 00 01 2F 01 26 01 2F 01 B7 01 2F 02 4A 01 2F 02 DB 02 02 01 26 3B 10 4E 74 68 4E 75 69 B6 60 62 B8 60 B8 61 62 B8 61 B7 53 AD 08 AD 06 CD FF 92 25 F9 81 55 7E CD FF 74 AD 0A 4E 7E 74 4E 7F 75 CD FF 56 81 4E 68 74 4E 69 75 AE 7A CD FF 65 4E 60 55 AD 0F 4E 61 55 AD 0A 4E 53 55 AD 0A AE 6A CC FF 65 6E 04 54 20 03 6E 02 54 B6 55 49 49 B7 55 49 A4 03 AE 04 42 AB 6A 97 CD FF 65 3B 54 EB 81

The bytes 3B 10 highlighted above govern the LeadOut value between frames. In my sampling of the IR signal from S3F80 code, it was seeing 13ms and the HCS08 code was seeing 30ms. The 30ms does match with the IRP notation and it was the value used in the S3F80 code so I left it. Perhaps cutting the 3B10h in half(1D 88) would help cut down on double-presses?

binky123 · Post by **binky123** » Wed May 02, 2018 6:30 pm

I changed the leadout timing to use 18 82 and the gap is now 13ms. I also removed some extraneous instructions and code is smaller than before. Timing looks the same.

Code.HCS08=20 1B 23 47 02 C0 00 01 2F 01 26 01 2F 01 B7 01 2F 02 4A 01 2F 02 DB 02 02 01 26 18 82 B6 60 62 B8 60 B8 61 62 B8 61 B7 53 AD 08 AD 06 CD FF 92 25 F9 81 55 7E CD FF 74 AD 04 CD FF 56 81 AE 7A CD FF 65 4E 60 55 AD 0F 4E 61 55 AD 0A 4E 53 55 AD 0A AE 6A CC FF 65 6E 04 54 20 03 6E 02 54 B6 55 49 49 B7 55 49 A4 03 AE 04 42 AB 6A 97 CD FF 65 3B 54 EB 81

alfi33 · Post by **alfi33** » Sun May 06, 2018 3:17 pm

binky123 wrote:I changed the leadout timing to use 18 82 and the gap is now 13ms. I also removed some extraneous instructions and code is smaller than before. Timing looks the same.

Code.HCS08=20 1B 23 47 02 C0 00 01 2F 01 26 01 2F 01 B7 01 2F 02 4A 01 2F 02 DB 02 02 01 26 18 82 B6 60 62 B8 60 B8 61 62 B8 61 B7 53 AD 08 AD 06 CD FF 92 25 F9 81 55 7E CD FF 74 AD 04 CD FF 56 81 AE 7A CD FF 65 4E 60 55 AD 0F 4E 61 55 AD 0A 4E 53 55 AD 0A AE 6A CC FF 65 6E 04 54 20 03 6E 02 54 B6 55 49 49 B7 55 49 A4 03 AE 04 42 AB 6A 97 CD FF 65 3B 54 EB 81

This seems to be working very well for me. Thank you!

maxim1 · Post by **maxim1** » Tue Jun 05, 2018 3:20 pm

Hi, google brought me here. Can you please advise if the IrScrutinizer protocol defined in IrpProtocols.ini would still be following or any change is required in regards to latest discoveries?

[protocol]
name=Xiaomi
irp={36k,290,msb}<2,-2|2,-3|2,-4|2,-5>(1000u,-2,D:8,F:8,C:4,2,^30m)+ {C=(D:4:4^D:4^F:4:4^F:4)} [D:0..255,F:0..255]

Thank you!

T_Tronix · Post by **T_Tronix** » Thu Jun 21, 2018 5:28 pm

I have this code for MiBox enter:

"CURSOR ENTER","sendir,1:1,1,38000,1,1,39,22,22,45,23,22,23,33,23,44 ,23,22,23,22,23,55,23,33,23,22,23,55,23,481","0000 006c 0000 000c 0027 0016 0016 002d 0017 0016 0017 0021 0017 002c 0017 0016 0017 0016 0017 0037 0017 0021 0017 0016 0017 0037 0017 01e1",,

But it doesn't play/pause...

anyone knows the code for these functions in that format?

mdavej · Post by **mdavej** » Thu Jun 21, 2018 9:05 pm

Looks like Global Cache format. Unfortunately, I have no idea how to generate that format. Have you tried getting the codes from their site (login required)?
https://irdb.globalcache.com/

Barf · Post by **Barf** » Fri Jun 22, 2018 12:03 am

"CURSOR ENTER","sendir,1:1,1,38000,1,1,39,22,22,45,23,22,23,33,23,44 ,23,22,23,22,23,55,23,33,23,22,23,55,23,481","0000 006c 0000 000c 0027 0016 0016 002d 0017 0016 0017 0021 0017 002c 0017 0016 0017 0016 0017 0037 0017 0021 0017 0016 0017 0037 0017 01e1",,

This is not a "format", it is a line of a CSV ("comma separated values") file, probably from irdb.globalcache.com. It contains a few fields, separated by commas (name, GlobalCache's sendir format, and finally the Pronto Hex format. In order to contain commas, the fields are enclosed by double quotes.

IrScrutinizer understands both the sendir and the Pronto Hex format. It can also generate them as export (select either "Text" or "Girr" as export format, check the "Pronto Hex" and/or the "sendir" checkbox"). It can also import such a CSV format file directly (Import -> Text -> Raw, select comma as "Field separator, Name as 1, "Raw signal col." as 3 (or 2)).

T_Tronix · Post by **T_Tronix** » Fri Jun 22, 2018 2:02 am

I know the format, I got all my codes already from global cache, I was just wondering if anyone had the ir codes for the play pause functions since they are not part of the global cache library.

yaworski · Post by **yaworski** » Fri Jun 22, 2018 2:49 am

All codes supported by the MiBox are already in this thread. These were extracted from the remote.conf file on the device.

There's no dedicated play/pause code. The device uses the OK button to act as play/pause in the players.

T_Tronix · Post by **T_Tronix** » Fri Jun 22, 2018 11:41 am

All is good got it to work, thanks.

MaskedMan · Post by **MaskedMan** » Wed Jun 27, 2018 12:51 am

Helping someone on RC, using this thread and its posted Hex commands.

http://www.remotecentral.com/cgi-bin/mb ... ,last#last

Post by **The Robman** » Wed Jun 27, 2018 8:40 am

Btw, Jeff2525 did eventually load up his RMDU file, but it appears that while this worked with the old OS, it doesn't completely work with the new OS:
http://www.hifi-remote.com/forums/dload ... e_id=25261