Xiaomi Mi Box Streaming Android TV Device

3FG · Post by **3FG** » Sat Feb 04, 2017 12:42 pm

The signals in the above ICT file are a Base 4 IR protocol:
{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}

Code: Select all

Menu      Xiaomi   Dev= 134   OBC= 4  
Up        Xiaomi   Dev= 134   OBC= 5  
Down      Xiaomi   Dev= 134   OBC= 6  
Exit      Xiaomi   Dev= 134   OBC= 7  
Home      Xiaomi   Dev= 134   OBC= 8
Left      Xiaomi   Dev= 134   OBC= 11 
Right     Xiaomi   Dev= 134   OBC= 12 
Select    Xiaomi   Dev= 134   OBC= 13 
Vol+      Xiaomi   Dev= 134   OBC= 14 
Vol-      Xiaomi   Dev= 134   OBC= 15 
Power     Xiaomi   Dev= 60    OBC= 204

The only correspondence I can see with the pasted file is that the signal contains 20 bits. I'm not aware of a standard executor that shoots this protocol, so Jeff would need to get a JP1 remote and interface cable or just get a OARUSB04G which doesn't need an interface cable. It also has the advantage that Base4 executors are easily written.

As I am writing this, I see Rob has replied with somewhat different advice. IMO, buying Nevo/Xsight remotes for use as a tester for IR signals is not a good idea. From what I can see, the quality as delivered is spotty, and getting connection to a computer may involve a registry hack and a firmware upgrade. Probably worth the trouble if the intent is to put the remote into daily use, but not as a test remote.
Edited to fix an error in the Dev number, and to provide a simpler IRP.

alanrichey · Post by **alanrichey** » Sat Feb 04, 2017 12:57 pm

Yes, there is little point in getting a Nevo, it will just generate the same ICT we have already. The Harmony does what we need. Although I would say that since they upgraded the software for IE11 it actually works very well now and doesn't require all the tweaks you are referring to.

Post by **The Robman** » Sat Feb 04, 2017 2:34 pm

alanrichey wrote:I did it yesterday and posted the link (6 posts back)

Ah thanks, I missed that part of the post. I had to edit the ICT file to make all the buttons show as I only saw 2 buttons when I opened it. I have also converted the file to IR format as this is much easier to work with:
http://www.hifi-remote.com/forums/dload ... e_id=14378

Using IR.exe, we can easily convert these signals into Pronto hex, which I believe you can use to add functions to the Harmony. If that is the case, here are a few samples. I am especially curious about the POWER button as it uses different codes to the rest.

POWER
0000 006C 0000 000C 0028 0015 0017 0016 0017 0038 0017 0038 0017 0016 0017 0038 0017 0016 0017 0038 0017 0016 0017 0038 0017 0038 0017 019E

MENU
0000 006C 0000 000C 0027 0016 0017 002D 0017 0016 0017 0022 0017 002D 0017 0016 0017 0016 0017 0022 0017 0016 0017 002D 0017 002D 0017 01F7

UP
0000 006C 0000 000C 0027 0017 0017 002D 0017 0016 0017 0021 0017 002D 0017 0016 0017 0016 0017 0021 0017 0021 0017 002D 0017 0038 0017 01E1

DOWN
0000 006C 0000 000C 0027 0016 0017 002D 0017 0016 0017 0021 0017 002D 0017 0016 0017 0016 0017 0021 0017 002D 0017 002D 0017 0016 0017 01F8

Post by **The Robman** » Sat Feb 04, 2017 3:03 pm

Regarding the #amlogic code posted earlier, I found the following on the web, don't know if this helps any.

https://forum.xda-developers.com/androi ... d-t3092663

https://www.reddit.com/r/androidapps/co ... ver_to_do/

https://web.archive.org/web/20141129124 ... oid-tv-box

mdavej · Post by **mdavej** » Sat Feb 04, 2017 4:44 pm

alanrichey wrote:Yes, there is little point in getting a Nevo, it will just generate the same ICT we have already. The Harmony does what we need. Although I would say that since they upgraded the software for IE11 it actually works very well now and doesn't require all the tweaks you are referring to.

Just a quick aside, as I don't want to take the thread off track. But I think you are mistaken about EZ-RC. I see no evidence that it has been updated in the last several years. It just happens to work in IE 11 compatibility mode. The site itself still claims that is only works in IE 7 or 8. And the registry hacks are still required in Win 8 and higher.

Jeff2525 · Post by **Jeff2525** » Sat Feb 04, 2017 6:47 pm

Apparently I did a decent job explaining on my last post, which unfortunately lead me to be completely submerged in information I do not comprehend! haha

Ignoring the JP1 for a moment, this is what I gathered:

-The file that I posted appears to be either one command or just the general commands that I already have access to.

-The only way that that file is helpful is if I edited it and add the additional commands myself. This would involve more Android knowledge, which I have some experience with (More basic knowledge of Android/Mac/PC code than remote codes, but I'm no expert). The problem for me with editing the file is that it would require rooting my box to access to edit and save protected files (which this is one of them). I do not believe the Mi Box US version has been rooted yet, and I have no inclination to root it.

-I am pretty much out of luck right now, unless I were to either edit the file or have a slight possibility if I were to have a JP1 or know someone with access to that and have the Mi Box.

As for the JP1, I am interested in getting it for help with this issue, as well as for possible future issues and personal interest. However my lack of funds means I will be needing to hold off on it (even if it means this issue can't be solved right away). Also, I switched to a Mac a few years ago. Since Mac is more sandboxed, if a hack is needed it would probably have to be done through Console commands to the app or OS itself rather than editing the Registry. Web browsers can change the client to IE, even if it isn't IE. That shouldn't be too much trouble, though it doesn't work 100% of the time. But right now that aspect, unfortunately, has to be more of a side note.

3FG · Post by **3FG** » Sat Feb 04, 2017 7:26 pm

Jeff,
I think among all of us there is still some confusion. So I'll set out here what I think is the situation:
1) You have a Xiaomi MDZ-16-AB
2) You also have a Harmony remote and Logitech supplies IR codes labeled as MDZ-16-AA but which work with your -AB model
3) You supplied Al Richey with your logon information so that he could clone your Harmony into his.
4) He captured the IR signals that you know work using a IR Widget, resulting in an ICT file.
5) The learns do not correspond to a known IR protocol.
6) I looked at the signals and described the IR protocol using the IRP notation, and listed the decoded signals.
7) The decodes do not seem to resemble the Android file that you posted, or the subsequent ones that Rob linked to. In other words, so far none of the Android files have provided value.

However, we know the form of the IR protocol, and we can generate test IR signals in a variety of formats. There are at least 256 candidate IR signals to try, and the fact that the Power function has different parameters to all the other signals means you could try another 256 candidates. Your choices for testing include:

1) Generate the candidate signals in Pronto Hex format (we would use IrScrutinizer to do that, and it is easy). If Logitech will load all of the candidate signals into a set of test remote configurations, then you have a way to test at no additional cost.
2) Obtain a JP1 capable remote, and we will provide to you a set of upgrade files that will shoot the test signals from the JP1 remote.
2a) The easiest way is to buy a used OARUSB04G for $8. This uses a USB connection and will work with a Mac.
2b) Another way is to buy a e.g. Comcast remote and also a interface cable This approach will also work with a Mac, but you'll need to wait for delivery of the interface cable from China.
2c) Buy a Nevo/Xsight class remote. I'm not sure if this will work with a Mac, but it uses a USB connection so no interface cable is needed. Again a wait for delivery from China

Jeff2525 · Post by **Jeff2525** » Sat Feb 04, 2017 8:05 pm

3FG, Thats mostly right except I didn't send in my login information, but he did seem to clone it himself.

I own an extra Comcast remote, however it looks like this:http://img.auctiva.com/imgdata/1/8/9/2/ ... 1941_o.jpg. Its the same, but a bit newer and looks to be just a color change. Could that work?

Also, I found a cheap interface cable shipped through Amazon . Again, would that work? It says for Raspberry Pi, but looks the same. http://www.amazon.com/JBtek-WINDOWS-Sup ... usb+to+ttl

I can't get it right away, but if I have 1/2 the stuff now, I might as well go in that direction.

I could also do that remote. That would probably be easier, but then again there is something nice about Frankenstien's monster electronics.

Post by **The Robman** » Sat Feb 04, 2017 9:18 pm

Jeff,
Your Comcast remote is a JP1 remote, but of course you will need to buy a cable for it. I don't know if the Amazon cable will work as nobody's tried it AFAIK, but maybe one of the Daves know different. Given that the cable is $7 and you can get the OARUSB04G for $8, I would go for the remote.

Back to your STB, am I right in understanding that you can use Pronto hex to add functions to your Harmony? If that is the case, could you try adding some of the learns that I posted and then confirm if they still do what they're supposed to do? If they do work, we can generate more of them for the rest of the 256 buttons in that remote signal profile for you to test.

I know nothing about the Harmony myself, but some Googling came up with the following:
https://www.avforums.com/threads/guide- ... e.1510519/

http://www.remotecentral.com/cgi-bin/mb ... d.cgi?7442

https://community.logitech.com/s/feed/0 ... 05AK2PsCAL

mdavej · Post by **mdavej** » Sat Feb 04, 2017 9:22 pm

Your remote is JP1, but that cable won't work. Unfortunately, our recommend cable from ebay has the wrong picture, so you can't go by that. 4 conductors won't work. We need the one with 6.

We're getting closer. If we had gotten lucky and this were a known protocol, you could have used your Comcast remote to discover new commands without needing a cable.

3FG · Post by **3FG** » Sat Feb 04, 2017 11:17 pm

It turns out that UEI has a setup code for this protocol. It is setup code Cable 3982 and uses executor PID 023B. In checking this setup code I realized that the Device numbers I originally posted were wrong, and furthermore, this protocol has 8 bits of Device and OBC, with a trailing 4 bit checksum. I've edited my earlier post to provide correct info and the better IRP.

The good news here is that the OBCs (except for Power) are 4,5,6,7,8,11,12,13,14, and 15. So a search for other functions would start with the obvious missing OBCs: 0,1,2,3,9,10. We don't know if other functions exist, but IMO this OBC list suggests that there are.

Regarding the interface cable, to add to mdavej's comment, we need Ground, TX, RX, and RTS lines. Additionally, we have had poor results with Prolific chipsets. FTDI chipsets work. Make sure to get a 3.3volt cable. If you get a 5 volt cable, make sure to never hook up the +5 to the remote. It is probably best to get a cable which has individual female connectors. If you order from China or Hong Kong, you can find ones on eBay for less than $5 shipped. From the US, it's faster but $8 or $9.

Is your Xfinity remote a JP2 or JP1.3 remote? You can use the 983 Signature blink back command to find out. If it is a JP1.3, the signature will be 3039. If it is a JP2, it will be 25562. A JP2 remote absolutely requires a FTDI chipset.

Post by **The Robman** » Sun Feb 05, 2017 11:06 am

Good find Dave. For anyone who's interested to see the binary for the signals, here it is...

Code: Select all

dev-bin  obc-bin  xsum dev obc Button
10000110 00000100 1010 134 004 Menu
10000110 00000101 1011 134 005 Up
10000110 00000110 1000 134 006 Down
10000110 00000111 1001 134 007 Exit
10000110 00001000 0110 134 008 Home
10000110 00001011 0101 134 011 Left
10000110 00001100 0010 134 012 Right
10000110 00001101 0011 134 013 Select
10000110 00001110 0000 134 014 Vol+
10000110 00001111 0001 134 015 Vol-
00111100 11001100 1111 060 204 Power

chuliu · Post by **chuliu** » Sun Feb 05, 2017 8:07 pm

The Robman wrote:Good find Dave. For anyone who's interested to see the binary for the signals, here it is...

Code: Select all

dev-bin  obc-bin  xsum dev obc Button
10000110 00000100 1010 134 004 Menu
10000110 00000101 1011 134 005 Up
10000110 00000110 1000 134 006 Down
10000110 00000111 1001 134 007 Exit
10000110 00001000 0110 134 008 Home
10000110 00001011 0101 134 011 Left
10000110 00001100 0010 134 012 Right
10000110 00001101 0011 134 013 Select
10000110 00001110 0000 134 014 Vol+
10000110 00001111 0001 134 015 Vol-
00111100 11001100 1111 060 204 Power

Thanks.
But should I go ahead and make a custom protocol to build a rmdu with the above info. Or are you making one? I am not sure if I can build that with a simple plain protocol builder within rm.

I am interested in being able to control the mi tv box with slingbox.

mdavej · Post by **mdavej** » Sun Feb 05, 2017 8:40 pm

Maybe this will be of use. It's an RMDU for the 3982 upgrade from UEI including the MAXQ622 protocol code (same thing 3FG found and posted earlier):

http://www.hifi-remote.com/forums/dload ... e_id=14380

Note, I assume this does not have the correct commands for the Mi.

3FG · Post by **3FG** » Sun Feb 05, 2017 11:06 pm

Dave,
The MAXQ610 executor isn't much help with a Slingbox. Sending a base 4 signal is more difficult with the S3F80, and so is sending a signal of 20 bits. The PID 00ED executor (Nokia) gives a good example of how to send a base 4 signal with a S3F80, but it sends 3 bytes rather than 2.5 bytes.

The 023B executor uses 0 fixed data bytes and 2 command bytes, which makes sense considering that both Device and OBC are changing in this IR protocol. I wonder if EZ-RC knows about Cable 3982? The Nevo/Xsight remotes use a S3F80.