New milight bulb - different protocol

For problems with LYT/WiFi itself, NOT your project

Re: New milight bulb - different protocol

Postby woodster » Tue Apr 26, 2016 9:31 am

Hi there,
Any Update, or ideas?
I tried to read out the firmware from the bulb, but it has the read out protection enabled.

Martin
woodster
 
Posts: 30
Joined: Mon Sep 07, 2015 5:54 am

Re: New milight bulb - different protocol

Postby pietromoscetta » Tue Apr 26, 2016 4:47 pm

Martin,

no news up to now.
Still very busy to complete new projects.
Don't know if krulkip has some news.

Pietro
pietromoscetta
Site Admin
 
Posts: 65
Joined: Tue Jun 30, 2015 3:03 pm

Re: New milight bulb - different protocol

Postby woodster » Sun May 29, 2016 9:51 am

Hi,

Any update or news Pietro / Krulkip?
Martin
woodster
 
Posts: 30
Joined: Mon Sep 07, 2015 5:54 am

Re: New milight bulb - different protocol

Postby pietromoscetta » Mon May 30, 2016 1:29 pm

Martin,

unfortunately no news.
We are totally busy to finalize another new product so no time to work on this matter.
As you can imagine it is not a priority.

Pietro
pietromoscetta
Site Admin
 
Posts: 65
Joined: Tue Jun 30, 2015 3:03 pm

Re: New milight bulb - different protocol

Postby woodster » Mon May 30, 2016 1:32 pm

Hi Peitro,

Ofcourse, no problem you have my full understanding.
New Product, what are you up to?
Martin
woodster
 
Posts: 30
Joined: Mon Sep 07, 2015 5:54 am

Re: New milight bulb - different protocol

Postby pietromoscetta » Mon May 30, 2016 2:59 pm

Hi Martin,

just few weeks and you will discover it....... :mrgreen:

P.
pietromoscetta
Site Admin
 
Posts: 65
Joined: Tue Jun 30, 2015 3:03 pm

Re: New milight bulb - different protocol

Postby woodster » Thu Aug 04, 2016 7:37 am

Hi guys,

It has been a while - I've been busy with work and vacation.... to little time for projects unfortunately....

@Pietro: I have seen your new app, it looks very promising.

@Pietro and Krulkip: Have you had the chance to look at the protocol, any news?

Martin
woodster
 
Posts: 30
Joined: Mon Sep 07, 2015 5:54 am

Re: New milight bulb - different protocol

Postby psherratt » Mon Sep 12, 2016 2:22 pm

Hey guys,

I purchased a bunch of these new bulbs and a little frustrated not having the wifi bridge yet.. I contacted Martin after reading his post on hackaday and have made /some/ progress:

Code: Select all
data = """
9B 59 E6 0B CF AD AD E3 84
9C 37 05 5C 82 ED 74 82 6D
A1 16 A8 49 D9 B8 24 DA 1E
A3 51 EE 13 C7 A5 84 DB 67
A4 2F ED 64 7A E5 63 7A 64
A8 33 F1 60 7E E9 64 7E 89
AA 00 A7 7B EA BD B4 AA 41
BA F0 D7 4B 1A ED D4 DA 31
C6 DC C3 5F 06 D9 BE C6 57
D4 FF BD 14 4A B5 3E 4A 9F
DE C4 7B 37 AE 81 75 6E 56
DE C4 7B 37 AE 81 84 6E 25
E5 CA 4C 15 8D 6C DF 8E 1D
E9 CE 50 11 91 70 FD 92 3F
EE B4 6B 47 9E 71 83 5E 34
EF FD 9A E7 73 51 42 87 51
F5 BA 7C E5 BD 9C D9 BE 77
F5 BA 7C E5 BD 9C EF BE 7D
FE A4 9B 17 CE A1 AB 8E 4C
"""

msgs = []

ML_XOR = [0, 4, 44, 24, 36, 36, 36, 36, 0]
ML_OFS = [0, 4, 44, 12, 44, 44, 44, 44, 0]

def ml_decrypt(payload):
  key = payload[0]
  for i in xrange(1,9):
    payload[i] += (key + ML_OFS[i]) ^ ML_XOR[i]
    payload[i] %= 256

msgs = []
for l in data.split('\n'):
  if len(l) < 26: continue
  payload = map(lambda x: int('0x' + x,0), l.split(' ')[:9])
  ml_decrypt(payload)
  msgs.append(payload)

for msg in sorted(msgs, key=lambda x: x[0] & 3):
  print msg


output:

Code: Select all
[156, 219, 233, 12, 110, 217, 96, 110, 9]
[164, 219, 233, 12, 110, 217, 87, 110, 8]
[168, 219, 233, 12, 110, 217, 84, 110, 49]
[212, 219, 233, 12, 110, 217, 98, 110, 115]
[161, 183, 137, 254, 194, 161, 13, 195, 191]
[229, 183, 137, 254, 194, 161, 20, 195, 2]
[233, 183, 137, 254, 194, 161, 46, 195, 40]
[245, 183, 137, 254, 194, 161, 222, 195, 108]
[245, 183, 137, 254, 194, 161, 244, 195, 114]
[170, 170, 161, 41, 220, 175, 166, 156, 235]
[186, 170, 161, 41, 220, 175, 150, 156, 235]
[198, 170, 161, 41, 220, 175, 148, 156, 29]
[222, 170, 161, 41, 220, 175, 163, 156, 52]
[222, 170, 161, 41, 220, 175, 178, 156, 3]
[238, 170, 161, 41, 220, 175, 193, 156, 34]
[254, 170, 161, 41, 220, 175, 185, 156, 74]
[155, 244, 209, 202, 178, 144, 144, 198, 31]
[163, 244, 209, 202, 178, 144, 111, 198, 10]
[239, 244, 209, 202, 178, 144, 129, 198, 64]


I need to take a break from this for a while. Hopefully someone is able to carry on or has any more clues.. it feels like the solution should be a little simpler.


Cheers,

Paul
psherratt
 
Posts: 1
Joined: Mon Sep 12, 2016 1:33 pm

Re: New milight bulb - different protocol

Postby carl0s » Sun Nov 13, 2016 12:40 am

Hey Woodster. Been following / waiting for something to happen on this front myself..

It seems I can do a replay on the 2.4GHz signal, for the RGB+CCT bulbs.

I suppose the next step is for me to figure out how to work Inspectrum and build something in GNU Radio that can control the bulbs. There is a little stick (not the yardstick) that can be programmed a bit like an SDR (but it's not an SDR.. it's hardware and has to be programmed with the parameters or something, and I can't remember its name), but maybe that could work.

Anyway here's a video I shot. Any ideas / tips for where to go next?

https://www.youtube.com/watch?v=NumDn1V0A38
carl0s
 
Posts: 1
Joined: Sun Nov 13, 2016 12:34 am

Re: New milight bulb - different protocol

Postby sidoh » Wed Mar 22, 2017 7:13 am

This was one of the places that showed up on Google often when I was researching the new protocol, so I wanted to share that I think I've managed to reverse the scrambling. I detailed the structure a bit here:

http://blog.christophermullins.com/2017 ... -protocol/

I put a bunch of packet captures and the hacky ruby scripts I used during the process here:

https://github.com/sidoh/milight_decoding

There's some Arduino code for encoding/decoding packets here:

https://github.com/sidoh/esp8266_miligh ... matter.cpp
sidoh
 
Posts: 1
Joined: Tue Mar 21, 2017 7:22 pm

PreviousNext

Return to Installation & Troubleshooting

Who is online

Users browsing this forum: No registered users and 1 guest

cron