Been playing around some and tried to figure out some more of the 5508 packets
on JA-100 serie
Found that Panels and Application are using byte5-6 as ID
Seems like all Apps have the same ID (verified with 2 different users and phones)
byte3 seems to be which alarmstate you are sending from the panel/app (armed_away, armed_home, disarm)
byte4 is the remaining mystery, seems to be related to the user I get 6c from app and 6d from any
of my panels, wife get 70 from app and 71 from any panel
No idea what I can use the info to but maybe someone has an idea.
Hey there,
Iâve installed this integration for JA-100 and use binary_sensor.py @davidzadrazilâs PR and alarm_control_panel from @hsk-dkâs fork. Thanks to both of you for your work.
However, I found out that @davidzadrazilâs _triggersensorupdate() makes all devices âoffâ on every call, which he states to be fix for PIRs (works for them), but I also have reed switches connected through JA-118M (8 channel reed switch bus âmultiplexerâ) in my windows and the fix makes me unable to tell if the windows are currently open or closed (I only get âonâ state when the state of the respective window changes).
I found out that all of the JA-118M channelsâ 4th bytes are always even. On opening the second lowest bit is 1 whereas on closing the second lowest bit is 0. My PIRs seem to report only on first detection of motion and their 4th byte is odd.
Does anyone sorted this problem out or do I need to find correct byte configurations and recode the thing on my own?
Modification for 0.103 from @mattsaxon and replacing the BinarySensorDevice for BinarySensorEntity in binary_sensor.py and AlarmControlPanel for AlarmControlPanelEntity in alarm_control_panel.py, should be enough to get it partially working. Then you probably need to watch the logs for âNew unknown packet:âŚâ for your sensors and add the missing packets to the line 398.
Thanks, I will try it, but i think that without the manual reset by @davidzadrazil my PIRs will get stuck in âonâ state as they seem not to send any packet when there is not movement anymore (Do yours do?). And as you said Iâve got it to partialy work (all sensors recognized, added and triggered correctly, alarm control panel working), but I want to know whether my windows are closed or open not just if they were open/closed, I mean their actual current state not just state change.
@Nath sure, looking at the 5508 packets in JA-100 the current code only looks at sensors. But there is actually also available info showing interactions with Controlpanels and the MyJablatron app.
By looking at it I found that each panel has their own ID and in that sense the MyJablatron app acts as a controlpanel. Based on this I can see where an alarm was triggered. (byte5 and byte6)
Byte4 which usually shows on and off behaves a little bit different when it comes to the panels/app and seems to show which user has interacted with the panel/app.
Byte3 is giving away the state you are sending to the alarm (Armed_home, armed_away_ disarmed)
Based on this I can get info on who interacted with the alarm and how.
Ok, get it now and itâs pretty interesting. I was aware of IDs for panels (and other) but didnât know about apps (actually i stopped paying⌠)
@vbrenik I noticed in the issues section that somebody was interested in slightly âupdatedâ @plaksnorâs code and there is a recent fork which apparently works. (everything i mentioned above) So you donât need to do it on your ownâŚ
Iâm running HA from Docker and I observe odd behavoir with the plugin for JA-100 series. If I reload docker container with HA then all is fine. However, if I restart HA from UI then my âalarm_control_panel.jablotron_alarmâ entity becomes unavailable. Yet, âhidrawâ device is still reachable from within container. And I can see all Jablotron sensor states and their changes. Itâs just the control panel that stops responding.
Hello,
Thank for your work.
Alarm work well in my case with JA-100 !
I was wondering, is the trigger for opening/closing a gate or garage door provided for in the roadmap?
Not sure if it was covered well enough here. Few of my observations on Jablotorn behavior.
alarm can be disarmed from HA only if access code in the plugin matches user with the same or higher level of privileges as the one who armed the Alarm. I.e. you canât disarm from HA using âUserâ code if the alarm was armed from panel using âAdminâ code.
sensor states are not udpated in HA if you use âUserâ code. It only works with âAdminâ or âServiceâ code.
Is there anyone who have sniffed the packages sent when arming/disarming from a controlpanel or application?
I wonder if the bytes i found are same for everyone or depending on the system.
I found byte3 to be
ae :âarmed_homeâ
0c :âdisarmâ
2e :âarmed_awayâ
Hi,
I have a OASIS 80 and iâm debugging with OLINK to improve the HA integration. I want to share some infos.
Sending
\x02\x01\x8E
\x02\x01\x8E
(that means # #) can put the alarm into normal mode, exiting service mode or mantainance mode.
Sending
\x02\x01\x8A
will retreive lot of info about the sensor: this equals to press âDevicesâ on the Olink software while running in Service mode. Anyone can help me to decipher protocol? Goal is to check battery info on all sensorâŚ
Yes, i have few thousands of lines sniffed (i mean the useful onesâŚ)
Well, basically it acts (i mean the communication) like a normal section it just needs one alarm panel in case of two or more physical sections (which probably will slightly mix up the code you use - i think itâll need yaml). So letâs say you hit arm_away, it normaly sends a0 as to âfullyâ arm the main section and Jablotron do the rest (*you will see all sections are arming/armed/disarmed/pending in 5122*********07 packet), same as if you send 90 it disarm all section (b0 is not in use), on the other hand in case of arm_home, it should send a1/91 which dis/arms only the subsection (packet 5122*********07 reports only the one section armed).
What about triggered? I donât know if anybody noticed it but @plaksnorâs code not only doesnât have it but it also interrupts a siren after few seconds. (hadnât time to find the culprit)
EDIT: I quickly looked into your code and donât see the triggered bytes neither (i was wondering if itâs same code as mine ) FYI: In my case from armed 51220300*******07 => 51221b05*******07 (means it uses 1b05 accordingly to the sections position)
EDIT2: Sorry i have to correct myself*. In case of arming/pending/triggered the main section still reports as armed. Btw: bytes 21 and 23 in position of the section reports error, like a dead battery(disarmed 51222100*******07, armed 51222300*******07)
Binded sections: Ok, Iâll try to add support in the future. Do I understand correctly that when you partly arm (home/night) the main section, then the binded sections are armed fully (away) ?
What about triggered ?
I didnât sniff any âtriggeredâ bytes in my alarmâŚ
bytes 21 and 23 in position of the section reports error
Interesting - I think i can add binary âproblemâ sensors.