433 Issues - Homey blocks all 433 communication

JurgenVanDeRijke

Currently I use 0.8.20 as firmware. 
Trying to get Somfy working I have encountered several issues (see thread Somfy range). Apparently Somfy influences KAKU. But also voice influences the functionality of KAKU. It all seems to be a general 433 issue. 
Yesterday I had to power down the Homey. Again KAKU stopped working (which it does after some time and/or issue as mentioned in the Somfy range thread). The main issue was however Homey disturbing all 433 communication which also blocked my kaku comm. So my kaku remote didn't work anymore (only when very very close to receiver, homewizard could not control kaku anymore.

It seems that there is a serious issue with 433 communication which makes the Homey send out garbage? in the 433 range preventing other devices to use it. 

If Athom needs a log files please let me know. Until a new version of (app|firmware?) I have to leave my Homey down

Reinier

@JurgenVanDeRijke ;
This problem appears when de microcontroller crashes when sending. When this happens the 433 carrier stays high causing communication problems with other 433 devices. I did a lot of testing yesterday and it looks like the instability issues are gone. Im also planning on implementing some sort of watchdog when de mcu crashes which will prevent issues like this. 

The instability issue is fixed in 0.8.21. 

JurgenVanDeRijke

@Reinier I will test and send results asap! Thanks for your swift reply. 

mruiter

.... Ship still did not leave port with 8.21 i'm sitting in the dark every night. Kicked out Homewizard and i'm to lazy to get off the couch to turn on the lights

itsamaik

Ran into this issue yesterday as well, I was a bit confused and thought my remote's battery was dead. Hope the new version will fix it

TheoDeKoning

Reading this thread now, I am really furious.

Knowing that this problem would be going on till 0.8.21 and than give 0.8.20 free on a Friday is stupid.
If I had known than I had stay on 0.8.17 without this problem. Then praying for luck that has not game automatic.

It make my Homey now worthless because one on command to a device and nothing from KaKu is working any more.
Even the wall switches and ICS-1000.

Athom has now crossing my already stretched line.
Now my Hhomey is switch off, I take a few days of from this form to calm down. There is nothing to see here now.

MarkVanDeVen

I think I now finally got rid of the issue. I took me at least 2 complete resets (including all saved settings), but after that i was able to properly register all my devices and it responds to the lights on/off commands.

will see if it remains stable, but at least KaKu works again...

AndyWilde

I think the issue is still there,  and it depends on how fast homey requests transmit signals to be sent,  I have notices that if I ask one signal to be sent before completing the fist I get this 433 lock up.  Thus is mainly when changing DIM values as it can send many signals

MarcoWijk

Emile said in 0.8.20 topic that it's only a few of us experiencing this. It seems to be a no big thingy, at least not for them.
And it should be fixed in 0.8.21,  but read: after very few tests we will send out a new version which absolutly does NOT ensure it's fixed.  It's the same story over and over again since they've been shipping and updating... after bricking Homey once and 1 version where 433 did actually work, I do not have faith it will be fixed accordingly. A "clever" jump from 0.8.17 to 0.8.20 ( i even doubt there is a .18 and .19) but it all keeps getting worse.

@TheoDeKoning I'm 100% in agreement with u dude!

mruiter

MarcoWijk said:

Emile said in 0.8.20 topic that it's only a few of us experiencing this. It seems to be a no big thingy, at least not for them.
And it should be fixed in 0.8.21,  but read: after very few tests we will send out a new version which absolutly does NOT ensure it's fixed.  It's the same story over and over again since they've been shipping and updating... after bricking Homey once and 1 version where 433 did actually work, I do not have faith it will be fixed accordingly. A "clever" jump from 0.8.17 to 0.8.20 ( i even doubt there is a .18 and .19) but it all keeps getting worse.

@TheoDeKoning I'm 100% in agreement with u dude!

Just a few is bullshit , he should trademark that term in combo with some off his posts lol
To bad "we ship in two weeks"  already has been trademarkt bij Butterfly labs and there bitcoin miners ( real geeks will know what i mean)

forum is filling up with kaku en rf problems. for instance look at zwave also crashing with lots of people and devices. Please put some effort in real dev, not keychains and other non important stuff.

Still loving my more death then alive homey

djesko

And blocks my car remote,  switch off homey and all good 

honey

Let`s install it in a car park. Can build a business model on it...  

bvdbos

So Homey hangs sending 433 signals, question is why, what process is hanging and how to trigger this situation. Also, Marco only had two apps installed iirc: wikipedia and kaku. Perhaps it''s a specific kind of kaku-device which causes this situation? For the people who have problems with kaku: What devices (remotes, sockets etc) do you have? And do you also have high processor-stress (load-avg above 2)?

MarcoWijk

BasVanDenBosch said:

So Homey hangs sending 433 signals, question is why, what process is hanging and how to trigger this situation. Also, Marco only had two apps installed iirc: wikipedia and kaku. Perhaps it''s a specific kind of kaku-device which causes this situation? For the people who have problems with kaku: What devices (remotes, sockets etc) do you have? And do you also have high processor-stress (load-avg above 2)?

@BasVanDenBosch ...eh if i'm the Marco you are refering to. I also had the Action and Elro app installed, next to those 3 no other 433 apps. 

The high avg load is gone for now. Someone contacted me and told me there was a nasty bug in his app, which is fixed btw. But still, after a reset 433 doesn't seem to work.

@djesko Funny, i also had the feeling my 'normal' 433 remotes were blocked for a few moments... Could it be that Homey continiously is sending bursts of just 1 433 signal, therefor blocking the airwaves?

MarcoWijk

@mruiter +1 on the keychain... Have a really expensive one on my cubboard at the moment (if i put a chain on it, that is)

MarcoWijk

Okay! Just found out that Homey is blocking every kind of 433. I noticed that my weatherstation, which works on the 433 frequency, lost contact with my outside sensors. Pairing them wasn't possible with Homey ON. After unplgging Homey they paired almost instantly. Homey is filling the airwaves with some kind of continious signal blocking the rest. A problem which should be adressed immediatly, not to get my KaKu devices working again, but because it could in potential be blocking all 433 apliances.. Read: Fire detectors, alarm, etc.!!

bvdbos

""Someone contacted me and told me there was a nasty bug in his app, which is fixed btw." -> what app is this?

As for the blocking, that's what I was saying, we need to find out why some of the Homey's keep sending the 433 signal and why the rest don't do this? What's specific for your situation? You mention you had an app which behaved badly? But yesterday you performed a full reset, only installed kaku and then load-avg was over 2? Might this high loadavg and the 433 hanging somehow be connected?

MarcoWijk

BasVanDenBosch said:

""Someone contacted me and told me there was a nasty bug in his app, which is fixed btw." -> what app is this?

As for the blocking, that's what I was saying, we need to find out why some of the Homey's keep sending the 433 signal and why the rest don't do this? What's specific for your situation? You mention you had an app which behaved badly? But yesterday you performed a full reset, only installed kaku and then load-avg was over 2? Might this high loadavg and the 433 hanging somehow be connected?

I guess the high load avg and 433 are connected, it is possible. As for the mentioned app, the bug is fixed and he wasn't sure it was the cause anyhow. I also guessed it wasn't because even when i didn't have it installed the load avg went beserk. It's regular now btw. My guess is as good as yours....  

bvdbos

@djesko @mruiter @TheoDeKoning ; @AndyWilde @JurgenVanDeRijke ; @itsamaik ;: we need to find out what apps / versions and devices you have in common. As @MarkVanDeVen somehow got past this point, there seems to be a solution. Perhaps it's something with a race-condition on a flow? Did you start over completely : no apps besides kaku, no flows, no other devices, firmware reset by turning upside down and not restoring data?

MarkVanDeVen

BasVanDenBosch said:

@djesko @mruiter @TheoDeKoning ; @AndyWilde @JurgenVanDeRijke ; @itsamaik ;: we need to find out what apps / versions and devices you have in common. As @MarkVanDeVen somehow got past this point, there seems to be a solution. Perhaps it's something with a race-condition on a flow? Did you start over completely : no apps besides kaku, no flows, no other devices, firmware reset by turning upside down and not restoring data?

I have bad news... I thought I had solved it by completely reinstalling/powercycling and pairing all devices again. And indeed, it worked.

When I just tried to turn my lights on again, it stopped again half way in the process. This time the 433mhz wasn't completely blocked, but homey didn't want to turn anything on/off either... So I guess i still have an issue...

bvdbos

What's your loadavg? Also: did you install any other apps or create any flows in the meanwhile? Something must have triggered the bad behaviour....

bvdbos

Or something was wrong already and was building up a queue...

MarcoWijk

This morning the load avg was about 0.4. I haven't installed any apps after the reset yesterday, aside from the kaku. No flows created after the reset. For now i unplugged homey, for i do not know what it is blocking or not. Nothing to do then wait for an update and pray that they did fix the issue.... or demand a refund  

[edit]

An update won't be released this weekend though. It's a weekend at the moment. All starting up a company and no play, isn't do-able ofcourse  

djesko

BasVanDenBosch said:

@djesko @mruiter @TheoDeKoning ; @AndyWilde @JurgenVanDeRijke ; @itsamaik ;: we need to find out what apps / versions and devices you have in common. As @MarkVanDeVen somehow got past this point, there seems to be a solution. Perhaps it's something with a race-condition on a flow? Did you start over completely : no apps besides kaku, no flows, no other devices, firmware reset by turning upside down and not restoring data?

i only have the kaku app installed and using a old NEXA socket  

MarcoWijk

BasVanDenBosch said:

Or something was wrong already and was building up a queue...

Could be. I haven't noticed it untill the RF bug surfaced. All in all, i think it could well be a combination of 433 issues and the flows triggering it. Alas with no further input from "the team" it's a guessing game. Unbricking a Homey  didn't take so long, but this bug is really, really bugging me (as you may have noticed  ) And without a log of some sort it's the trial and error method. And i'm running out of ideas....

Reinier

Guys,

I first want to say that you guys are doing a pretty good job testing and helping me fix this issue!

I've looked into it and it seems that the problem arises when sending long signals following by a reception event(For example, some noise). The only way to fix it is by power-cycling your homey. Unfortunately it slipped through the testing stage and ended up in your homeys. I've fixed the issue in the next update and made some changes to my test scripts.

My sincerest apologies and again,
thanks for helping me out!

AndyWilde

@Reinier  It is also possible to block the 433 communication by homey sending a tx before the previous tx has completed.  IS this also addressed in the next update?

MarcoWijk

  @Reinier Glad to hear the issue is fixed... again... for you said it was already fixed on march 1.https://forum.athom.com/discussion/comment/12822/#Comment_12822

   All i can conclude that al of our whining helped to fix a bug that otherwise would not have been fixed in the next update. So i'm a bit sceptic...

And i would like to know, and many of us i believe,  WHEN the update will be deployed? 

djesko

MarcoWijk said:

  @Reinier Glad to hear the issue is fixed... again... for you said it was already fixed on march 1.https://forum.athom.com/discussion/comment/12822/#Comment_12822

   All i can conclude that al of our whining helped to fix a bug that otherwise would not have been fixed in the next update. So i'm a bit sceptic...

And i would like to know, and many of us i believe,  WHEN the update will be deployed? 

We need this new update ASAP !  ( and the zwave fix so i can replace my HC2 ) 

bvdbos

@MarcoWijk : I think you see this the wrong way, whining doesn't work but providing as much information as possible and preferably give a testcase does...

Emile

The update will be available Monday (tomorrow) if all our tests pass. Yes we do run tests It's just a tradeoff between shipping fast and testing throughly, and during this period our focus is more on shipping updates faster.