Worrying Bug

[Simon Still]

I'm assuming everyone has seen this already?

http://www.loxone.com/blog/enuk/first-aid-miniserver-inaccessible/

It's a pretty horrible failure mode that leaves the miniserver inaccessible like this.   I'm not sure I'm completely happy about the miniserver firmware updating automatically - there's a lot to be said for a stable installation that runs without issues.  Loxone are quite likely to introduce new bugs as well as solve old ones with updates.  Is it preferable (or even possible) to diable auto updates?

[TomM]

Yep, I agree this is not ideal.  whilst I don't have an issue with the fact that a bug slipped through, as a software guy myself I know that however much you test and test you can't always catch every single bug, I am concerned that this update could be pushed to the miniserver without my knowledge.

I rarely let any software updates install themselves automatically on any of my devices, not that i don't trust them, more so that I know they are happening according to my schedule and I can therefore react accordingly if an issue does arise.  I certainly think that for something as core as the Loxone Miniserver, which controls the complete infrastructure of my family home, it is fairly irresponsible for me to allow the miniserver to update automatically.

Incidentally, my miniserver is still on v7.1.9.30 and therefore I don't think I've had any auto updates anyway.  I am not aware that there is currently a way to disable the auto updates, until there is, I won't be upgrading the firmware of my miniserver any further just in case auto-updates deploy a serious bug again - of course this will happen when I'm not around and the family will be left without lights/heat/hot water!

On a completely different subject I spent all last night listening to the extractor fan from the bathroom because the toilet ventilation block is a pain to get on with.  I actually think it was my logic that was flawed to a certain extent, but I also think there's an intermittent issue with the logic block which causes the fan to remain on full time.  All good fun.

[Simon Still]

On Tuesday, 5 January 2016 11:20:35 UTC, TomM wrote:

Yep, I agree this is not ideal.  whilst I don't have an issue with the fact that a bug slipped through, as a software guy myself I know that however much you test and test you can't always catch every single bug, I am concerned that this update could be pushed to the miniserver without my knowledge.

I rarely let any software updates install themselves automatically on any of my devices, not that i don't trust them, more so that I know they are happening according to my schedule and I can therefore react accordingly if an issue does arise.  I certainly think that for something as core as the Loxone Miniserver, which controls the complete infrastructure of my family home, it is fairly irresponsible for me to allow the miniserver to update automatically.

Incidentally, my miniserver is still on v7.1.9.30

So is mine - I think that's the version with the bug - they've not yet rolled out an update.  Presumably if you're using the alarm clock block you'll have disabled it, if you weren't using it there's not an issue so the update is non-urgent (only now an issue if you've not seen the warning and add for the first time in the interim).

 

On a completely different subject I spent all last night listening to the extractor fan from the bathroom because the toilet ventilation block is a pain to get on with.  I actually think it was my logic that was flawed to a certain extent, but I also think there's an intermittent issue with the logic block which causes the fan to remain on full time.  All good fun.

That's interesting - I've not added that block to my bathrooms yet.  I've got whole house ventilation largely driven to 'boost' by humidity sensors but I'd like to add a boost based on time of occupancy as well.  It struck me that the toilet block will only work if you have a simple switch for the room lights.  If you use the push button to control scenes (and my bathrooms have different light levels) that block won't work.  

[TomM]

so the ventilation block is doing my head in.  Key issue in mine is that I have a central inline extractor that ventilates all 1st floor bathrooms.  I have tried to use a ventilation block per bathroom and just output the control to the extractor actuator - this is the scenario which was causing the fan to be on all night without explanation.  I'm sure my issue is just my poor logic and not the controller block.

That said, you can easily utilise the lighting controller with the ventilation block.  Just setup one of the lighting controller outputs to be the extractor fan (switch) and then connect a wiping relay (to generate a pulse as opposed to permanently on) which feeds the Tr of the ventilation block.  Annoyingly the trigger needs a pulse instead of a permanently on signal and its a signle pulse for 'start of session' and single pulse for 'end of session' - not the double pulse as it used to say in the docs.

You can of course connect anything to the outputs of the lighting controller so in the bathrooms we have outputs for the extractor fan and mirror heaters in addition to the lighting channels too.

[Simon Still]

I've treated my ventilation and heated mirror slightly differently - the AQs output from the lighting controller reports whether any lights are on.  I've taken a memory flag from that output.

The humidity sensor is linked to a Threshold switch which lets me set a level for 'high humidity' - again the output is a memory flag

The heated mirror is controlled by an AND block - if lights are on AND humidity is high then the mirror heats.

MVHR is on boost so long as Humidity is 'high'.  The output for boost is linked to 2 humidity readings, a 30 minute manual boost controlled by a user interface button and a signal from the cooker hood.  Multiple inputs controlling a single output appears fine.  

 

I'm just setting up the toilet ventilation boost - I'm trying to use a push switch block on the AQs output of the lighting block to get a pulse when lights go on and another when they go off (allowing the room to have scenes).  However, the issue comes when the miniserver reboots - the lighting block remembers the status of the lights (so if you reboot with them on they come back on) but there isn't a state change to generate a pulse to the toilet ventilation block so my sessions are getting out of sync.  Could your issue be similar? (and how do I solve this?)

[Simon Still]

On Saturday, 9 January 2016 13:49:56 UTC, Simon Still wrote:

II'm just setting up the toilet ventilation boost - I'm trying to use a push switch block on the AQs output of the lighting block to get a pulse when lights go on and another when they go off (allowing the room to have scenes).  However, the issue comes when the miniserver reboots - the lighting block remembers the status of the lights (so if you reboot with them on they come back on) but there isn't a state change to generate a pulse to the toilet ventilation block so my sessions are getting out of sync.  Could your issue be similar? (and how do I solve this?)

The solution I've come up with is to taken the remanence off the lighting block.  A bit of a bodge but not really an issue for a bathroom light (especially as I'm planning on putting Loxone on a UPS so power interruptions should be planned.  

[TomM]

Simon,

The humidity sensor is the component missing in my setup.  I only have 1-wire or IP easily available to me in the bathrooms, however I don't think there's a native 1-wire humidity sensor so I don't think I can add one without running more cables, not an option at this point. If there's an IP connected humidity sensor then maybe I could install that with a bit of work.

I am also finding sync issues with the fan and bathroom session [discussing bathroom 'sessions' on a tech forum amuses me! - sometimes I wonder whether I'm going to far with the home automation, I know my wife does!]  my solution to this thus far is a manual reset on the app.  The sync issue doesn't happen that often so not really bothered at the moment but I would like a better way to do it.  I think in an ideal world the ventiallation block would actually be integrated into the lighting controller somehow but I doubt that's about to happen any time soon so I'll think up some more logic instead.

[Simon Still]

On Saturday, 9 January 2016 21:25:49 UTC, TomM wrote:

I am also finding sync issues with the fan and bathroom session [discussing bathroom 'sessions' on a tech forum amuses me! - sometimes I wonder whether I'm going to far with the home automation, I know my wife does!]  my solution to this thus far is a manual reset on the app.  The sync issue doesn't happen that often so not really bothered at the moment but I would like a better way to do it.  I think in an ideal world the ventiallation block would actually be integrated into the lighting controller somehow but I doubt that's about to happen any time soon so I'll think up some more logic instead.

I've just realised there's a simple solution to this as I'm now getting the same issue. 

Take a Memory Flag from AQs on the Lighting controller.  This gets an analogue output of the lighting scene but you can switch the memory flag to be digital in properties.

Feed this memory flag to the input of the toilet ventilation block.  That means the ventilation is controlled by light status not switching status and you then shouldn't need to reset any longer.  (It probably doesn't need the memory flag - you could just run the ventiation block directly from AQs but all my ventialtion is on a separate page)