Controllers behave weirdly without internet

  • Thread starter Thread starter Raiden
  • Start date Start date
  • Tagged users None

Users Who Are Viewing This Thread (Total: 0, Members: 0, Guests: 0)

Yeah not sure how to replicate this for others outside of that is what happened to me and logs do show input activation in 3 separate collections.
 
You might want to setup this type of input on each of the 0-10v inputs you have push buttons or switches connected to. It will graph the voltage that is on the input but will not affect the other input settings you have setup but you will have a graph of the input voltage. It may help in troubleshooting the issue.

IMG_1725.png
 
Yeah, I suppose, although not sure to want end. Support does not seem interested.

So far what I got from support is this is a feature a not a bug. So when my internet goes out I am expected to have random audio alarms going off from leak sensors. Good job, Hydros lol.
 
Did you make you on switch box or was it purchased. The information may help if figuring out what is causing the inputs to trigger since it graphs the voltage that is applied to the input. As far as alarms I have my audible alarm set to off for all alarm modes but I do have lights that turn on depending on the alarm level in my living room so I still get a visual alert if I am home. Those do light up sometimes when I loose internet but they usually go back off and I have not noticed any pumps going off. My flow sensor on my return pump does not look like the pump ever went off and the power to it is from my kraken. The control signal has to go through the Sicce cloud from Hydros so that would not happen if internet is down but the pump seems to continue.
 
Switches I made, but we can even ignore that and just focus on rope leak sensor from Hydros, it is acting the same way.

And sure some things wouldn't work, like WiFi outlets or WiFi feeders. I expect that.

Just want to press more investigation from firmware team into the regular wired sensors, why loss of connectivity to Hydros servers affects sensor readings intermittently.
 
Raiden said:
Switches I made, but we can even ignore that and just focus on rope leak sensor from Hydros, it is acting the same way.

And sure some things wouldn't work, like WiFi outlets or WiFi feeders. I expect that.

Just want to press more investigation from firmware team into the regular wired sensors, why loss of connectivity to Hydros servers affects sensor readings intermittently.
Click to expand...
Unfortunately I do not have the rope leak detector but there is a sensitivity adjustment, It may be set too sensitive. I have seen where if they are set too sensitive they they can get set off when there is no leak. Here are graphs of my leak detectors but they are the point leak detectors. As I said in earlier post I had quite a bad day Sunday with internet going out but it normally goes out several times a day.

IMG_1730.png
IMG_1731.png
 
It only goes off like that during internet outage. In fact I have to get them quite wet for real alarm to happen.
 
The point leak detectors use the same input port on the controller. Not sure the exact differences that cause an issue with a rope detector and not a point leak detector. Looks like your TDS might have some points on it but I think mine is way too low to show anything. The graph is empty. That only gets a sample once an hour anyway I think.
 
Raiden said:
TDS is pointless, but yeah just floating in the sump of freshwater tank.
Click to expand...
I think it is really for the TDS readings on a RODI unit. That is where mine is. probe one is at output of DI and probe 2 is at output of RO unit.
 
Yeah. Originally it was for planted tank reasons. Where you can kind of tell when to do a water change based on TDS reading vs dosing fertilizer, etc.

sites.google.com

Perpetual Preservation System - TDS Meter

TDS meter Total dissolved solids TDS meter is an electronic device measuring electrical conductivity of water. Pure water has low conductivity and water with dissolved inorganic ions has conductivity higher. Organic contaminants, such as from fish and plants don’t have much impact on TDS
sites.google.com sites.google.com
 
Raiden said:
Yeah. Originally it was for planted tank reasons. Where you can kind of tell when to do a water change based on TDS reading vs dosing fertilizer, etc.

sites.google.com

Perpetual Preservation System - TDS Meter

TDS meter Total dissolved solids TDS meter is an electronic device measuring electrical conductivity of water. Pure water has low conductivity and water with dissolved inorganic ions has conductivity higher. Organic contaminants, such as from fish and plants don’t have much impact on TDS
sites.google.com sites.google.com
Click to expand...
I was referring the the Hydros meter. It was ment for the RODI monitor with the two probes. The original monitor on my RODI unit had 2 probes the same as the Hydros and they mount the same way.
 
So cause of this as far as I was told

When there is a connectivity issue, controller reboots eventually (haven't clarified no WiFi vs no Internet but WiFi).

If this controller is the ones with sensors, switches, etc. those seem to go unknown for a moment, which can end up turning some things on depending on fallback states, in case of leak detector I guess that's ON and you can't really do anything about. In case of something depending on sensor, you can potentially say turn on if unavailable, although this might not always be a good idea in my opinion.

I did find it strange that whole controller would reboot due to connectivity issues versus just some sort of reset (if that's needed even) of the WiFi module. Not quite clear if this is hardware limitation or just firmware choice. After all normally you laptop does not reboot if it fails to connect to the Wifi, and even simpler things like TVs or smart switches do not seem to do that.

Other possibility is to add some sort of hysteresis to the unknown sensors, i.e. if I can read leak sensor while its controller reboots, maybe I stick with its old reading for few seconds or whatever the reasonable limit would be. Maybe in something like ATO refill/drain it being a bit more sensitive to water level reading being more up to date, but mostly it does not need to be this panicky about loosing sensor momentarily.
 
Raiden said:
So cause of this as far as I was told

When there is a connectivity issue, controller reboots eventually (haven't clarified no WiFi vs no Internet but WiFi).

If this controller is the ones with sensors, switches, etc. those seem to go unknown for a moment, which can end up turning some things on depending on fallback states, in case of leak detector I guess that's ON and you can't really do anything about. In case of something depending on sensor, you can potentially say turn on if unavailable, although this might not always be a good idea in my opinion.

I did find it strange that whole controller would reboot due to connectivity issues versus just some sort of reset (if that's needed even) of the WiFi module. Not quite clear if this is hardware limitation or just firmware choice. After all normally you laptop does not reboot if it fails to connect to the Wifi, and even simpler things like TVs or smart switches do not seem to do that.

Other possibility is to add some sort of hysteresis to the unknown sensors, i.e. if I can read leak sensor while its controller reboots, maybe I stick with its old reading for few seconds or whatever the reasonable limit would be. Maybe in something like ATO refill/drain it being a bit more sensitive to water level reading being more up to date, but mostly it does not need to be this panicky about loosing sensor momentarily.
Click to expand...
You can setup a delay by using a virtual output between the output you are using for the depends on to turn off the return. and use the depends on it with the original output. Set up the dependency mode to off if off if the original output is normally off and off if on if the original was normally on. Then set a minimum off time to the delay you want. Then use the new output in the depends on of the return. I use the return output to shut off the others so that would be the only one I would have to add this. I am going to do some testing this weekend if I can get the chance.It will try powering down my cable modem for a while and later do the same with the router to see what happens in each case. In the mean time I will see if I can get an output setup to do this on the return pump power output.
 
It needs to be fixed at firmware level, not us doing hacks to work with it. And in case of leak sensor, nothing can really be done.

But yes in the mean time until it is fixed in any manner, if it will be at all, will need to keep that sort of possibility in mind when setting up systems with Hydros.
 
I do not consider doing what I suggested as a hack. Since there is no coding they do give us ways of customizing things for our unique situations. Not everyone would be happy having a delay on alerts such as leak detectors. I already had an output I had named tank leak setup to turn on if the tank leak detector came on so I added another output and used the tank leak output to hold it in off with the depends on setting. I also added a minimum off time of 30 seconds for now I may drop that to 15 seconds. I also checked the Tank leak lof and there are no turn on or turn off times logged. I think the log goes back a month. I know I have had lots of internets down times in a month. I also have an output for leaks in the garage where the RODI, mixing station, ATO pump and AWC pumps are located. It has in time on and off on August 3 but that one was me. None for any internet outages. Not a day goes by that my internet is not out even if it is just a few seconds at a time. I have not had any wifi connection issue since I got my new router. I know I have rebooted it in the last month so apparently a wifi reboot did not trigger any on or off sequences for those two outputs on my system. I also checked the controllers for last reboot. All but one are at the same time on the 25th which is when I did a reboot collective. Not sure why I did that. One was rebooted on the 26th not sure if I did it for some reason or not. I will keep an eye out for that. Also the output settings are below. The only one not shown is the actual Sicce output but I was already using the one setup as return to control the Sicce. I also can probably get rid on that one and control directly with the return off output. The reason for using a depends on forced off is the minimum off time does not start until the depends on is released. Normally the minimum off time starts as soon as the output turns off.

IMG_1740.png
IMG_1741.png

IMG_1747.png
IMG_1748.png
IMG_1749.png
 
Last edited:
Ok, I was able to power off the router for 5 minutes so that would be no wifi for 5 minutes. When I powered it back on there was no internet and that lasted for another 5 minutes. I did have 2 controllers out of 15 reboot during that time. One of them had 2 leak detectors connected to it. The RODI leak detector and the air conditioner leak detecter. I know that is not aquarium related but it gives me an alarm if the drain were to get plugged up on the AC so I can react before it gets under the wooden floors in the sunken living room. That would be a disaster if it did not get caught in time and it has happened once in the years I have been monitoring it. Neither of those inputs have anything on their graphs and I did not get any alerts for leaks. The only ones I got were for controllers being offline too long and alerts when they did connect back up. I did not see anything on any of the logs either. The only thing I can think of is I do have a collective with multiple controllers so it may react differently than a single controller but so far I cannot duplicate the issue you are having on mine.

IMG_1751.png
IMG_1750.png
IMG_1752.png
 
I check the graphs again and there is a spike on the input but the output for the two inputs on the controller that reset but I think I was making it more complex than it needs to be. The output I had setup as leak garage has both of those inputs as inputs but I had the if input/ dependency unavailable to turn off. So the output never turned on. So using just that output in the depends on for another output the wifi or internet issue would not have turned the output off. So it would only require setting up a output using the leak detector input and then set the if input/ dependency unavailable to off. Then the output would not come on and trigger the output off if it is because the input is unavailable as it would not be if the controller is rebooting. below is the graph and the output log and the setup of the output. So it would be simple enough to do this on a output like the return. Hope this helps. If you have questions let me know. Looks like I have it setup to block a reboot all along since I usually use this output to control the outputs in the garage so they will not come on if there is a leak detected by either of the leak detectors. I also had a 10 minute minimum on time setup so there would be no question that it triggered since it would stay on for at least 10 minutes.

IMG_1765.png
IMG_1766.png

IMG_1763.png
IMG_1764.png
 
Raiden said:
It needs to be fixed at firmware level, not us doing hacks to work with it. And in case of leak sensor, nothing can really be done.

But yes in the mean time until it is fixed in any manner, if it will be at all, will need to keep that sort of possibility in mind when setting up systems with Hydros.
Click to expand...
Raiden,
I know this is an old thread, but the internet issue described here happened to me as well. I experienced a power outage few weeks ago. When the power was restored, my internet service remained down for an additional four days.

After power returned, I noticed that my return pump began shutting off every few minutes. Since my skimmer depends on the return pump, it would also turn off and stay off for about two minutes before restarting. My ATO controller, powered by an XP8 would also tune OFF then ON and pump water to the sump every few minutes.

I have five controllers in the collective, and I could tell they were rebooting or power cycling because the MinNow pumps cycled every time it powered on.

The return pump, a Varios 4, is connected directly to the WE ver2 and depends on a virtual generic output with two level sensors: one for sump low and another for sump high.

To address my issue, I unplugged the Hydros system and ran the return pump directly from its pump controller. I also added two old wave makers to keep the tank somewhat going. Once the WiFi was restored, I reconnected the Hydros system, and everything started working perfectly again.

I’ve made a change where the return pump is no longer dependent on the virtual output. Unfortunately, this means it won’t shut off when the sump water level is too low or high.
The issue is that when running the return pump directly from the WE, there’s no option to assign a level sensor, as there is with the X4.

Thank you for explaining that the root cause is that sensors or switches go unknown to the controllers for a split second (moment) at the time the controller reboot.
It is just so weird that it happens when the collective is powered ON without internet service.
 
You must log in or register to reply here.
Back
Top