reef-pi :: An opensource reef tank controller based on Raspberry Pi.

[GaryE]

Do you have ambient temperature reading? Any chance the heater is not able to keep up ? Or you think it’s a software issue ....?

I have the exact same setup in my second tank, and am not seeing this issue.

I don't believe it's ambient temp related at all. I can tell that my second heater is not coming on during the dip period. So I suspect it may be related to an issue with the kaza smart plug.

I'm planning to get a relay based power strip soon. The relay strip I built has mostly failed.

 

[Ranjib]

Do you see any error logged in the UI? if reef-pi is having a kasa driver issue, it will log an error that you can see through UI. Theres also a metric for the equipment state, to confirm if reef-pi was instructing the heater to be on or off , if you have mqtt or grafana/prometheus setup , this will be easier to confirm. Lastly, if its a kasa power monitoring capable switch, you can measure the current draw from the corresponding outlet to confirm if the heater was on or off.

 

[GaryE]

Do you see any error logged in the UI? if reef-pi is having a kasa driver issue, it will log an error that you can see through UI. Theres also a metric for the equipment state, to confirm if reef-pi was instructing the heater to be on or off , if you have mqtt or grafana/prometheus setup , this will be easier to confirm. Lastly, if its a kasa power monitoring capable switch, you can measure the current draw from the corresponding outlet to confirm if the heater was on or off.

No UI errors

Yes, during the period when the temp drops, one of my heaters does not come on.

this is my heater state graph. the yellow is my sump heater. It is off during he drop out, well, as you can see, it's off most of the time. Seems very slow to respond to the drop in temp.

I wonder if I should control that heater with the tank temp probe also. As I have it setup now, that heater is controlled by the probe in the sump.

 

[Ranjib]

Makes sense. Im trying to understand, if this indicated a logic bug that we should be fixing in reef-pi or its as expected. We have to consider the probe that's controlling this heater, current temp reading, hysteresis and check interval.

 

[GaryE]

Makes sense. Im trying to understand, if this indicated a logic bug that we should be fixing in reef-pi or its as expected. We have to consider the probe that's controlling this heater, current temp reading, hysteresis and check interval.

That probe could be at issue, however, it it reports a temp below the threshold for calling for heat, the heat should come on. If there was something in the logs to tell me, I'll look..

 

[Matevz Savarin]

Hi all anyone has an idea why my doser dosed 1.5l of soda it just ran and ran. After reset of pi doser works.. i did wc after 20 minutes. kh is 14 all alive for now. Reef pi works for 2 years without problems..

 

[Matevz Savarin]

Can I make an emergency macro and how that doser stops after lets say 10 secunds. I am dosing 3x4s daily..

 

[Ranjib]

Hi all anyone has an idea why my doser dosed 1.5l of soda it just ran and ran. After reset of pi doser works.. i did wc after 20 minutes. kh is 14 all alive for now. Reef pi works for 2 years without problems..

Can you share the details of your settings, and circuit/equipment used? any possibility that its a hardware issue and not software.

 

[Matevz Savarin]

I have jebao 3x dosing pump connected to pwm mosfet and connected to reefpi mikes hat. Its drivenwith led driver. And like I said it worked ok for 2 years and after reset it works again. Maybee there was something wrong with mosfet, so pwm signal turned off but mosfet stayed on..

 

[elysics]

Now that i read that, something similar happened to me. Alkalinity doser just wouldnt stop pumping, no matter what i did in the ui, luckily i was right there next to it on the couch when it happened, so i yanked the dosing line out of the reservoir in time before anything bad happened. Motor wouldn't stop turning until i pulled the system's (including the pi and the motors) power, and resumed turning when i plugged it back in until reefpi initialized. My chain was pi--i2c->pca9685-->lm298--> dc pump.

My theory at the time was that the command to set the pca9685's duty cycle for that channel back to zero got lost somehow and was never resent. But i forgot and never posted about it, hasn't happened since.

Thinking about it a bit further, the enable lines could be all tied together to a pin on the pi that would be directly controlled by reefpi and only be active when it is supposed to dose something. Would at least shift some of the error modes towards not dosing enough instead of dosing too much, and i think switching a direct pin is less likely to fail than switching something over i2c

 

[stefanm]

I have jebao 3x dosing pump connected to pwm mosfet and connected to reefpi mikes hat. Its drivenwith led driver. And like I said it worked ok for 2 years and after reset it works again. Maybee there was something wrong with mosfet, so pwm signal turned off but mosfet stayed on..

Are you using those MOSFET boards with jumper wires? If so desolder the header pins and solder the wires directly. Every problem I've had with reef-pi has been down to jumper wires the same goes for my previous Arduino controllers.

 

[GaryE]

Are you using those MOSFET boards with jumper wires? If so desolder the header pins and solder the wires directly. Every problem I've had with reef-pi has been down to jumper wires the same goes for my previous Arduino controllers.

This disaster of jumpers has been working quite well for over a year now.. I'll eventually get it all replaced, but I found a bit of lazy that I've been working on lately.

 

[attiland]

This disaster of jumpers has been working quite well for over a year now.. I'll eventually get it all replaced, but I found a bit of lazy that I've been working on lately.

This reminds me to to this

 

[Carlo Rios]

Ranjid I have been wondering for a while me should make a post where we can share our Reef-Pi. What do you think.

 

[Alaa]

Hi, hope everyone is fine
Regarding the dosing pump malfunction which is a serious issue, I was wondering what you are doing to safeguard against it. Do you run your dosing pump on timer to turn on just before the schedule then turn off again, is there a way to check the status and even better the duration that the dosing pump has been running and setting alarm if passed.

In a non related issue my pH graph shows spikes related to the dosing pump function, i tried to change the wire path but with no luck, also it shows wrong reading if i turn off my return pump mostly related to high water level in the sump during that time. is there a way to automatically adjust for those anomalies like setting upper and lower limit in the graph function to accept values with x % of mean
Regards

 

[attiland]

Hi, hope everyone is fine
Regarding the dosing pump malfunction which is a serious issue, I was wondering what you are doing to safeguard against it. Do you run your dosing pump on timer to turn on just before the schedule then turn off again, is there a way to check the status and even better the duration that the dosing pump has been running and setting alarm if passed.

In a non related issue my pH graph shows spikes related to the dosing pump function, i tried to change the wire path but with no luck, also it shows wrong reading if i turn off my return pump mostly related to high water level in the sump during that time. is there a way to automatically adjust for those anomalies like setting upper and lower limit in the graph function to accept values with x % of mean
Regards

I am working on a little different approach. I want to put dosing to a microcontroller. This will most likely an Arduino since I have most experience with that. Reef pi would only send a command via serial to dose so much with pump x. The reason for this if power resets the pump would not stuck in on position or any failure on communication could result in an emergency stop.
in the program I would keep checking scheduler status and get PI to send heartbeat and microcontroller to do the same. If either of them lost I would reset the microcontroller either via power on a relay or program on the microcontroller.
I am hoping from the above an extra safety layer.

 

[dmolavi]

What the frickity frank happened to my temp probes? Nothing changed, the system has been humming along for a couple of years and now the probes are off the charts....any ideas or do I need to get new ones?

 

[dmolavi]

Hmm this may lend a clue...low battery on the RTC. I am using the pico board. Could it be time to change the cell on it?

Log is full of failure to read probe messages.

 

[dmolavi]

More breadcrumbs...both sensors went south at the same time, and now pi can't detect them. They are connected to the pico board...could it be the problem with the low battery? I'm replacing that today to see if that resolves it...

 

[dmolavi]

Yeah def a probe issue but strange they would both tank at the same time. I'm hoping my pico didn't take a dump. Waiting on the stores to open to get a battery...