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

[ryangonzalezrz]

So, I came to tell you what happened. My light control was not turning off completely and was not turning on the light at maximum power. The problem with the maximum power was the VCC of the PCA9685 board, which was connected to 3V3 and should have been connected to 5V, so it started turning on the light at maximum power. The problem with the light not turning off 100% was on the PCB board. Somehow, some rails were leaking power to other rails, and so it was always running even though the MOSFET was in low mode. I had to test the best rails to use and in the end it worked. Thanks for all your help @BenB

 

[bishoptf]

Does anyone know what changed in bookworm to hose the GPIO stuff up? Why oh why would they change that....when I have more time and it gets colder I will try to figure this out and see what we can do to update it...should work on the new OS...

 

[Simonv92]

Hi All! I'm currently doing some safety test on my Reef-Pi controller..
I've seen that if I disconnect the temperature sensor which controls the heater this one kicks in without any control.
How can I add a safety feature to disable the heater if the temperature reading goes below 2° - which means sensor disconnected of faulty?
Thank you!

P.S. I already tryed to create a second temperature controller but the second one disable the action of the first..

 

[TastyScrants]

I’m thinking of building one of these to learn about programming.

I wondered if anybody knows if it’s possible to set up some redundancy in the DIY doser. If the motor fails and sticks on, is it possible to have the doser shutdown after running for a set period of time?

Thanks

 

[biznatch]

I’m thinking of building one of these to learn about programming.

I wondered if anybody knows if it’s possible to set up some redundancy in the DIY doser. If the motor fails and sticks on, is it possible to have the doser shutdown after running for a set period of time?

Thanks

You won't really learn about programming running reef-pi. There's no programming to be done, unless you want to fork the code and make your own changes.

And the doser pump runs by getting power from reef-pi. You set it to run for x seconds, then it stops. It's literally not possible for the motor to 'stick' on, since it's not getting any power.

 

[elysics]

I’m thinking of building one of these to learn about programming.

I wondered if anybody knows if it’s possible to set up some redundancy in the DIY doser. If the motor fails and sticks on, is it possible to have the doser shutdown after running for a set period of time?

Thanks

Yes it is possible. Whatever motor driver chip/board you use has a pin labeled something like "EN" for enable.

Rather than just wiring that pin to 3.3v, making the chip constantly turned on, you can wire that pin to a GPIO and set a timer in reefpi that only enables that chip at certain times for a certain duration.

 

[iamdan]

Hi all, having some issues with an Eheim feeder with Robo-tank hardware and reef-pi…

When I initially setup reef-pi/robo-tank a few years ago, I bought a feeder, soldered the wires as per instructions to the feeder circuitry, setup the pca9685 in reef-pi under Drivers with address of 64 and frequency of 150 and called it “Extra Ports”, then under Connectors > Outlets setup “Feeder” using the “Extra Ports” driver and pin 0 and all worked good.

Shortly after that I put the feeder away as I went a different direction with tank/fish and feeding pellets wasn’t going to be needed.

EDIT: all hardware etc was purchased during pandemic and in Western Australia because a lot of things weren’t available or able to be delivered etc etc whenever I bought something I bought at least 2 so I have 2 raspberry pi’s, robotank hardware and Eheim feeders incase something goes wrong and I have extra equipment to diagnose issues or replace etc

I also tested the 2nd Eheim feeder, raspberry pi, robotank hardware and reef-pi and all was working. Once I had the main one set up I imaged it to the backup and tested and then put away.

Fast forward to now and my tanks inhabitants have changed again and I was wanting to use the feeder, plugged in, toggled the “feeder” and nothing happens. Power is on the feeder unit and pushing manual feed works.

Power-cycled robotank/reef-pi and no change, tried the 2nd feeder and no change

Connected up the backup robotank/reef-pi and tried both eheims with that and no change.

Running the latest released version of reef-pi with Rasbperry Pi Bullseye - kernel 6.1.21-v8+

Is there any update or anything in bullseye that could cause this issue on both boards?

 

[robsworld78]

Hi all, having some issues with an Eheim feeder with Robo-tank hardware and reef-pi…

When I initially setup reef-pi/robo-tank a few years ago, I bought a feeder, soldered the wires as per instructions to the feeder circuitry, setup the pca9685 in reef-pi under Drivers with address of 64 and frequency of 150 and called it “Extra Ports”, then under Connectors > Outlets setup “Feeder” using the “Extra Ports” driver and pin 0 and all worked good.

Shortly after that I put the feeder away as I went a different direction with tank/fish and feeding pellets wasn’t going to be needed.

EDIT: all hardware etc was purchased during pandemic and in Western Australia because a lot of things weren’t available or able to be delivered etc etc whenever I bought something I bought at least 2 so I have 2 raspberry pi’s, robotank hardware and Eheim feeders incase something goes wrong and I have extra equipment to diagnose issues or replace etc

I also tested the 2nd Eheim feeder, raspberry pi, robotank hardware and reef-pi and all was working. Once I had the main one set up I imaged it to the backup and tested and then put away.

Fast forward to now and my tanks inhabitants have changed again and I was wanting to use the feeder, plugged in, toggled the “feeder” and nothing happens. Power is on the feeder unit and pushing manual feed works.

Power-cycled robotank/reef-pi and no change, tried the 2nd feeder and no change

Connected up the backup robotank/reef-pi and tried both eheims with that and no change.

Running the latest released version of reef-pi with Rasbperry Pi Bullseye - kernel 6.1.21-v8+

Is there any update or anything in bullseye that could cause this issue on both boards?

Hi Dan, I just replied to your email, I was thinking it was a related to an updated version of the Eheim feeder however now I don't know what to think. If you tested them both and they were working very odd both aren't. You also mentioned you tested a DC port with a pump and then tried using it's GND as a signal for the feeder but that didn't work either. You did the correct test, the feeder port is nothing more than a low powered DC port. The signal pin is a Ground being switched.

I can't imagine it's a problem with reef-pi as many use a feeder, Bullseye works good, the controller hardware seems ok, the feeders seem ok, hmmm... Did you check continuity between connector that plugs into controller and feeder PCB on all 3 wires to make sure there's no breaks. Sometimes those green connectors can cut the wire so it's only left with a strand or two holding it in place making a bad connection but it looks ok. Maybe try removing the connector and inspect the wires.

 

[iamdan]

Hi Dan, I just replied to your email, I was thinking it was a related to an updated version of the Eheim feeder however now I don't know what to think. If you tested them both and they were working very odd both aren't. You also mentioned you tested a DC port with a pump and then tried using it's GND as a signal for the feeder but that didn't work either. You did the correct test, the feeder port is nothing more than a low powered DC port. The signal pin is a Ground being switched.

I can't imagine it's a problem with reef-pi as many use a feeder, Bullseye works good, the controller hardware seems ok, the feeders seem ok, hmmm... Did you check continuity between connector that plugs into controller and feeder PCB on all 3 wires to make sure there's no breaks. Sometimes those green connectors can cut the wire so it's only left with a strand or two holding it in place making a bad connection but it looks ok. Maybe try removing the connector and inspect the wires.

Yeah checked the connector and wires and all good there I checked between ground and signal with continuity mode on my meter on the feeder port and can see when I turn the feeder switch in reef-pi on and off and I see the same results happen when I test my spare reef-pi/robotank board combo

Checked the continuity of all wires to their respective solder joints and no issue there.

Maybe the little circuit boards on the feeders are just crap.

Now if toggling manually the feed button on the Eheim feeder this morning the screen like pulses for I’d say like 10ish seconds and then stops doing that and then sometimes initiates the feeder and sometimes doesn’t.

Very very weird and I’ve given up as it’s driving my blood pressure up lol

Thanks anyways :)

 

[robsworld78]

Yeah checked the connector and wires and all good there I checked between ground and signal with continuity mode on my meter on the feeder port and can see when I turn the feeder switch in reef-pi on and off and I see the same results happen when I test my spare reef-pi/robotank board combo

Checked the continuity of all wires to their respective solder joints and no issue there.

Maybe the little circuit boards on the feeders are just crap.

Now if toggling manually the feed button on the Eheim feeder this morning the screen like pulses for I’d say like 10ish seconds and then stops doing that and then sometimes initiates the feeder and sometimes doesn’t.

Very very weird and I’ve given up as it’s driving my blood pressure up lol

Thanks anyways :)

Sorry to hear that, if you want to give it another go later let me know.

 

[Raf714]

Does anyone know what changed in bookworm to hose the GPIO stuff up? Why oh why would they change that....when I have more time and it gets colder I will try to figure this out and see what we can do to update it...should work on the new OS...

Can this be of any help?: Failed load raspberry pi driver. Error:can't build hal pin 5: strconv.Atoi: parsing "62+rpt": invalid syntax

 

[bishoptf]

Can this be of any help?: Failed load raspberry pi driver. Error:can't build hal pin 5: strconv.Atoi: parsing "62+rpt": invalid syntax

I havent had time to fully dig into it but the new os changed how to access gpio, that whole substructure needs to be re written for the new OS. Its on my to do list to try to figure out the code but I havent gotten to it yet.

Trying to see if I can get @Ranjib to give me some pointers as to what sections of the code need to be updated.

 

[5ft24]

I havent had time to fully dig into it but the new os changed how to access gpio, that whole substructure needs to be re written for the new OS. Its on my to do list to try to figure out the code but I havent gotten to it yet.

Trying to see if I can get @Ranjib to give me some pointers as to what sections of the code need to be updated.

Part of the issue is they completely removed the libGPIO support from the kernel.
You can get GPIO back by downgrading the kernel to 6.1.74 (last version with LIBGPIO)
sudo rpi-update d86b5843d68b9972a5430a6d3da1b271cfc83521 then rebooting
It won't work with the Pi5 though, as the separate I/O chip has a different address internally for the GPIO

 

[Ranjib]

Can this be of any help?: Failed load raspberry pi driver. Error:can't build hal pin 5: strconv.Atoi: parsing "62+rpt": invalid syntax

This is a raspberry pi os issue . The new raspberry pi is ships with newer Linux kernel that has deprecated the old sysfs interface that reef -pi is using . I’m in process of updating the code , right now there’s no way I know of reef-pi can work on the new os. You have to use the older raspberry pi os while I fix the gpio driver issue

 

[Ranjib]

I havent had time to fully dig into it but the new os changed how to access gpio, that whole substructure needs to be re written for the new OS. Its on my to do list to try to figure out the code but I havent gotten to it yet.

Trying to see if I can get @Ranjib to give me some pointers as to what sections of the code need to be updated.

Spot on. There’s lot of changes that needs to be addressed ,, I’m working through it

GitHub - reef-pi/rpi at pi5updates

Physical drivers for Raspberry Pi (I2C, GPIO, PWM) - GitHub - reef-pi/rpi at pi5updates

github.com

If you are interested in the progress

 

[Ranjib]

Part of the issue is they completely removed the libGPIO support from the kernel.
You can get GPIO back by downgrading the kernel to 6.1.74 (last version with LIBGPIO)
sudo rpi-update d86b5843d68b9972a5430a6d3da1b271cfc83521 then rebooting
It won't work with the Pi5 though, as the separate I/O chip has a different address internally for the GPIO

Yes .. I’m in process of updating those things (rewriting a lot of stuff actually ). The challenge I think will be that after the update we won’t be backward compatible , the new reef-pi won’t work with older raspberry os

 

[bishoptf]

This is a raspberry pi os issue . The new raspberry pi is ships with newer Linux kernel that has deprecated the old sysfs interface that reef -pi is using . I’m in process of updating the code , right now there’s no way I know of reef-pi can work on the new os. You have to use the older raspberry pi os while I fix the gpio driver issue

Id love to help, i was looking through the code a couple months ago and found some sub routines that pulled in some gpio stuff but wasn't sure where was best place to start.

Let me know how I can help. In regards to backwards compatibility, yeah will loose that but appears that with the new OS its the only way forward unless we write some middleware based on OS but I think its just easier to say if your using OS x then use this, older OS use y. At some point they will stop supporting the old OS plus I think we should try to support the rpi 5 since Inwould guess that will be the basis for the new architecture going forward.

 

[bishoptf]

Spot on. There’s lot of changes that needs to be addressed ,, I’m working through it

GitHub - reef-pi/rpi at pi5updates

Physical drivers for Raspberry Pi (I2C, GPIO, PWM) - GitHub - reef-pi/rpi at pi5updates

github.com

If you are interested in the progress

Did this move somewhere, was trying to get to the URL and its no longer found...

 

[kboogie]

Did this move somewhere, was trying to get to the URL and its no longer found...

It looks like a deep link issue. If you go up to the project root, https://github.com/reef-pi/, it works.

 

[Ranjib]

Did this move somewhere, was trying to get to the URL and its no longer found...

last night I merged the pi5 support branch. with this we should be able to start testing pi5 and the new raspberry os, but this also means we won't be able to run this version of reef-pi on the older raspberry pi os versions. Depending upon how much there's appetite for new pi5 vs older pi+raspberry pi os , I have to decide if I should back port any fixes and released a 6.1 version or just go with 7 only,

none the less, wanted to keep you folks in the loop that we are slowly working towards the next release,