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

[Ranjib]

Hoping someone will chime in on ATO alerts. I created a low ATO level float that lights an LED when the water level is getting low in the ATO container. That works but would like to get it to alert via email when triggered. I've tried lots of things, I think it would be nice to have it record when it is triggered, and I can enable control equipment and then have nothing for control pump. If I do that then it just keeps counting up and up unless I set the disable on alert after 'X" seconds. That also works but I then have to remember to go and reset it when I have added water, I know I am lazy. I am looking for something that just records on and off and will reset when water is added etc.

Maybe ATO is not the best function to use, if anyone has any additional thoughts let me know, thanks.

Update this may be doable via the new ph-analog input route but trying to understand how that works, looking thought the release notes its touched on, if there is any additional documentation on the ph-analog inputs let me know.

Thanks

This should work. Can you share your ato settings?
There was a bug that cause alert to not fire if it’s not during the first hour. I have fixed it and it will be available in next immediate release. I think you are hitting that. Check if you do get alert if it’s in the first hour , that will confirm you are hitting that bug,

 

[narkotix]

reef-pi 3.3 onwards we have file driver which can be used to prototype /test any new driver. Just write your sensor value into a file and reef-pi will read that file and treat the value as analog input which can be hooked up to a ph module for controlling equipment etc. for example one can use circuit python based adafruit gas sensor code to maintain a file and write the sensor value there and then configure reef-pi with that file. This will allow controlling a air purifier with gas sensor data using reef-pi

Hi @Ranjib im experimenting with this feature but I don't quite understand how it works. In this case I`m using it for a PH probe that isn't supported (ads1115 instead of ads1119) but I`m able to output the analog value to a text file. The file is a single line containing a rounded up version of the PH with 2 decimals and nothing more. Each time I measure the PH, i overwrite that text file.

Beyond that - I've set up an analog file driver to read that file (my python script outputs the value to /home/pi/ph.txt) and then i've added an analog input under connectors and then finally used that analog input for the PH page. Am i doing this right?

*edit* i should confirm that the PH probe is functional using my python script and is able to measure different values with my test solutions.

 

[Ranjib]

Hi @Ranjib im experimenting with this feature but I don't quite understand how it works. In this case I`m using it for a PH probe that isn't supported (ads1115 instead of ads1119) but I`m able to output the analog value to a text file. The file is a single line containing a rounded up version of the PH with 2 decimals and nothing more. Each time I measure the PH, i overwrite that text file.

Beyond that - I've set up an analog file driver to read that file (my python script outputs the value to /home/pi/ph.txt) and then i've added an analog input under connectors and then finally used that analog input for the PH page. Am i doing this right?

*edit* i should confirm that the PH probe is functional using my python script and is able to measure different values with my test solutions.

yup. i think you figured it out already.

 

[narkotix]

yup. i think you figured it out already.

Fantastic! thank you and yes i just did! Just working on integrating it as a bootscript (fingers crossed the ramdrive that i created reinitialises and the script can write to it).

Also some progress on my ammonia detector.

So I think i've reverse engineered the seneye ammonia/ph probe. Basically those strips that you need to purchase on an ongoing basis are nothing more than what an Seachem ammonia and ph alert badge is. So i've purchased a TCS34725 rgb sensor (which is i2c) and the idea is to have the ammonia badge submerged in the sump (or a visible area inside a tank) with a black box to enclose it and stop refugium light hitting it but let water flow thru alongside the TCS34725 outside the tank mounted as well but the sensor of the TCS34725 over the ammonia alert sensor. Basically i`m interpreting the colours on the ammonia sensor and then mapping those colours to an ammonia value.

Apologies if this doesnt make sense but I just got the sensor hooked up today but I need to experiment on how this sensor is supposed to work because i am getting inconsistent values against demo code. I suspect my focal point is not correct so the sensor is detecting light from other sources.

 

[Zippyfear]

This was a bug that I fixed recently. It will be fixed in new release, please help us testing it

I will be testing this fix as well, as my ATO alert/stop is critical given it's pumping in from a 75 Gal holding tank. lol. Thanks!

 

[robsworld78]

Fantastic! thank you and yes i just did! Just working on integrating it as a bootscript (fingers crossed the ramdrive that i created reinitialises and the script can write to it).

Also some progress on my ammonia detector.

So I think i've reverse engineered the seneye ammonia/ph probe. Basically those strips that you need to purchase on an ongoing basis are nothing more than what an Seachem ammonia and ph alert badge is. So i've purchased a TCS34725 rgb sensor (which is i2c) and the idea is to have the ammonia badge submerged in the sump (or a visible area inside a tank) with a black box to enclose it and stop refugium light hitting it but let water flow thru alongside the TCS34725 outside the tank mounted as well but the sensor of the TCS34725 over the ammonia alert sensor. Basically i`m interpreting the colours on the ammonia sensor and then mapping those colours to an ammonia value.

Apologies if this doesnt make sense but I just got the sensor hooked up today but I need to experiment on how this sensor is supposed to work because i am getting inconsistent values against demo code. I suspect my focal point is not correct so the sensor is detecting light from other sources.

A few years ago someone on a forum, can't remember who right now, was attempting basically the same thing. He came very close but think other things came up and he couldn't continue so all the best. I think a big problem was the rgb sensor wasn't sensitive enough and something with an IR filter. I see this has that, would be cool if you can get it, if you need some testers let me know.

 

[narkotix]

A few years ago someone on a forum, can't remember who right now, was attempting basically the same thing. He came very close but think other things came up and he couldn't continue so all the best. I think a big problem was the rgb sensor wasn't sensitive enough and something with an IR filter. I see this has that, would be cool if you can get it, if you need some testers let me know.

Thats great to know - if this fails, I was going to take a more exhaustive approach using a small camera (connected via usb or raspi) and then use something like opencv to detect the sensor state. its more heavy duty but it could also work (ableit the difficulty of squeezing in a camera somehow). I've built a facial recognition product using opencv and a movidius neural compute stick on a raspi and the jetson platform so im up for the challenge. Will let you know how i go with testing but commit to releasing this as open source. I may need someone however to translate my horrible python code to something better integrated with reef-pi.


PS for those with the ADS1115 and PH4502c ph sensor, hit me up and ill give you the code I have plus instructions to run it as a rc.local job (will move it to init.d) and have the file writing to ramdisk rather than the sd card. Its a hack job compared to the fine work the others have done but in trying times (and the 4 hours i spent messing around) where getting parts sent from overseas might be a challenge, this does the job until ADS1115 is supported as part of ph-probe.

 

[mr.salty]

Sorry if I haven't read the original post properly or things have progressed but how many outlets or channels etc can you have with reef pi? Is there a set limit or is it expandable? I'm making two 8 channel ac outlet powerbars and then want temp, pH, orp.

I also want the ability to control solenoid valves etc for a liquid feeding rig so can all this be done or will I need multiple raspberry pi and reef pi?

 

[GaryE]

only limited by how many ULN2803 chips on board and how many GPIO pins you have available to control those chips.

the board I'm developing has 1 2803 but 10 unused GPIO pins. I could easily add another 2803 for 16 total outlets controlled by relays. My setup doesn't really need that.

 

[BenB]

@Ranjib sorry for the amount of questions ;Happy There's a software limit in the maximum number of Outlets and Jacks? Because I'm designing a power bar using I2C bus. I'm using a PCA9685 extender do drive 8 AC plug and 8 DC mosfet output to drive fans, led strip, ecc In the future I'll add a second PCA9685 on a separate board to only drive the led drivers... Thank you!

No problem, I have 2 PCA9685s on different addresses (decimal 80 and 83) and they work fine together (do not forget to solder the hex address on the PCA9685 itself), I am planning to add another 2 PCA9685s for lights later on.

 

[Ranjib]

Hive mind,
I have released reef-pi 3.4. Following is the summary:

• Introduce sht31d humidity and temperature sensor support
• openapi documentation
• journal subsystem api and wip UI

Bugfixes:

• Fix ato usage alert and disable on alert behavior
• Display control errors out due to new line character
• Dont break connector ui when associated driver is missing
• Fix camera ui when no images are present
• pH calibration ui bug in firefox browser

Chores

• address all react 16 lifecycle method deprecation warning

release builds can be found in usual location: https://github.com/reef-pi/reef-pi/releases/tag/3.4

This is a non-breaking minor release. All 3.x users should be able to upgrade without any issues.

Again, this is minor release with sht31d humidity sensor support and few critical bug fixes (ato disable on alert and broken connector ui when associated driver is deleted etc). Please try it out if you are encountering any of those bugs, or want to use the new humidity sensor. I have also added a new subsystem called "journal" for manual testing like workflow where users want to record specific values on-demand, manually. At this time its API ready with minimal UI support. i.e users can only create/update/delete the journal parameters (e.g. alk test results) but not enter their values via UI (API is done, UI still pending). We intend to wrap up the UI in next release.


stay safe
ranjib

 

[Ranjib]

Sorry if I haven't read the original post properly or things have progressed but how many outlets or channels etc can you have with reef pi? Is there a set limit or is it expandable? I'm making two 8 channel ac outlet powerbars and then want temp, pH, orp.

I also want the ability to control solenoid valves etc for a liquid feeding rig so can all this be done or will I need multiple raspberry pi and reef pi?

Theres no software limit from reef-pi side. You are limited by rpi GPIO or your circuit. Since 3.0 we also support smart plugs and powerstrip, which is not limited by circuit. I am pretty sure if we keep on adding new drivers at some point the system will bog down, but i have not tested it that far. I have used three hs300 strips at the same time at max , which is 18 outlets. I think it will be just fine with dozens of such power strips.

 

[robsworld78]

Hive mind,
I have released reef-pi 3.4. Following is the summary:

• Introduce sht31d humidity and temperature sensor support
• openapi documentation
• journal subsystem api and wip UI

Bugfixes:

• Fix ato usage alert and disable on alert behavior
• Display control errors out due to new line character
• Dont break connector ui when associated driver is missing
• Fix camera ui when no images are present
• pH calibration ui bug in firefox browser

Chores

• address all react 16 lifecycle method deprecation warning

release builds can be found in usual location: https://github.com/reef-pi/reef-pi/releases/tag/3.4

This is a non-breaking minor release. All 3.x users should be able to upgrade without any issues.

Again, this is minor release with sht31d humidity sensor support and few critical bug fixes (ato disable on alert and broken connector ui when associated driver is deleted etc). Please try it out if you are encountering any of those bugs, or want to use the new humidity sensor. I have also added a new subsystem called "journal" for manual testing like workflow where users want to record specific values on-demand, manually. At this time its API ready with minimal UI support. i.e users can only create/update/delete the journal parameters (e.g. alk test results) but not enter their values via UI (API is done, UI still pending). We intend to wrap up the UI in next release.


stay safe
ranjib

Much appreciated and great job @Ranjib and @Michael Lane! I love the fact you haven't lost the passion to keep doing this.

I'm working on a large system that will push reef-pi where it's probably never been. This is basically the goal.

64 outlets
12 pH circuits
12 temp sensors
16 dosing pumps
16 stirrers
10 ATO's
at least 16 PWM channels
12 flow meters when it can

Haven't settled on how I'm going to tackle it yet but will be fun.

 

[Simonv92]

@Ranjib, Testing the new version 3.4 looking good so far Thanks also to @Michael Lane!
Under the driver tab, what is "dli-wpsp" driver?
Thank you again!

 

[Simonv92]

Today I had a strange problem, maybe a quif fix problem. I have a quite old iPad with iOS 9.3.6 which works fine. When I try to reach the reef-pi dashboard with my local raspberry pi address it shows a white page with the reef-pi label at the top of the page (in the name of the tab) on all of my other device works fine. Do you know what could it be? It's an iPad related problem for sure, maybe someone has already encountered this problem..
Thanks all!

 

[AbjectMaelstroM]

I think there was someone on here with a similar issue, but it was a while ago. Not sure if there was a fix.

I don't have Apple devices, but it may be an outdated browser issue.

 

[b4tn]

Yup, I actually tried again the other day just because I am stuck at home anyway and got the same thing on my older ipad. You cant try any of the 3rd party browsers because they only support ios 10 and above. It kind of sucks because the ipad is not used for anything else and would be perfect to set up next to my reef. But something in the reefpi web browser is not supported. I think I did a quick try jailbreaking it with no luck also but I may try again one day.

 

[ChrisNH]

I have a pro-tip. If you ever need an aquarium heater quick, or to cauterize a wound, just reverse the data and the +V line on your temp sensor.

I got confused by the image on the adafruit build guide and thought it had the ground up instead of being at the bottom when orienting the power and data lines. Good times.

 

[Simonv92]

Yup, I actually tried again the other day just because I am stuck at home anyway and got the same thing on my older ipad. You cant try any of the 3rd party browsers because they only support ios 10 and above. It kind of sucks because the ipad is not used for anything else and would be perfect to set up next to my reef. But something in the reefpi web browser is not supported. I think I did a quick try jailbreaking it with no luck also but I may try again one day.

Thank you very much! I was also thinking about the jailbreak, but I don't know if that can fix.. I was able to try Firefox but I had the same issue... We have to find a way to install iOS10

 

[b4tn]

Thank you very much! I was also thinking about the jailbreak, but I don't know if that can fix.. I was able to try Firefox but I had the same issue... We have to find a way to install iOS10

If you happen to find a way please tag me lol. I am trying to jailbreak mine right now with not much success so far.