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

[pdisner]

Man, you guys have been doing awesome work. I’ve been following along, reading about all your current issues that you’ve been working on. What a blast! Right now I’m trying to breed several species of fish/inverts and it’s been taking up all my reserve time and then some. It’s paying off. Two species have spawned and I am having success In raising some. When I join, there are a few features that I’d like to work on with my Pi some having to do with first foods like pods and rotifers and mysids. Here’s my clown gobies that spawned:

Here are my striped Blennies that have spawned:

And baby Bangaii Cardinals:

I’d like to add some feature to this project, but there are some apps that are only used for breeding purposes but can be adapted to other aspects of this whole hobby. Please keep up the good work and hope to participate soon!

 

[pdisner]

I think i have the weekend to work on my Reef-pi Modules.. I Have Relay Circuit working for 3 months now.. I have been working on Long 1-wire for temp sensors.. With the above help i have that Reliable at the distance from my Network closet to my basement sump and my Display tank to my Network closet.. My plan is to put the reef-pi computer remotely away from the damp saltwater . This has been the challenge... Since my Tanks and equipment is so spread out..

off topic..

I have been working with a Investor to expand my Business..(ericaremodel.com our website in progress) I Have found a Reliable Silent partner , So we are in Verbal negotiations and Very close to drawing up Contracts. My only weekend Goal is to Inventory all of my Equipment , This includes Vans Tools and so on. Once i get that to the Business assessment auditor (she has the financials) We will put a Value on what the business is worth.. Then figure a Percentage of ownership for my new partner . His Investment over the next 18 months should let me put on 2-4 new 2to4 people Crews .. This will allow me to turn my Work Van over to a employee and Focus just on Doing estimates and insuring things are Complete to my STANDARDS . I am getting older and with my Bad arm and shoulder i MUST Change how i interact with customers and STOP Physical work.... Anyway .. This is why i have not been as active..
Life stuffs...


Those who are patient have better results in the end..

Have you guys ever played around with the wireless types of sensor components ? I think that the future will be in all wireless peripherals. It would be neat to design a total wireless system, to include each sensor.

 

[Ranjib]

Man, you guys have been doing awesome work. I’ve been following along, reading about all your current issues that you’ve been working on. What a blast! Right now I’m trying to breed several species of fish/inverts and it’s been taking up all my reserve time and then some. It’s paying off. Two species have spawned and I am having success In raising some. When I join, there are a few features that I’d like to work on with my Pi some having to do with first foods like pods and rotifers and mysids. Here’s my clown gobies that spawned:

Here are my striped Blennies that have spawned:

And baby Bangaii Cardinals:

I’d like to add some feature to this project, but there are some apps that are only used for breeding purposes but can be adapted to other aspects of this whole hobby. Please keep up the good work and hope to participate soon!

If you get time, can you list down specific features you think will be useful? I'd love to be aware of those, may be some of them can be implemented easily,
thanks for sharing the breeding photos, they look awesome. I love those clown gobies, they are perfect for small tanks

 

[Erica-Renee]

Have you guys ever played around with the wireless types of sensor components ? I think that the future will be in all wireless peripherals. It would be neat to design a total wireless system, to include each sensor.

I am not sure any wireless capability is reliable enough , When you add water, Well Salt water ,humidity, Lots of other electric components you are asking for a Failure.We have issues from time to time with normal wireless networks (they have been evolving for a decade or longer) And still far from what i call 110% reliable.. This 110% is what a aquarium component needs in my opinion... meaning 95% reliable 15% over head safe guards to insure something will not interfere ,Over run , shut off unexpected without a safe guard in place to insure no loose of livestock..

 

[sjeff35]

The jagged charting .... I had mine doing the same... The Driver (1 wire protocol) has a up and down timing memory for each sensor remembering the Distance from Master to slave for each probe and this required a end probe so to speak..

if they are wired in a star network ( Like i had them in that black break out box) you will get data dropping and multi digit temps .. I Might be explaining this wrong..
Wrong way to wire them

correct wiring.... There must be a master and a slave Master is your gpio pin. Slave is the sensor on the end of the one wire .. The stubs need to be from my experiment 6 inches from the master but appear to not matter distance between them. There wire length needs to be shorter then the master.. If you use long wires.. you MUST Put a 150 resister on each of the Data signal... I have 4 probes wired one is approx 20 feet the others are between 2 feet and 14 feet ..

Again this is from my testing and reading the 1-wire guide ..

https://www.maximintegrated.com/en/app-notes/index.mvp/id/148

Thank you for making my head hurt this early on a Friday...

I have two short runs from the RJ45 jacks to the board where they are wired. This is a very crude diagram of how they are wired.

I'm confused about this whole thing so don't count on me making sense.

 

[pdisner]

That’s what I’m trying To focus on, small species for nano tanks. It’s the future of the hobby. Less long term expense and effort. Yes definately will chime very soon.

 

[sjeff35]

I deleted my tank temp config and recreated it and now no more jagged lines...

hmmmm

 

[pdisner]

I am not sure any wireless capability is reliable enough , When you add water, Well Salt water ,humidity, Lots of other electric components you are asking for a Failure.We have issues from time to time with normal wireless networks (they have been evolving for a decade or longer) And still far from what i call 110% reliable.. This 110% is what a aquarium component needs in my opinion... meaning 95% reliable 15% over head safe guards to insure something will not interfere ,Over run , shut off unexpected without a safe guard in place to insure no loose of livestock..

Well, this is one of my ideas, though I’m sure you guys have discussed somewhere here. Haven’t seen yet. But if someone has a 3D printer, a nice watertight and corrosion resistant case can be designed to protect components. I’m sure that there’s something that can be done to prevent corrosion, don’t you? And what problems exactly, regarding WiFi and it not being reliable enough? What have you played around with as far as wireless sensors? I’ve just skimmed web pages recently. I’ve had to put my reef pi in hold for the moment.

 

[Erica-Renee]

Have you guys ever played around with the wireless types of sensor components ? I think that the future will be in all wireless peripherals. It would be neat to design a total wireless system, to include each sensor.

I am not sure any wireless capability is reliable enough , When you add water, Well Salt water ,humidity, Lots of other electric components you are asking for a Failure.We have issues from time to time with normal wireless networks (they have been evolving for a decade or longer) And still far from what i call 110% reliable.. This 110% is what a aquarium component needs in my opinion... meaning 95% reliable 15% over head safe guards to insure something will not interfear

Thank you for making my head hurt this early on a Friday...

I have two short runs from the RJ45 jacks to the board where they are wired. This is a very crude diagram of how they are wired.

I'm confused about this whole thing so don't count on me making sense.

That is actually correct .. but.. the distance from gpio to the first sensor needs to be more then 5 inches (why this matters ,Charge and discharge , ON A Very short length with more then one sensor. there is still current (charge) in the wire before the second discharge signal is read . ) Also the distance between sensor 1 the sensor on the stub.... and the slave sensor. (sensor on the end of wire) needs to be twice as long as the distance between gpio and sensor one.. so if the first sensor is 6 inches from the reister the slave sensor needs to be 12 inches.. I Found this info from a biologist that studies caves actually. He did a very detailed and Technical write up about a 300 meter 1wire sting of temp sensors each sensor 6 feet apart with a 150 resister between each stub sensor and his slave sensor on the end was 12 feet from the last stub....

With the above said it may not be explained well. It might be partly wrong or not completely required for very short runs.. Also the slower data collection rate does help... for very long runs above 100m you must move up to 5v and different resisters... The Geeky Blonde chic who is prob wrong...

 

[Erica-Renee]

I deleted my tank temp config and recreated it and now no more jagged lines...

hmmmm

That is Great.. Leave it running for 2-3 days and see if it still works and again my SUPER LONG Overly complicated post may not even apply to the length of your 1-wire setup....I Do know when i had mine wired in a star topology ,I could delete all sensors and they would work fine for a hour or even a day then stop dropping ,Then for no apparent reason start working correctly again..

as i said my build is different then most because of the distances i am trying to achieve..

 

[sjeff35]

Let me think about it....

Thanks

Well, now that I have a Kessil 360 on the way (Thanks @philshel )

I need to get busy and wire up a control for it!

This is a never ending project... ;Woot

 

[philshel]

Well, now that I have a Kessil 360 on the way (Thanks @philshel )

I need to get busy and wire up a control for it!

This is a never ending project... ;Woot

Happy to complicate your life!

 

[HM3105]

This is cool
My bare steel probes are still working smooth, no rust yet.

I was listening to this pod cast with the guy from Printed Reef Solutions he made an interesting point. The magnets he uses aren't completely encased so they aren't water proof. He thinks this is a good thing because if it starts to rust you can see it and take action. Whereas if it is completely incased you can't and if it ruptures you release all the contents of the enclosure holding the magnet into the system vs. hopefully just a bit. Just a something I heard.

The jagged charting .... I had mine doing the same... The Driver (1 wire protocol) has a up and down timing memory for each sensor remembering the Distance from Master to slave for each probe and this required a end probe so to speak..

if they are wired in a star network ( Like i had them in that black break out box) you will get data dropping and multi digit temps .. I Might be explaining this wrong..
Wrong way to wire them

correct wiring.... There must be a master and a slave Master is your gpio pin. Slave is the sensor on the end of the one wire .. The stubs need to be from my experiment 6 inches from the master but appear to not matter distance between them. There wire length needs to be shorter then the master.. If you use long wires.. you MUST Put a 150 resister on each of the Data signal... I have 4 probes wired one is approx 20 feet the others are between 2 feet and 14 feet ..

Again this is from my testing and reading the 1-wire guide ..

https://www.maximintegrated.com/en/app-notes/index.mvp/id/148

Mine is wired using the second picture and I haven't seen those data drops.

 

[sjeff35]

Dang, just got started and already stuck...

Can someone take a look at this picture and see if they can spot anything wrong?

If I followed the wiring correctly and my thought process is correct, the voltage at the 3.5 mm jack should change as I move the sliders??? Correct??? I don't see it change, I just get 12v across the ground and both sides.

It looks like the Reef-pi part of this is working...

 

[Ranjib]

I am not sure any wireless capability is reliable enough , When you add water, Well Salt water ,humidity, Lots of other electric components you are asking for a Failure.We have issues from time to time with normal wireless networks (they have been evolving for a decade or longer) And still far from what i call 110% reliable.. This 110% is what a aquarium component needs in my opinion... meaning 95% reliable 15% over head safe guards to insure something will not interfear


That is actually correct .. but.. the distance from gpio to the first sensor needs to be more then 5 inches (why this matters ,Charge and discharge , ON A Very short length with more then one sensor. there is still current (charge) in the wire before the second discharge signal is read . ) Also the distance between sensor 1 the sensor on the stub.... and the slave sensor. (sensor on the end of wire) needs to be twice as long as the distance between gpio and sensor one.. so if the first sensor is 6 inches from the reister the slave sensor needs to be 12 inches.. I Found this info from a biologist that studies caves actually. He did a very detailed and Technical write up about a 300 meter 1wire sting of temp sensors each sensor 6 feet apart with a 150 resister between each stub sensor and his slave sensor on the end was 12 feet from the last stub....

With the above said it may not be explained well. It might be partly wrong or not completely required for very short runs.. Also the slower data collection rate does help... for very long runs above 100m you must move up to 5v and different resisters... The Geeky Blonde chic who is prob wrong...

I read the exact blog post. I think now collectively we also know a whole lot about the ds18b20 , deployed on duty.

 

[Ranjib]

Dang, just got started and already stuck...

Can someone take a look at this picture and see if they can spot anything wrong?

If I followed the wiring correctly and my thought process is correct, the voltage at the 3.5 mm jack should change as I move the sliders??? Correct??? I don't see it change, I just get 12v across the ground and both sides.

It looks like the Reef-pi part of this is working...

The wiring looks ok, but its hard to properly see their connection.
Can you check the voltage where pca9685 pwm is connected to the resistor, is that correct (for example 45% should give around 2.3V). Test the transistor circuit only with 1 channel first. Can you share the resistor values as well?

 

[sjeff35]

The wiring looks ok, but its hard to properly see their connection.
Can you check the voltage where pca9685 pwm is connected to the resistor, is that correct (for example 45% should give around 2.3V). Test the transistor circuit only with 1 channel first. Can you share the resistor values as well?

Thanks @Ranjib

The resistors are 1k and 10k , I just followed the guide here.

I took a better pic from directly above the board. I have 12 volts coming into the rail. When I measure the voltage where the resistors meet the PCA board, I get nothing.

 

[Erica-Renee]

I read the exact blog post. I think now collectively we also know a whole lot about the ds18b20 , deployed on duty.

Yes I think this can be considered solved , The 1-wire circuit can in fact be used in long distances . I am going to work on Modular circuit board , Revisit my Relay wiring . Then Move to the Dosing Circuit (which is simple) Ohh and rebuilt my ATO Circuit . It works flawless but it looks kinda clunky ...

I am changing from using Jacks for the ato and temp to using screw terminals. I think this will make it easier for when running the actual cables. to proper length. The down side is they take up more space inside a housing..

 

[janos]

Hi Erica if you have some Circuit from dosing please show that.Used your circuit for the relay board and working perfect.

 

[Ranjib]

Thanks @Ranjib

The resistors are 1k and 10k , I just followed the guide here.

I took a better pic from directly above the board. I have 12 volts coming into the rail. When I measure the voltage where the resistors meet the PCA board, I get nothing.

Circuit looks ok to me. So, if you detach the transistor part and just check the pca9695 output, its not generating pwm ? If thats the case then we have problem with the pca9685 board itself