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

[MaccaPopEye]

Its strange that tech things are so costly in australia. Is it because of lack of maufacturing facility? or some weird taxing system ?

I can only really guess but I would say it's a combination of a lot of things. We call it the "Australia tax", but I don't believe it is a government issue.

Goods and services are taxed at a standard 10% nationwide, so not extreme. That tax is also included in the advertised price (a difference that took me by surprise when I visited the states haha). We also generally rely on imports for tech products, not much is manufactured in Australia (as far as I know).

I think the main thing is Aus has pretty strict standards for tech products and everything sold in Aus has to be certified. So Aus shops can really only sell brand name products with high quality control that have gone through the effort to prove they meet the standards. These brand name products are more expensive to get than no-name ones, then throw in the exchange rate and reliable / fast shipping and the shops profit margin and the price goes up fast.

Brick and mortar stores are also expensive to run in Aus (or so I have been told) so they have to put a large mark up on things in order to keep their doors open, then the cost of living is also fairly high so they need to be able to make a decent living themselves.

The downside to that is it turns a lot of people (like me) online to buy uncertified no-name brand components from china and it becomes a matter of get it now and pay massive mark-up or get it from china and wait 1-2 months. There is also no import tax for purchases under $1000.

One example of the price difference is the RJ-45 keystone sockets I recently ordered. The cheapest I could find them for locally was $3 PER socket (some place they were $7 each). Or I could get a packet of 25 from china for $25. It sucks waiting and one or two of the sockets might not work, but to me that’s better than paying 3-7x the price (and usually the QC of products I get from china is pretty dang good).

However that mostly only applies to no-name tech components if you can find them.

Importing brand name tech items from the US or Europe is very expensive and usually I will be able to find them cheaper in Aus (if I can find them). The main cause of that seems to be the exchange rate and (reliable) international shipping. A good example is the Raspberry Pi 3, I looked all over online and once you factor in international shipping and exchange rates it was the same price/cheaper to buy one from an Aus shop (for $50 AUD). But then the range in Aus shops is very limited, I couldn't find any of those perma proto boards locally (only the normal boards).

I ended up getting 2x half size perma proto boards (one with a space for a raspberry pi and one normal) and a 40-pin female GPIO header to solder onto the raspberry board from Amazon US. After international shipping + exchange rate it cost me $40 AUD. In comparison Amazon Aus (or any other Aus shops I could find) didn't even have the same products, but the adafruit perma proto boards they did have were $50 for a pack of 3 small ones..... ebay does have them for a reasonable price (around $5 AUD each) but international shipping is $30-40 AUD.

At the end of the day things are quite expensive here, but our wages are also quite high to match. Both my wife and I have good full time jobs and we are able to live comfortably enough, and I still usually have enough on the side to spend irresponsibly on my hobbies :p

 

[Ranjib]

In the above board.. you have one gpio pin for 3 temp sensors. Or am i missing something.. Also i see a ic chip at the bottom of the board... Am i missing something ?

I am about ready to start assembling mine Renovating the Basement sump/fish room all new electrical plumbing and so on . Want to add this to process ...

You are right about both. DS18b20 sensor is a 1-wire protocol based sensor, which is like i2c, and multiple devices can be connected. I am not aware of any theoretical limits in the protocol, but the way linux kernel is implemented it is known to work with as many as 20, and the limitations arise from physical setups (wire length), its is known to work easily with as many as 7 sensors all tied to the same one wire bus. So, yes all sensors connected to GPIO4 or the one wire bus (configurable by devicetree/dtoverlay)

The IC on the bottom of the perma proto board is an eeprom. It is not used here at all. Adafruit perma proto board are compliant to raspberry pi hat specification which involves using an eeprom to autodetect things. If we load custom device code, we can use the eeprom to design a setup that Pi will automatically load on boot.

 

[cheeseclock]

I should be ready to test it tomorrow.

 

[Ranjib]

pH monitor entered operation yesterday night. The coding work and initial prototype testing was done for almost a month or two before. I had to punt on the actual deployment due to missing probe holder in my initial order. Then I got busy with 2.0 features. Anyway..

I was very happy to see the calibration features work smoothly, no major issues. I will update the UI to make things more intuitive, but I am over all happy. Will post the graphs soon, its very interesting to see the graph as I perform the calibration process.

This is also the smallest reef-pi build :)

Calibration ..

 

[cheeseclock]

pH monitor entered operation yesterday night. The coding work and initial prototype testing was done for almost a month or two before. I had to punt on the actual deployment due to missing probe holder in my initial order. Then I got busy with 2.0 features. Anyway..

I was very happy to see the calibration features work smoothly, no major issues. I will update the UI to make things more intuitive, but I am over all happy. Will post the graphs soon, its very interesting to see the graph as I perform the calibration process.

This is also the smallest reef-pi build :)

Nice rug!

Do reef people use pH for dosing or is it just to have the data? I ask because with a planted tank, you could potentially meter CO2 based on pH.

 

[Ranjib]

Nice rug!

Do reef people use pH for dosing or is it just to have the data? I ask because with a planted tank, you could potentially meter CO2 based on pH.

Dosing.
Two part or kalkwasser dosing is sometime controlled by ph feedbacks. I am learning sps husbandry , which requires some of these. Once 2.0 is out , I might focus on some ph driven dosing system

 

[cheeseclock]

They all light. I was confused why I had no control over the brightness. It appears that the LDD drivers vary the cathode voltage as opposed to the anode. I'll have to adjust the wiring accordingly.

 

[Ranjib]

Too many build pics ,

Here is my bubble coral, caught off guard at night

Micro goniopora at night with sweeper tentacles out

Rainbow acan

 

[cheeseclock]

Those are gorgeous! You're making me want to try saltwater.

 

[Ranjib]

Here are my learnings during the ph monitor build calibration process.
This is the ph chart for the whole process,

- Out of the box , the ph probe is fairly accurate. The storage solution itself has ph 3. This is the very first phase in the chart. Rest three phases (up, down and up) represent three calibration phases.
- Atlas Scientific ph kit ships with three ph calibration solution. Mid (7), Low(4) and High(10)
- Without any calibration the probe reads 6.9 for mid, 4.1 for low and 9.7 for high. Here are some screen shots with the exact values.

- I ran the calibration using reef-pi ui. In order of progression: Mid(7),Low(4) and High (10)

I have also manually verified that the calibration was done indeed, using atlasscintific example code

Over all, it was a fairly straight forward thing for me. I am planning to create a build guide for this. Probably after Makerfaire,
Finally, this is current trend of my 29G biocube's pH

 

[Ranjib]

Got some time today to take a few macro shots for my tanks.

Bubble coral (earlier photo was at night)., fully extended during daytime

Zoa garden is out of real estate

One of the pico tanks

Ricordia

 

[Cary]

Ranjib, Good morning. I may have missed this in the previous 200+ pages, if so I am sorry. Did the two customers (Public Aquariums?) get their systems set up and if so what do they think of Reef-pi?

Also, any pictures on them?

 

[Ranjib]

Ranjib, Good morning. I may have missed this in the previous 200+ pages, if so I am sorry. Did the two customers (Public Aquariums?) get their systems set up and if so what do they think of Reef-pi?

Also, any pictures on them?

I wont call them customers, they are just users (since I have not sold anything to them :-)). One of them is using it to research aging process in killi fish (??), where reef-pi is hooked up with cmu cams and it can quantify the movements of fish. He is running it on-demand (the fish lab is far from his actual work desk, and there are many fish that goes through the experimentation in parallel), this is a local university. The other setup is temperature monitors. They have not complained about anything yet. Though I think the public aquarium setup could benefit a lot from the 2.0 release. I'll check with them for pics after the makerfaire is over

 

[Cary]

thanks for correcting me. Look forward to the pics.

 

[Cary]

BTW, If anyone is looking for a rock flower anemone, @cultivatedreef (an R2R sponsor) has a sale going on. I am not affiliated with them and don't receive anything. I have purchased several zoas\palys and a beautiful Cyphastrea and they were all packaged and delivered promptly.

 

[Ranjib]

Today I deployed latest reef-pi (master) to all of my 5 reef-pi controllers. All are up :) . I was very happy to finally customize the dashboard of each of the controllers. This was probably the widest hardware test I did in a single run, throughout reef-pi development. Here is a rundown of these controllers,

1. The power controller (adj srp 8 power strip). Dashboard is configured to show only equipment and cpu/memory utilization. Dashboard

Equipment

Connectors

Dashboard configuration

2. The kessil controller, which controls day light cycle of all of my three tanks. Runs on pca9685 and couple of lm2596 (12 -> 5v and 12->10v)
Dashboard

Dashboard configuration

3. The multi temperature monitor , no extra hardware other than pi and minimal circuit required to support temp sensor (1 4.7k resistor). It has three connectors for three sensors, currently only two of them are in use. Dashboard:

dashboard settings:

4. ph Monitor. Uses Pi + atlas scientific ezo circuit. The smallest form factor build. Dashboard

Dashboard settings

5. The all in one controller (sort of :) ). Runs pi + one temp sensor + one ato sensor + kessil controller based on pi pwm, not pca9685.

Dashboard configuration

Very happy with the outcome. I'll take down aio controller and revamp it to add few things (uln2003 and ph circuit) to support power controller and ph monitoring capabilities.

 

[MaccaPopEye]

Today I deployed latest reef-pi (master) to all of my 5 reef-pi controllers. All are up :) . I was very happy to finally customize the dashboard of each of the controllers. This was probably the widest hardware test I did in a single run, throughout reef-pi development. Here is a rundown of these controllers,

1. The power controller (adj srp 8 power strip). Dashboard is configured to show only equipment and cpu/memory utilization. Dashboard

Equipment

Connectors

Dashboard configuration

2. The kessil controller, which controls day light cycle of all of my three tanks. Runs on pca9685 and couple of lm2596 (12 -> 5v and 12->10v)
Dashboard

Dashboard configuration

3. The multi temperature monitor , no extra hardware other than pi and minimal circuit required to support temp sensor (1 4.7k resistor). It has three connectors for three sensors, currently only two of them are in use. Dashboard:

dashboard settings:

4. ph Monitor. Uses Pi + atlas scientific ezo circuit. The smallest form factor build. Dashboard

Dashboard settings

5. The all in one controller (sort of :) ). Runs pi + one temp sensor + one ato sensor + kessil controller based on pi pwm, not pca9685.

Dashboard configuration

Very happy with the outcome. I'll take down aio controller and revamp it to add few things (uln2003 and ph circuit) to support power controller and ph monitoring capabilities.

That's looking amazing mate!

Have you ever considered trying to combine everything from all of your 5 systems into 1?

Not really for any reason other than because it would be a cool build to see. I know it' probably easier prototyping with individual controllers though.

 

[Ranjib]

That's looking amazing mate!

Have you ever considered trying to combine everything from all of your 5 systems into 1?

Not really for any reason other than because it would be a cool build to see. I know it' probably easier prototyping with individual controllers though.

Yes. The 4th build, which is AIO will go through some changes, definitely a power strip integration, and may be ph monitoring as well.
It will be hard to combine everything though, because reef-pi is going through heavy development (and this is likely to be the case for quiet some time.. dosing, wavemaker is still in prototype phase), so there are some things thats always a work in progress.. but definitely two all in controllers are in my mind, each of them will completely control one pico tank., and have temperature, ato, kessil and power controller capabilities (reef-pi 1.0 specs).

 

[Ranjib]

Also to reduce the number of controllers I have overall :0) . Currently this is beneficial as this setup allows me to test things one at a time, while i have multiple backups. It also ensures that one of the original promise of reef-pi, that is modularity is preserved (i.e. each sub-system can be run independently without needing anything else).

 

[Ranjib]