Robo-Tank DIY Aquarium Controller Forum
Robo-Tank 2.0 - Updated Controller - Printable Version

+- Robo-Tank DIY Aquarium Controller Forum (https://www.robo-tank.ca/forum)
+-- Forum: Robo-Tank (https://www.robo-tank.ca/forum/forum-1.html)
+--- Forum: Current News and Design Ideas (https://www.robo-tank.ca/forum/forum-8.html)
+--- Thread: Robo-Tank 2.0 - Updated Controller (/thread-1239.html)



Robo-Tank 2.0 - Updated Controller - Rob F - 09-06-2021

Hi, I've just finished a hardware update for the controller, I've been selling this one for nearly a year and half now so figured it was time to make some changes based on feedback and what I've seen. 

Aside from the changes I was able to add extra filtering making the board less susceptible to noise and ripple from external DC equipment. Now when DC pumps or lights that draw an amp or more turn on the power doesn't drop momentarily and the +5v rail remains clean. I suspect this is what originally was causing the pH circuit to corrupt and cause a possible error on the DS18B20 sensor. Here's a video showing some measurements on the bench.



Physically the controller is the same size but ports have been changed, here's the list.
  • PCB is now 4 layers instead of 2. This provides a nice ground and power plan and helped reduce noise overall.
  • Controller can now be powered using a 24v DC power supply, this is nice if you want to plug 24v accessories in the DC ports.
  • Added a real time clock, RTC, so the controller isn't reliant on a network, with an RTC you can run it off line and overall makes a more reliable system as controller is in trouble if no time is available.
  • Replaced the USB connector for the sensor extension with an Ethernet RJ45 connector, this was done due to some USB cables don't work.
  • Added three 3.5mm phone connectors, these are connected to 6 PWM/Analog outputs, 2 per plug, and can be used for Kessil lights, Jabeo pumps and any other equipment that has this connector which is quite a bit, this makes it plug and play. The two 4pin plugs that are used today are still on the controller, these are just extensions.
  • Added 2nd 1-wire bus for DS18B20 temp sensors, this means temperature port 1 is isolated from temperature port 2 and 3, they now work independently 
  • Added extra jumper to the water sensor ports so they can be used for DS18B20 and DHT22 temp sensors as well. This means it'll be possible to run up to eight 1-wire buses for DS18B20's, no more needing to share a bus for multiple sensors, this will greatly improve DS18B20 reliability. 
  • Added 2 backup ports for 2 AC outlets, connect a float switch or regular switch to bypass the controller and control the AC outlet. These have the same function as the 3 backup ports for DC ports.
  • Removed one DC port due to no space on board, now there's a total of 6 + the feeder port.
  • Added speed control to Raspberry Pi cooling fan port, allows fan to be slowed down so it runs quieter.
  • Replaced the DIP switches for PWM/Analog ports to pluggable jumpers. Unfortunately this isn't an improvement, I prefer dip switches but it wasn't possible to use them due to size and automated mounting constraints. 
Here's some pictures of the prototype I have, on the 3rd and 4th picture you can see a location to plug in a DS3231 Real Time Clock, I've now decided to install a DS1307 on the board permanently, instead there will be a small CR1220 battery in that location to retain the clock when controller isn't powered up. Any battery will last over 5 years, some up to 10. With this it's possible to run the controller without a network connection and increases reliability.

A 20mm x 20mm cage fan will still fit in the same location and where the Pi plugs in there's no longer a bunch of header leads so I believe a low profile heat sink will fit on the Pi without an issue. I'm not 100% sure just yet as I don't have a heat sink but I'm going to make something the same size and verify but it looks like it'll be ok.

The extra expansion and extensions I have available will still work with this board however all the pin assignments for ports have changed for them and the controller.

Unfortunately I do need to raise the price $10-$15 due to 4 layer PCB and some extra components. Due to that I'm going to add a basic version of the controller, it'll probably be around $75 assembled. I expect to have these available in about a month.

If anyone has any questions or suggestions let me know.

[Image: CAM02429-optimized.jpg]

[Image: CAM02435-optimized.jpg]

[Image: CAM02440-optimized.jpg]

[Image: CAM02441-optimized.jpg]


RE: Robo-Tank 2.0 - Updated Controller - JFReyes - 09-06-2021

Great Rob!

Please document all these changes so I can adjust reef-pi accordingly.

Thanks,


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 09-06-2021

Thanks Jose, I will definitely do that.


RE: Robo-Tank 2.0 - Updated Controller - pwest - 09-12-2021

Hi Rob:
These all sound like great updates! Concerning:
- Added a real time clock, RTC, so the controller isn't reliant on a network, with an RTC you can run it off line and overall makes a more reliable system as controller is in trouble if no time is available.
Could you make a compatible plug-in/accessory module available so those of us who have the old controller can get that functionality?

-Phil


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 09-12-2021

Hi Phil, yeah you can add an RTC to the current controller, either of these will plug into the I2C port on the controller. The DS3231 is more accurate.

DS3231

https://www.amazon.com/gp/product/B082G6MXVR/ref=as_li_qf_asin_il_tl?ie=UTF8&tag=5302020-20&creative=9325&linkCode=as2&creativeASIN=B082G6MXVR&linkId=8b85c5c380873b768c181f048bfaa01b

DS1307

https://www.amazon.com/gp/product/B07XB2265S/ref=as_li_qf_asin_il_tl?ie=UTF8&tag=5302020-20&creative=9325&linkCode=as2&creativeASIN=B07XB2265S&linkId=52724100bee1754aeb25800659caa1da


RE: Robo-Tank 2.0 - Updated Controller - fietsenrex - 09-17-2021

If the ds3231 is more reliable why did you chose the ds1307?


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 09-17-2021

It's not reliability, just accuracy. I didn't go with the DS3231 because they are about 18 times the price as DS1307. It's cheaper to buy the DS3231 on a module but I just want the chip. Also the DS3231 chip is iffy to get right now with the chip shortage.

Because 99% of users will be connected to a network I can get the date/time every day or whatever and update the DS1307. If I remember correctly I think DS1307 can lose 5-10 minutes a year.


RE: Robo-Tank 2.0 - Updated Controller - fietsenrex - 09-20-2021

That makes sense, it’s nice to have a back-up for when the internet goes down or if you don’t want it connected to the internet at all


RE: Robo-Tank 2.0 - Updated Controller - fietsenrex - 09-24-2021

Is it available in the shop? :p


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 09-24-2021

Not yet but I'm planning to order this weekend or early next week. Also for now if anyone wants the new controller they can order the current and save a few dollars as I will be raising the price a little bit when I list them.


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 09-24-2021

Just got the face plate finished.

[Image: Deluxe_Faceplate_RevA.PNG]

For the sensor extension I changed that up a little. The picture is the one used today, I have two versions, one for optical sensors and another for non-contact sensors. The board does have solder jumpers so a person can reconfigure port for either sensor type.

The new sensor extension looks the same but more spacing between ports and it has 3 XH sockets on two sides of the extension so 6 sockets total. One side is configured for opticals the other side for non-contacts so still only 3 sensors can be connected but no more solder pads to mess with. The danger to this is if an optical is plugged into a non-contact port it will be damaged, non-contact in an optical won't damage it though, it just won't work. On the top of case each side will say the sensor type so if a person looks they should be safe. I did it like this because it was a real nuisance for users to cut and solder micro pads. Each port still does have one solder pad though, if it's soldered the port on the non-contact side will work with a DS18B20 temp sensor or DHT22 temp/humidity sensor. Not really necessary but why not.

[Image: CAM01424-Optimized.jpg]


RE: Robo-Tank 2.0 - Updated Controller - Dom59 - 10-20-2021

Rob, with your new board, could it be possible to take on/off the sdcard without open the case? With an extender.

Dom.


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 10-20-2021

Hey Dom, that's not easily possible however you should be able to add one. I don't know how the Pi would react to one, probably be ok.


RE: Robo-Tank 2.0 - Updated Controller - Dom59 - 10-21-2021

(10-20-2021, 02:17 PM)Rob F Wrote: Hey Dom, that's not easily possible however you should be able to add one. I don't know how the Pi would react to one, probably be ok.
Tried this morning and seems ok.


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 10-21-2021

Good to know. Now you just need to break out the glue gun to secure it in the case. My dream is to one day find a person to embed the Pi as they are getting rather expensive, I wanted that for the Arduino controller and now I can do it but don't need to haha.


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 10-25-2021

Alright v2 is finally ready, cases are looking good, website has been updated and new manuals are ready to rock.  K05164 

Unfortunately I did screw up on the sensor extension, you think I would have learned by now but no, not yet, maybe now...

Originally I was just going to stick with the optical extension but after I got the test board I came up with the great idea of having of having a plug for optical and non-contact sensor on a single port so a user could use one or the other. I waited till last minute to make it and didn't think at all, as I put non-contact socket and optical socket on the same data line there's a pullup and a pulldown resistor fighting each other, dohhhhhhhhh.  K0505 Lucky for me one set of ports works as is, the optical side however the non-contact ports don't work unless that pesky little resistor on the optical side is removed. Almost like the old optical extension, in the end the set I have are no different. In a week or two I'll be ordering more boards so it'll be sorted then.

I think I also decided to add a faceplate to the top of the sensor extension. To make it work correctly I do need to add a pluggable jumper so it can be set for one or the other and it would be nice to have a little note on it explaining why dual ports and the importance of not mixing them up.

Here's the various manuals.

This one goes over the controller ports and some diagrams to connect sensors.
https://docs.google.com/document/d/1FtkRidTqk-7t8PHlatf8LvtjlZwUSyRf/edit

This one shows how to install your Raspberry Pi on the Plug and Play controller.
https://docs.google.com/document/d/1tu_5IL4rXsMMK8oiXy_lNPrj1aJW6yBJ/edit

This one show you how to assemble the DIY controller
https://docs.google.com/document/d/1YVbEq8JJhm4dYXYOn43gIUxoNJsnaqwp/edit

Cases can be found here, one of these days I'm going to make this page look better, kind of hard to navigate.
https://www.robo-tank.ca/Robo-Tank_3D_Printable_Cases

I have 7 or 8 of the old ones remaining if anyone is interested, I'm only blowing them out as DIY kits.
https://www.robo-tank.ca/Reef-pi-DIY-Hardware/Reef-pi-DIY-Deluxe-Aquarium-Controller-v1

And here's some pics, the power bar hasn't changed, just posting the full package. You can see the faceplate isn't sitting perfectly square, I'll work on for next order.
[Image: CAM02609.jpg]
[Image: CAM02545-optimized.jpg]
[Image: CAM02584-optimized.jpg]
[Image: CAM02618-optimized.jpg]


RE: Robo-Tank 2.0 - Updated Controller - pavol14pavol - 10-26-2021

Ahoj Rob ako je to riesene s chladenim? Konektor pre malinovku by som viac zdvihol alebo použil inú redukciu na otočenie maliny hore nohami kde by sa dalo pripojiť lepšie chladenie


RE: Robo-Tank 2.0 - Updated Controller - Rob F - 10-26-2021

Hi Pavol, sorry just seeing your message. The fan does help quite a bit but definitely not as good as it would be if blowing directly on the Pi chips. There is more space between the Pi and the controller board as there's no components in the way, I think there's enough room for a low profile heatsink on the Pi main chip but it's close. Unfortunately I don't have one to test.