My Neptune Apex web interface compatible DIY reef controller

Come to think of it, it probably won't work. I think they sell propellers for drones specifically for clockwise or counter clockwise rotation. The blades of the propeller has to change direction for this to work.

Paddle wheel boat can go opposite direction because the paddle is flat and not angled. Maybe if the propeller of cp-40 is just flat across it will push water in either direction equally.
 
I'll probably make my circuit similar to the stock controller.

The stock controller uses P and N channel MOSFETS.
As for power, it drops the 24v to 15v and then to 5v. The LM339 is powered 15V but output is pulled up to 5v. The brushless motor board I got from ebay also drops the input power to 15V then to 5v, but it powers the LM339 using 5v. The 15v is used to power the IR2101 that drives all N channel MOSFET. I thought about doing this, but the retail price of IR2101 is just ridiculously high from digikey, like almost $2 each. I'll simplify my circuit by just using a 7815 and 7805 regulator.

I'm still weighing between using LM339 or just feed BEMF analog signal directly to arduino just like in the Atmel application note.

Using arduino as motor controller, you can change feature as needed by simply updating the program. Like for overcurrent control, you can easily change the value via a config parameter so it works properly according to your load (2amps or 4amps).

I'll order parts by Wednesday so I can build the prototype this weekend.
 
maybe you are right, it is really not needed.

For example, the white board I got, the overcurrent limit is 4.5amps. I think if my pump runs dry, the power supply will shutdown as it is rated 2amps only. So perhaps really not a big deal.

As to no soft start, that can be programmed in the controller. I highly doubt the $125 icecap interface can do soft start. I really don't see why anyone would need to pay $125 to do something that can be done under $20.

The signal pin, based on the description, you are supposed to get data if the motor is still spinning or not, and even calculate RPM value. From looking at the signal in an oscilloscope, the signal seems random to me, or at least I could not make sense of it to figure how to calculate RPM. if the pump stops, the signal is low, but not always, sometimes it still puts out a square wave.

But regardless, if the motor stops spinning physically, BEMF signal stops and 24v output will automatically stop. So the pump will be safe regardless.

The ability to know the RPM value is nice, but not absolutely necessary.

From more searching/reading, I think those ebay brushless motor modules are used OEM to control electric bicycles and scooters, small motor operated tools, I would not be surprised if hoverboards use them.

To do it for multiple pumps, an arduino mega or a teensy 3 is needed, as there is only enough pins on arduino uno to control one motor. If going with a teensy 3, a wifi interface can be added. But again, the wifi interface is probably best done in the controller. The fish street T1 is marketed as a stand alone wifi controller, which I think does not really make sense. You really need to integrate it to a main controller.

I may just do it just for the sake of doing it. :)
 
Last edited:
So on the idea of a reversible flow powerhead. I think this is really a big deal. I think part of the reason reefs succeed or fail is flow pattern and most systems only go around in one direction creating detritus traps or killing microfauna against meshes or screens...

Take a look at this

http://m.marinedepot.com/products/t...-programmable-aqua-pump-powerhead?group=false

Taam-Rio-Polario-Turbine-22ML-Dual-Action-Programmable-Aqua-Pump-Powerhead-99.jpg


It was discontinued, but with the right parts and a 3D maching, we might be able to make our own??

I think screw impeller pumps are also reversible? (but they're not DC)

https://www.amazon.com/Cal-Pump-T10000-Gallons-Torpedo/dp/B000EJ0UA2

81YenxD58zL._SL1500_.jpg
 
If only someone was really good at 3D printing... and could build a robust variable pitch propeller for reverse flow...

<iframe width="420" height="315" src="https://www.youtube.com/embed/uTOJEP8jIGA" frameborder="0" allowfullscreen></iframe>

<iframe width="420" height="315" src="https://www.youtube.com/embed/DkHlOt2onB4" frameborder="0" allowfullscreen></iframe>
 
because we don't know it can be done for under $20 .... :)

In fact I think the $12 board is even better, because it has the signal pin to tell you if the pump is really running, which I think is valuable data, as you want to know if your return pump is not really running. You can't do that with icecap interface. I have to study the signal to figure out what it does exactly. There is really no information available on the motor controller chip (did not find any datasheet).

The chip is JY01. If anyone finds any info, like a datasheet, please let me know, so we can push this module to the limit of its features. It is a 16 pin chip. All the other pins are pretty self explanatory except for the signal pin.

Pin 1 is reference voltage for overcurrent protection
Pin 2 is Current sense voltage from current sense resistor.
Pin 3 is Gnd
Pin 4 is 5v
Pin 5 is Z/F direction
Pin 6 is signal pin
Pin 7 is A high
Pin 8 is A low
Pin 9 is B high
Pin 10 is B low
Pin 11 is C high
Pin 12 is C low
Pin 13 is Hall Sensor A
Pin 14 is Hall Sensor B
Pin 15 is Hall Sensor C
Pin 16 is PWM speed control input

You can feed BEMF signal from LM339 to Hall sensor input.
 
I got the higher current $12 board. Can't wait to wire it up to my Apex and DCT15000. If that works, then every other pump should work too. It's rated to 6A, 150W.
 
no date, but I myself want to replace the original controller with the new one. :)
as this is a hobby, I am not able to work on this full time.
I just need to go through the pcb design one more time to make sure I did not forget anything and can submit the order.

The last component I had to test on was for the brushless dc pump. Lots of mis-information all over the internet. :)

I also just learned the esp8266 firmware fixed the SSL issue, so I am going to test that as well. That is required to access most any email account to send and receive email. There are fewer and fewer email servers that allow non ssl connection.

(almost) final design
6 pwm lines - 4 individually selectable 5 or 10v, 2 fixed at 5v
4 doser ports, but 2 lines (5v lines) are shared with pwm. meaning if you use 4 dosers, you are left with 4 pwm.
temp
atlas sensors
16 outlets - if not all 16 are used to cotrol AC relays, spare lines can be used for I/O
ultrasonic sensor
2 ato, will support optical sensor
2 fans
auto feeder
current sensor
and of course, high speed wifi

what I have so far:
schematic
schematic_v2.png~original


pcb - this image is actual size.
pcb.png~original


let me know if I missed something.
 
Last edited:
right, through the motor controller module, and with the possibility of using one of the spare outlet IO line to control direction for gyre.

with DC pumps, you pretty much don't need controllable outlets anymore, so 8 outlets should be more than enough for most tanks, so most likely there will be spare lines that can be used for general purpose IO.

One can use 2 pwm to control WP,RW or CP pumps, 1 DCT return pump, and 2 for kessil LED light.

I try to keep the design compact and has most if not all functions needed.
 
How about a battery backed up DC bus for direct power to DC devices?

I'm thinking a 24VDC marine battery stack to drive DC pumps in case of power loss, for example.... Kinda specific for an example :D
 
That can be done completely independently from the controller so I can keep things simple and stable. It's open source so you can do whatever you want for your own requirements.
 
Ok. I was thinking of having a high current flow through but trickle-charge charging controller. Maybe even a battery charge and health indicator.
 
Back
Top