My Neptune Apex web interface compatible DIY reef controller

[theatrus]

The way that these sensors work is typically there is a tiny prism reflecting an LED light back onto a sensor that diffracts when submerged if I recall correctly, so you are correct you only will get a few millimeters of resolution. (Which can be a good thing, depending on application ) This is same concept as Tunze osmolator ATO uses to know when to top off.


Sent from my iPhone using Tapatalk

Yeah, just no prism but usually just using the change in refraction from the water hitting the dome. They're very reliable.

The Spectrapure UPLC appears to use a pressure sensor connected by airline tubing. I haven't cracked mine open to see exactly how it's setup or what sensor they use (cause I kinda want it to always work ). It can give about 4-5inches of water level difference for the on/off pump control.

 

[d0ughb0y]

In this case, the optical sensor can just be a direct replacement to the float sensor.
you need to get the one with the module board that converts the analog to a digital high low. So I can still use the pins intended for ultrasonic as a water level gauge.

 

[d0ughb0y]

Speed controls for the DCT6000 work fine. My DC6000 died, and I just bought a DCT6000, and swapped it out. Only thing I noticed, was the DCT6000 was pushing more water at the same setting, so I lowered it from 80% to 75% (Your application percentage). I could run it from 0-100 with no issues.

Yes it came with 10 Speeds.

FYI - I just sold my take over the weekend, but I did run the DCT6000 for almost a year.

On fish street, they sell the S1 wifi controller that works with RW pump.
Then they sell the T1 that works with DCS, DCT, Cross flow and maxspect gyre.

I know I read somewhere the DCS and DCT pumps use 50 hz pwm like servo control, so perhaps the 488hz pwm just worked by chance.

Anyway, I am planning on getting a DCT pump so I will be able to know for sure how it works. And based on the description of the T1 controller, the pwm signal should work with DCS, DCT, CP and Gyre. The T1 controller description says Gyre requires a different cable.

 

[shiftline]

My Neptune Apex web interface compatible DIY reef controller

Well I have a DC300 DCT4000 and. DCS7000 pump. I could likely unscrew the control boxes and hook up the scope (still learning how to use it)to compare what the difference is in the outputs. Oh and now a cp40!

 

[d0ughb0y]

Well I have a DC300 DCT4000 and. DCS7000 pump. I could likely unscrew the control boxes and hook up the scope (still learning how to use it)to compare what the difference is in the outputs. Oh and now a cp40!

The DC, WP and RW controller outputs analog 0-5v, but internally it is 488hz pwm. On this diy, I did not convert the pwm to analog voltage, and that is the only reason I can think of why it worked on the DCT pump. And sort of explain why Sorscode says he had to scale the percentage back, because 100% on a servo PWM is not quite 100% duty cycle. I am guessing that's why fish street is claiming 120% overclock (I find fish street tend to stretch the facts quite a bit).

But yes, you can look at the signal on the center pin first. If you can't get much out of it, then you need to open the control box. But for the DCT and cross flow, I suspect there is no need to open the box since the output is presumably pwm already.

I only need the info so I can make a DIY controller that more people can use. I personally do not need it as I don't own a gyre or cross flow, at least not anytime soon.

If you are able to reverse engineer the signal, you can publish the info on your website.

 

[shiftline]

My Neptune Apex web interface compatible DIY reef controller

Sounds good. I'm pretty sure I can actually record the output from the rigol scope too. Just need to figure out how

So 3 pins. Power/pwm and ground. What's the third wire? Would it be for reverse flow. Or would pwm and power be separate(doesn't make sense that way )

 

[theatrus]

Sounds good. I'm pretty sure I can actually record the output from the rigol scope too. Just need to figure out how

You can just pop in a USB stick to the front on most of the Rigol scopes. In the settings you can record images and usually a CSV dump of the actual values.

Also, turn on the 20MHz bandwidth filter

 

[shiftline]

Bandwidth filter?! All these new things lol. It's a 50mhz one I modified to 100mhz

 

[theatrus]

Bandwidth filter?! All these new things lol. It's a 50mhz one I modified to 100mhz

Yup! In each channel settings, you can turn on a bandwidth limiter. Unless you really need to be looking at signals or harmonics above 20MHz, your life is a lot easier and traces cleaner with that on

 

[shiftline]

Do i have to do it every time or just once? i have been trying to figure out how to clean up the signals.. here my first attempt at it .. hopefully it helps!

http://fishtankprojects.com/diy-aqu...g-and-diy-controller-reverse-engeneering.html

 

[theatrus]

It should keep the channel setting, even if you mash auto etc.

Make sure the ground signal is good and ideally close to the measurement point as well.

 

[theatrus]

My Neptune Apex web interface compatible DIY reef controller

Looking a bit at your boards, the controller and the power / FET bridges are integral to the controller board. The motor is also a three phase delta or wye motor (the W Y U terminals are a dead give away). You can probably measure the resistance between the poles when powered down to confirm exact topology.

In effect, in order to use none of the controller you're rebuilding a three phase inverter from scratch. You can probably draw some parallels from this guide (AC motor section, since "brushless DC" is basically AC just with permanent magnets). http://www.nxp.com/documents/application_note/APPCHP3.pdf

It may be simpler to just hot rod into the potentiometer (replace it with a voltage signal most likely) and let their motor driver do the actual work.

 

[theatrus]

If you're interested in rebuilding the driver, using 6 N FETs and a motor gate controller like http://m.ti.com/product/drv8303 would be one option. Not for the faint of heart.

 

[karimwassef]

It may be simpler to just hot rod into the potentiometer (replace it with a voltage signal most likely) and let their motor driver do the actual work.

That sounds like the most practical way to do it.

 

[theatrus]

That sounds like the most practical way to do it.

Wholly agree. Measure the voltage across the potentiometer legs (going to be 3.3V or 5V), and confirm the center leg to one of the others varies as you rotate.

Add a DAC to your Arduino (https://www.adafruit.com/product/935 for one on a nice breakout), which feeds the center pin of the (now removed) potentiometer.

If it's not actually a potentiometer and instead an encoder, you can use a series of pulses to emulate it's behavior.

Changing mode will require replacing the button(s) with a digital output pin most likely. The hard part will be figuring out what mode you are already in.

Of course, for full control, replacing the motor driver is the best but far most complicated setup.

 

[d0ughb0y]

shiftline, can you open up the DCT controller and see how it compares to the CP controller?

 

[shiftline]

Wow I missed a lot today! I have to keep up. Laying in bed today but I'll pop open the RW contoller tomorrow and see (or did you want dct specifically. )

 

[karimwassef]

you don't need to "test" what mode you're in. You can keep track of the first power-up event since the pumps start up at full speed. Replace memory for sensing...

 

[karimwassef]

for full disclosure, I have 3 DCT15000 and I've been "hankering" to hook them up to my Apex for a year... this is most excellent!

 

[d0ughb0y]

Wow I missed a lot today! I have to keep up. Laying in bed today but I'll pop open the RW contoller tomorrow and see (or did you want dct specifically. )

DCT.

Can you read what motor controller chip they used on the CP-40 and DCT controler?