Use PWM to control your Jebao DCT pump using brushless motor controller from ebay

karimwassef said:
The reverse feature could be very useful with a jammed impeller. If the signal pin indicates that it's not turning, a few seconds in reverse could help unjam it.
Click to expand...

True. But if it's a dead fish, it'll spread coarse mangled pieces of dead fish all over your tank. Not only that, if it's something hard like a snail shell it could wreck your propeller and severely damage your pump.
 
karimwassef said:
So. Anyone willing to share the wiring diagram Arduino to Jebao motor shield?
Click to expand...

not sure if anyone is working on that.

mine is essentially what I already posted.
while waiting for the fets, I am going to start working on the pcb routing since now I have a pretty good idea what the circuit will be like.
 
I'll probably use a headphone jack for the control cable connection. I think they are now pretty standard. I'm still trying to figure out if I can use a 4 contact TRRS jack so I can include the Tach signal output pin. I think if I put a 4.7k resistor in series, it should not affect the Arduino pin if the tach out gets connected to ground. For controllers that can actually read input IO signals, the tach pin can be useful to determine RPM of motor, or at least tell if the motor is spinning if there are pulses coming out of that pin. Say if you connect to Apex IO port (which I think can really do input only), you can read this periodically to make sure the pump is spinning. If say you get no reading after 10 tries, then you can make the controller send an alarm.


I am making the pwm input voltage jumper selectable 5v or 10v and can be calibrated to get maximum value.
 
I'm looking for new pump

I'm looking for new pump

I got a reef angel controller and can do 5v pwm/anolog or 10v pwm /anolog

What is the minimum settings this way and the news updates and pumps for return around 500gph(for current tank) or less - 2000gph or more (for future tank)

Was thinking dct6000 or 9000 or dcs?
But they don't go low enough
 
d0ughb0y said:
I'll probably use a headphone jack for the control cable connection. I think they are now pretty standard. I'm still trying to figure out if I can use a 4 contact TRRS jack so I can include the Tach signal output pin. I think if I put a 4.7k resistor in series, it should not affect the Arduino pin if the tach out gets connected to ground. For controllers that can actually read input IO signals, the tach pin can be useful to determine RPM of motor, or at least tell if the motor is spinning if there are pulses coming out of that pin. Say if you connect to Apex IO port (which I think can really do input only), you can read this periodically to make sure the pump is spinning. If say you get no reading after 10 tries, then you can make the controller send an alarm.


I am making the pwm input voltage jumper selectable 5v or 10v and can be calibrated to get maximum value.
Click to expand...

I've always used cat-5 connectors. They are a little bigger, though.
 
I got the 5 way tactile switch. I don't think it will work as well as regular tactile switch, and the size is really small. So I'm going back to using 3 regular tactile switch. For 3 functions, left, right, mode (just like dct controller).

Since it appears CP and DCT pump timing are different, I am going to program them separately, but user selectable.

When when plugging in the controller, if the mode button is held down, it will switch between CP mode or DCT mode.

I am using two 74hc595 shift registers so I have 16 LEDs. 10 will be used for speed display, and the 6 for mode.
In CP mode, Ext, H1, W1, W2, W3, ELSE
In DCT mode, Ext, H1
press and hold mode button will switch to feed mode for 10 minutes. During feed mode, I will make the 10 leds display the count down remaining time. I hate it when I don't know how much time is left during feed mode. During feed mode, I will make all 6 mode leds ON.
In either DCT or CP mode, if you switch to Ext mode, the pump speed is controlled by the PWM input.
In CP mode, press and hold left button to switch to time mode (equivalent to the knob on the CP controller), then press left/right to adjust time. Press hold left button to switch back to speed mode.

I can still use the right button press and hold for other things, like pwm calibration, change direction for CP mode, etc.

While pump is running, I will make the 10 leds display the actual speed. So for H1 mode, it is fixed, but for wave modes W1, W2 and W3 and likewise for Ext, I will make it display in real time the actual speed.

sort of like this
<iframe width="420" height="315" src="https://www.youtube.com/embed/dvPxNqefzbQ?rel=0&controls=0&showinfo=0" frameborder="0" allowfullscreen></iframe>

https://youtu.be/dvPxNqefzbQ

I am still waiting for the FETs parts to arrive, I hope to get them by next weekend.
 
Boards arrived, not all the parts yet so I haven't had a chance to populate anything. The board however does fit into the stock Jebao case:

boards.jpg
 
stage3-s4 said:
Very nice. Do you solder all those components by hand with a re-work gun?
Click to expand...

I stencil print (prototypes I'll cut with the laser from kapton, larger runs I'll buy a stainless steel), hand place, and toss in the reflow oven. Doing it was a hot air gun also works, but its harder to get everything consistently hot.
 
karimwassef said:
Which reflow oven? How do you do bottom side placement?

Would a hot plate work?
Click to expand...

The cheap and cheerful T962A IR. Toaster oven and skillet work great.

I can do both sides but just running it through the process twice, but in this case I'll do the one or two "top" side parts by hand or hot air.
 
I'd personally just use a skillet at prototype stage - the Hamilton Beach one has a nice flat bottom. You can watch the process through the lid :) Crank it to full, wait for everything to be shiny, switch it off and open the top.

The T962A (A at the end) is important - the none-A just doesn't have enough punch for its volume to do lead free right, or massive parts. The A however pretty much needs its own 20A breaker when heating.
 
theatrus said:
The T962A (A at the end) is important - the none-A just doesn't have enough punch for its volume to do lead free right, or massive parts. The A however pretty much needs its own 20A breaker when heating.
Click to expand...

But lead is the good stuff =P.

I've only been partially following this and yesterday i saw a post about the DCP line of pumps, which seem to be running full sine waves instead of square wave to closer emulate real 3 phase power. I know 3 phase AC is butter smooth with the overlap and all that. Any thoughts on going to that? I can imagine how a square wave will basically hammer the rotor around instead of the rotating field 3 phase sine waves make.

I actually wonder if they are really doing a sine wave or just strapped some big caps on the output, so it charges slow and then cut off early so it drains the rest of the time... Or maybe they are just splitting the output into sections and running lots of square waves and smoothing that out a bit (or not at all).
 
I've done a few of the mods, but haven't done the whole firmware replacement yet. The forced air convection during heat is my next mod on the list :)


Sent from my iPhone using Tapatalk
 
I got the fets and wired it up, still no go.
In the process, I have burned out a few of mosfets (the current one and the previous to-220 ones) due to misconnections, wrong pin assignment (had to move pwm pin to another one, and forgot to update my schematic), etc. For this last incident, I had the fet pair clamped together (to connect both drains), and it got so hot, it literally welded the two fets together. I have fixed my schematic, wiring and software to now all match. :) The fet output is switching, but I do not see any PWM on the output. If I just wire the N channel by itself in a test circuit, I can see the output has 24v pwm signal. So not sure why when wired as high/low driver, there is no pwm in the output. I am thinking it may be due to stray capacitance, so I'll try to lower the pwm frequency, which is currently at 15khz. I can see the 15khz square wave on the fet input, so I'm not so sure if it is really due to capacitance. The output signal looks like a deformed trapezoid wave, so I suppose there is capacitance. The stock controller output signal is a nice trapezoid wave chopped up by pwm.
 
I think I'll order a small prototype pcb containing only the fet and bemf circuit. I think a big part of what's causing my problem is my circuit is on a breadboard.
 
So I decided to hook up my breadboard circuit to the pump, and was surprised to see it is actually working!! So I guess I don't have to use a prototype pcb for now. I was going by the output waveform thinking it won't work. Don't want to hook it up to the pump and burn out the coils. So I checked the stock pump output waveform, and it is the same as my breadboard circuit waveform. I must have hooked up the oscilloscope to another test point when I got the trapezoid wave with pwm pulses on the stock controller. Anyway, I can now continue on to the bemf circuit and code.

This is my open loop start up code. I run it between the single and double beeps.
<iframe width="560" height="315" src="https://www.youtube.com/embed/lxx-Tj7QpzY?rel=0&controls=0&showinfo=0" frameborder="0" allowfullscreen></iframe>

https://youtu.be/lxx-Tj7QpzY
 
You must log in or register to reply here.

Similar threads

R
Ph Control Set Up
Replies
0
Views
171
rdavidw
R
B
JBJ Arctica 1/10 hp Chiller Modification for External Control
Replies
9
Views
748
griss
griss
T
New Deltec Skimmer
Replies
1
Views
90
Vinny Kreyling
Vinny Kreyling
S
New to Manifolds - Return Pumps
Replies
4
Views
650
BeanAnimal
B
N
Is this a design flaw from Red Sea? Please help me with this sump upgrade Max e170 (Sump and tank connection)
Replies
1
Views
370
wvned
wvned
Back
Top