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

I think the reason you can only use 24v is the mosfet max Vgs is 20v so the pre driver must not use >20v. I think it uses 18v.
Click to expand...

Thats 20V above source, which for low side switches is easy, and its why N-fet high-side drivers reference off of the source line when boosting. Of course using P-fets as the high side switches avoids the need for the boosting gate driver.

Usually drivers would only need to deliver about 10V above, anything else is diminishing returns in terms of FET switching for logic-level gates.

d0ughb0y said:
Apparently, the reason the ebay sellers dropped the price of crossflow is jebao is coming out with a new version. No further info on what changed.
Click to expand...

Probably a 0-10V speed control input :lolspin:
 
Last edited:
theatrus said:
Thats 20V above source, which for low side switches is easy, and its why N-fet high-side drivers reference off of the source line when boosting. Of course using P-fets as the high side switches avoids the need for the boosting gate driver.

Usually drivers would only need to deliver about 10V above, anything else is diminishing returns in terms of FET switching for logic-level gates.
Click to expand...

I'll see what voltage the npn pre driver is connected to. I don't think it can be connected to 24v, as that will exceed the 20v Vgs to the p-fet.

Probably a 0-10V speed control input :lolspin:
Click to expand...

I am guessing a wireless link between 2 or more crossflow, just like the latest incarnation of the RW pump (which they call PP)
 
I found this circuit for driving p-fet

mosfet6.jpg


It adds a 10v zener to keep the Vgs from exceeding -20v.

I'm thinking if I use a smaller value, say 9v zener, the circuit will still work using a 12v power supply, given min threashold Vgs is -2v
 
P-FETs are very lossy and will thermally limit output power unless the duty cycle is low.

All n-channel on a bridge usually allows for the highest efficiency and power throughput
 
I traced the cp-40 circuit to p-fet, and it just use a transistor (presumably npn) with 1.5k collector resistor and parallel capacitor to 24V and collector is directly connected to p-fet gate. The transistor then has a 620 ohm resistor from emitter to gnd and a 1k base resistor. It did not use 18v source in the circuit.

The low side to n-fet is just 100 ohm from mcu pin to gate and 100k to gnd.

I'll use this same circuit for starters.

I think the 1.5k and 620 ohm sort of form a voltage divider, less whatever collector to emitter voltage when transistor is on. so Vgs would be a little below -17v on a 24v supply. and should still work on 12v supply, Vgs will be less than -8.5v

I'll have to power it up and check the voltage on oscilloscope to see how it really works. The 620 ohm resistor should have 3-4v only if there is 5v going to base resistor.
 
Last edited:
I traced the white ebay board, and it is a simple npn circuit, with 4.7k base resistor, and 2 1k in series on collector and the midpoint is connected to p-fet gate.

so for this circuit, Vgs will be half of 24v - collector-emitter voltage of transistor. so approximately 11-12v. I'll try this circuit as well and see which one works better.

I'll order the parts tomorrow and plan to wire a circuit on a breadboard.
 
no. I'll try when I get the chance. I think it will work. I think the cp-40 does not have soft start like the dct pumps, or at least I didn't notice it.
 
d0ughb0y said:
no. I'll try when I get the chance. I think it will work. I think the cp-40 does not have soft start like the dct pumps, or at least I didn't notice it.
Click to expand...

Its probably much faster of a startup curve. I think pretty much all pumps run open loop for at least a few revs.
 
I'm going to use the program from atmel application note as a starting point.
this is the motor driver circuit from the application note. I think there is a typo on the fet part number.

I am going to use a voltage divider on the npn collector resistor to stay below Vgs max.

atmel_an.jpg~original
 
Last edited:
I want to keep the circuit as simple as possible.
The jebao circuit does not use push pull yet the waveform and performance is textbook perfect.
 
I connected the cp-40 controller to my dct-4000 pump, and although I can hear the motor spin (or at least that's what the whining sounds like), it was not pushing water, or very little water.

Not sure why it did not work. We know fish street sells the T1 controller that works with cp-40 and dct/dcs pumps.

Anyway, I may pull out the pump from my sump one of these days and test it in a bucket to find out exactly what is happening.
 
I captured the commutation signal of the cp-40 controller while connected to cp-40, and it looks very similar to the dct controller. I captured the H1 signal, and the only difference is, there is definitely no soft start routine, and the lowest setting starts at 58% duty cycle instead of 70%.

When the cp-40 controller was connected to the dct pump, when I increased the speed to 100%, after a few seconds, the pump stopped and the warning light started blinking. I'm guessing maybe the cp-40 may be configured to run at a lower current. I can try to capture the cp-40 signal while it is connected to dct pump maybe tomorrow.
 
d0ughb0y said:
I connected the cp-40 controller to my dct-4000 pump, and although I can hear the motor spin (or at least that's what the whining sounds like), it was not pushing water, or very little water.

Not sure why it did not work. We know fish street sells the T1 controller that works with cp-40 and dct/dcs pumps.

Anyway, I may pull out the pump from my sump one of these days and test it in a bucket to find out exactly what is happening.
Click to expand...

the cp driver may be too weak.

i meant the opposite - run the cp off the dct controller
 
The cp-40 and dct controller are essentially identical, parts wise.
My theory is, the soft start routine may be what is needed (possibly due to head pressure) to drive the dct pump properly, hence they turn a necessary software procedure into a feature.
I'll check the power usage to see if it exceeds 50watts to cause the warning light to blink.
 
So can you run the CP on the DCT controller? That should have power and soft start. Might want to set it to the lowest setting though :D
 
I'll try later. I'll also try the white and green ebay boards.

This is exactly why I want to use mcu to control commutation timing, it gives flexibility compared to using a dedicated motor controller. I'm pretty sure the unidentified 28 pin chip on the jebao controller is an mcu and not a dedicated motor controller chip.
 
cp-40 controller on cp-40, max power 43 watts
dct controller on dct pump, max power 35 watts
cp-40 controller on dct pump, at 9 leds, power was 40watts, at 10 leds, controller shuts down.
dct controller on cp-40, max power goes to 28 watts only, and cp-40 just mildly pushes water.

so clearly there is a mismatch between the controllers and pumps, despite them having the same FETs and bemf circuit.
 
You must log in or register to reply here.

Similar threads

R
Ph Control Set Up
Replies
0
Views
154
rdavidw
R
B
JBJ Arctica 1/10 hp Chiller Modification for External Control
Replies
9
Views
729
griss
griss
S
New to Manifolds - Return Pumps
Replies
4
Views
644
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
353
wvned
wvned
andiserax
OCEANBERRY - The DIY aquarium controller
Replies
4
Views
2K
Misled
Misled
Back
Top