POE to run a DC motor and controller?

[pratiken]

Hi guys
I'm tinkering with an idea that I'd like to get your thoughts on.

I'm trying to separate the microcontroller portion from the motor and controller apparatus.

Can I have a PIC board running my code a few feet away from the apparatus with a 5 foot ethernet POE connection over to the DC motor and motor controller?

The PIC board would have a 12V DC input that I'd run over the ETH cable to the controller.

What do you think? Is this a silly idea or could it work? It looks like I'd have 1,2 and 3,6 lines on the RJ45 to send pulses to the controller, right? Then 4,5 and 7,8 would be used for V+/V-

PIC board with RJ45 & 12V input ---------5ft CAT7 --------> Motor Controller --> 12V DC Motor

[uncle_bob]

Hi

It probably will work with some size of motor. It is likely to be trouble on a 10A 12V motor. You will spend some time doing filters to take care of the noise.

Bob

[pratiken]

Interesting! Same goes for this motor which I'm going to use?

https://www.servocity.com/html/45_rpm_hd_precision_planetary_.html#.V2iWB_krKUk

Thanks

[Ian.M]

It would be workable up to 1.5A if you use premium high temperature 22 AWG Cat7 cable. Cheap 24 AWG Cat5 cable may only be good for half of that.  At 0.75A/pin expect a lot of connector trouble . . . . .

You can forget about trying to make logic level signalling sharing a return with the power/ground work, especially if you are splitting pairs to get four H-bridge drive signals down two pairs!  You'll probably need to opto-isolate at the motor end to overcome ground bounce, and may need to power the opto LEDs from >5v so there is enough drive to avoid the ground bounce modulating the LED intensity excessively.
If however you are using two signals, each on a separate pair so its got its own return, its somewhat easier as you no longer have to contend with ground bounce directly in the signal loop, though you are still likely to need either balanced line transmitters and receivers or opto isolation at the motor end.

However, with your motor's rated load current of 2A, and stall current of 20A, IMHO you are asking for trouble.  You *may* get away with bucking a 48V supply over the cable to 12V at the motor end, which would give you about 4A for the  motor + careful current limiting to avoid over-current during acceleration.

[pratiken]

Thanks for the clear explanation! Would an RS232 cable be more reasonable? Or any other cable for the matter?

[Ian.M]

Data cables in general tend not to have enough copper, however a thick (22AWG) fully wired 25 pin DB25 parallel cable, using 10 pins for power, 10 for ground, and the remaining 5 for control signals and signal ground with optoisolation at the motor end to cope with the ground bounce, should be good for around 4.5A - 5A continuous.

Stick an anti-parallel diode and a 1K resistor across each Opto LED so any ringing in the cable doesn't blow the LED and drive then at around 75% of their max continuous If rating, with the current limiting resistor split equally between the LED end and the driver end.

Alternatively, use RS485 or similar differential serial balanced line for a two wire data link and put a slave MCU at the motor controller end. N.B. you'll need to be careful of the transceiver's common mode rating due to ground bounce in the power wiring.

[kc8apf]

As a rough guideline, traditional POE (802.3af) provides ~15W while POE+ (802.3at) provides ~24W.  Those standards use a wide voltage range (40-60V at the injection end) and allow for significant voltage drop over the cable.  If you are starting with 12V, you're going to have much lower overall power delivery.

[Pack34]

Is there a reason this system is designed this way? A cleaner solution would be to have a battery and the motor controller near the motors then trickle charge the battery down the PoE cable. Then send the motor commands using Ethernet.

[uncle_bob]

Hi

You have a cable that is designed for ~15W and a motor that can pull 240W. Right there you have the makings of a fire. You would need a good switching supply to go between the motor and the cable as well as first class filtering. If you are looking at this as a way to save money ... it's not going to do so. If you have a few hundred dollars per system, there are ways you can do things:

1) Put in a rechargeable battery at the motor end.
2) Run a switcher between the cable and the battery.
3) Run the motor on the battery
4) Spend real money on a (likely custom) filter to keep the crud off the cable
5) Insulate the entier "far end" of the system so it can float off ground.
6) Run the motor at a duty cycle that the charger (etc) can keep up with.

It's never a matter of "never", it's always a question of budget.

Bob

[pratiken]

Thanks for all the excellent information!!

What about separating the motor driver from the motor and using a heavy gauge wire from the driver to the motor?

MCU+Driver  ---14AWG wire---> DC Motor
http://www.mouser.com/ProductDetail/Amphenol-SINE-Systems/P29008-M6/?qs=sGAEpiMZZMv%2f60yVGPe5wCGX83opBZCworP18vMo4f0%3d


Using this as a reference
http://www.powerstream.com/Wire_Size.htm



OR


Use a separate DC power cord for power only?
http://www.mouser.com/ProductDetail/Molex/79516-1041/?qs=sGAEpiMZZMv%2f60yVGPe5wOIdTIaR%252bw9Xedw2yAYJqg0%3d

Then use the Ethernet cable for the motor controller data only?

[uncle_bob]

Thanks for all the excellent information!!

What about separating the motor driver from the motor and using a heavy gauge wire from the driver to the motor?

MCU+Driver  ---14AWG wire---> DC Motor
http://www.mouser.com/ProductDetail/Amphenol-SINE-Systems/P29008-M6/?qs=sGAEpiMZZMv%2f60yVGPe5wCGX83opBZCworP18vMo4f0%3d


Using this as a reference
http://www.powerstream.com/Wire_Size.htm



OR


Use a separate DC power cord for power only?
http://www.mouser.com/ProductDetail/Molex/79516-1041/?qs=sGAEpiMZZMv%2f60yVGPe5wOIdTIaR%252bw9Xedw2yAYJqg0%3d

Then use the Ethernet cable for the motor controller data only?

Hi

Moving power over power cables is always better than moving power over data cables

You still will need to worry a bit about ground issues. You have (even with bigger cable) significant current in the motor. That gets you current in the ground lead. Inductance and resistance in the ground lead gets you noise on the ground side of the motor power. An isolated controller is a good solution to these sorts of problems.

Bob