3
\$\begingroup\$

Background

I'm trying to build a small hobby grade ROV for an underwater robotics competition held at the university I'm studying. And I have used eight brushed dc motors each of which consumes 8A on full load.

The whole system might consume about 75-80A at maximum usage. But we are pretty sure that we are not running all motors at the same time at full load and hence average current usage would be around 50-65A.

I have included eight 12V 8Ah SLA batteries inside our ROV and hope it might run for like one and a half hour at most on average.

Problem

One of the issues I'm having right now is with designing a PCB layout where wires from eight batteries are merged onto the PCB Board.

Please note that I'm not an electronics/electrical engineering student that I do not have proper and solid knowledge on power circuits and calculating thermal dissipation.

I just don't want to be a dumbass letting the PCB burns into smoke and flame from overheating. I am afraid that might happen when the amount of current stated above have to be passed through the plate.

I have space limitation and maximum dimension of the PCB for the distribution panel should not exceed 6"x6" dimension.

Power Distribution

Above is the overview of the power supply and distribution concept for the ROV. After the summation of total current usage and wattage, to a shocking moment, it's around 1kW and 80A of DC 12V. Is it such a high usage of power in sense of power circuits? I never had an experience with circuits that use over 12V 35A. I'm afraid that I might be wasting a lot of power. But my ROV is a bit big in size, about 3'x2'x1.5' dimension of steel structure. But I'm pretty sure we are going to operate only four motors at the same instance and at 70% of full load at max. So I guess 50A is maximum for normal operation conditions.

My Questions

  1. How should I design my PCB for power distribution purpose only?
  2. Shall I use a typical thick PCB or a copper block with heat sink?
  3. Is the total overall power usage huge and not suitable?
  4. Your advice on making things better.
\$\endgroup\$
6
  • \$\begingroup\$ So the power dissipated over a resistor is P=I^2*R. If the resistance of the "trace" is small, there will be low power loss and therefore it won't heat up. So what you need to look into is current carrying capacity of traces/pins/wires and figure that out to see what your design can handle. Start with this 4pcb.com/trace-width-calculator.html. I think this will set you on the path to solve this. But you are correct to be concerned about traces/wires to burn up (aka fusing). \$\endgroup\$
    – klamb
    Nov 4, 2016 at 13:08
  • 2
    \$\begingroup\$ 80 A is a lot for a circuit board. You will definitely be better off not running 80A through your circuit board. I would run wires directly from batteries to speed controller for the motor, or use some type of high-current bus bar or something. \$\endgroup\$
    – user57037
    Nov 5, 2016 at 4:09
  • \$\begingroup\$ Have you told us everything? Why is there a heat sink suggestion for the PCB? Are there motor controllers /other circuitry on it? \$\endgroup\$
    – Paul Uszak
    Jun 25, 2017 at 12:02
  • 2
    \$\begingroup\$ Consider whether a PCB is the right structure for distribution of high current loads. They have a big problem; all the conductors are flat. That means it's hard to get a big enough cross sectional area in your conductors for a sufficiently low resistance. You might want to consider using wires and connectors, with only low current control signals on your PCB. \$\endgroup\$
    – Ian Bland
    Jun 25, 2017 at 12:54
  • 1
    \$\begingroup\$ This is a damn serious system design. But one thing I can tell for sure is that brushless motors would be much better, because their coils are on the stator and are easier to cool down, especially under water. \$\endgroup\$
    – user76844
    Aug 25, 2018 at 21:27

2 Answers 2

Reset to default
1
\$\begingroup\$

1- You should think of terminals and connectors while designing your PCB, you should also think about the width of traces since those traces will support high current (8A per every trace) and I'll give you an example: Every battery pack will be connected to the PCB by a XT-60 connector. Then connected by traces with largest width possible (say 10mm) and then branched to the 8 motors which connect to the motors wires (say using a fork terminal).

2-(from my experience with a similar situation) a thicker copper PCB is just fine 1 or 2 oz is more than fine, the most important thing is trace width, you can calculate what width you will need at this website.

3-No Power usage is just fine, I made an ROV with 6 thruster where each thruster can use up to 17A, using 4 arms using DC motors each can use up to 8A, so don't worry. If you mean power dissipation in the PCB then don't worry too, I tried it before and at 1oz thickness 8A won't dissipate that much heat, you won't need a heatsink.

4-The ultimate way to make this power distribution PCB is to design it and fabricate it and then solder a copper wire on the traces already there on the PCB, it will make it uglier and beefier but it's safe and will support as much current as you will need. I hope I answered your question :)

\$\endgroup\$
-1
\$\begingroup\$

Design Considerations: Cable power loss %, Battery energy density (Wh/kg), Buoyancy ( air presure force vs battery weight), wall pressure, choice of motors, battery and controller.

At least learn Ohm's Law V=IR and \$Pd=I^2R=V^2/R=V*I\$

  • Get a DMM (cheap) and measure Motor winding Resistance or DCR or get locked rotor=surge current spec.

Cable Loss You can look up AWG tables to get Ampacity @60'C or compute loss, Pd from above using \$\Omega/m\$ vs cable cross-section area or gauge or measure flat braid 1cm*1mm wide for example. There will be voltage drop. You want your cable loss to be <2% of surge current to reduce loss of torque.

  • motor torque = current ( prop load and acceleration of rotor mass)
  • (no load) motor speed = Voltage (V/RPM)
  • Motor power rating = rated current*V

  • Surge current = 8 to 10x rated current.

    • e.g.
    • worst case surge from I=V/R (for Vbat and DCR of motor coil) is up to 8x to 10x8A or 64A to 80A per motor !! So start up slowly

    • if you already have motors picked out OK, although I would use two >=21V drill motors, battery controller and all modified to work inside the sub with quadruple "O-ring seals" around the prop shaft made for high pressure.

    • Best wiring is multiple strands of insulated magnet wire called Litz Wire due to low inductance, which can cause EMI interference with electronics unless shielded and ferrite filtered in pairs.

    • next best is heavy braid like car battery cables.
    • Then use Terminal blocks for distribution like fuse panels in cars
    • Traces on PCB may need solid copper AWG 16 with low R = RdsOn to get same low loss as per above equations.

Buoyancy

  • \$F_B = ρ_f V_f g \$
  • \$F_B\$ is the buoyant force, \$ρ_f\$ is the density of the displaced fluid, \$V_f\$ is the volume of the displaced fluid, and \$g\$ is the acceleration due to gravity, \$9.8 m/s^2\$.
  • It's very important to remember that the density and volume in this equation refer to the displaced fluid, NOT the object submerged in it.

Consider Water pressure is 1Atm per 10m depth so 100m depth is 10 atmospheres or 150 psi, ( I recall) and see if SLA will float or sink vs LiPo. Then make structure like a tank to survive this pressure and the best seals to match...

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.