Electric propulsion design process

Discussion in 'Electric Propulsion' started by Will Fraser, May 17, 2019.

  1. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I am in the process of creating a video series on the design process that I followed for my solar kayak. It will also include investigations into batteries, motor controllers etc. I hope the test demonstrations and discussions will be a useful design guide for other diy enthusiasts.

     
    Dejay and BlueBell like this.
  2. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    To make testing easier, I am using the small toy motor and a 3W panel.
    The torque measurements need refinement but I am very pleased with how well the simple optical tachometer works.
    The results nevertheless clearly show the mismatch between the motor and panel. Best motor efficiency corresponds with an under-utilised panel and maximum panel power corresponds with an overloaded motor. In a follow-up video I will show how to calculate this motor-panel compatibility without the need for tests as well as options to remedy a mismatch.

     
  3. Dejay
    Joined: Mar 2018
    Posts: 348
    Likes: 48, Points: 28
    Location: Europe

    Dejay Senior Newbie

    Wow, amazing video! Finally got around to watching it :) I roughly understood the steps before but this lays it out very nicely. Thank you very much.

    Looking forward to your followup video on motor efficiency, that is the area where I understand the least so far.
     
  4. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    The motor instalment.
    I did do a constant voltage test on the little 7V toy motor and got better results, including running torque at various loads.
    The efficiency is very low and as such the little motor does not lend itself to the simplified analysis described in the video.
    Fortunately, almost any motor powerful enough to be practical as a boat motor would be of high enough efficiency to use these simplified equations and ratios.

    I have since also built a bench-top "dyno" to test slightly larger motors, and there is a shiny new 2hp brushless motor waiting its turn. I am looking forward to doing some first-hand efficiency comparisons between brushed and brushless motors.

     
    Dejay likes this.
  5. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    DSCF1404-1.JPG DSCF1405-1.JPG My hacked-together dyno is up and running!

    Quick overview: the watt-meter reads power in from a 13V LiFePo battery.
    The motor is a brushless Keda 195KV, rated for 80A at 22V. A servo-tester connects directly to the ESC to control the "throttle".
    The motor is attached to the wooden lever resting on the scale. The scale reading gets converted to torque by multiplying with the lever arm.
    The blue 18650 cell powers a photo-reflective diode positioned between the motor and the scale. This circuit creates a pulse every time the white masking tape passes the diode.
    The pulse frequency is measured by the red multi-meter. Multiply reading (Hz) by 60 to get rpm.
    The load is a brushed motor driven as a generator. I am using a cordless drill gearbox in reverse to step up the rpm to the generator.
    The generator output wires feed into a drill pwm trigger which in turn powers two headlight bulbs. These are attached to a volt/ammeter just out of curiosity.
    The pwm trigger allows me to vary the voltage to the bulbs. This in turn controls the current and torque on the generator so that I can get a range of readings at a given throttle setting.

    Some notes on readings I have taken:
    The motor on its own consumes 20W at WOT. Once installed, a heavy duty 12mm sealed bearing takes up another 10W.
    The gearbox is not very well aligned and there is some eccentricity on a coupling I had to make on the generator side. Together, these two components push up the power to 100W, most of it noise from the gearbox!
    With the load drawing maximum current, power-in goes up to 156W. I calculated power-out at 137W. That is 88% efficiency with the motor still operating well under-loaded.
    I expect efficiency to peak somewhere between 200-25oW at this voltage. I have a few more bulbs that I can add in parallel to draw more current and increase the load. If that is still not enough I will just short out the generator, that should really put the brakes on it.
     
    Dejay likes this.
  6. gonzo
    Joined: Aug 2002
    Posts: 13,537
    Likes: 368, Points: 93, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    Take into consideration that the stall torque is not going to be the same than the available torque at maximum RPM.
     
  7. Dejay
    Joined: Mar 2018
    Posts: 348
    Likes: 48, Points: 28
    Location: Europe

    Dejay Senior Newbie

    I'm sure you're doing something clever here but I have no clue what.

    Oh I see both the motor case and the motor axis can both rotate so the load measured with the scale at a certain RPM will be equal to the load of the motor. Very clever! Or maybe it's just normal clever. Or maybe it's really simple and I'm just a bit stupid ;)
     
  8. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    I expect torque to increase linearly as rpm decreases.
    The torque produced in the generator is directly proportional to the current in its windings. As the resistance across its terminals decrease, such as by adding more bulbs in parallel, current will increase.
    The resulting increase in torque will slow down the brushless motor, resulting in it also experiencing an increase in current and torque.
    The reduced rpm decreases the generator's ability to generate current and torque.
    The rpm will continue to decrease until the increasing torque on the brushless is in equilibrium with the decreasing torque on the generator.
    This system is therefore incapable of stalling. Even with the generator shorted, it can only produce a torque load if it is actually turning.
     
  9. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    Exactly that.

    The aim is to obtain efficiency curves at different input voltages. As voltage increases, the power at which peak efficiency occurs should also be higher.
    The results can be used to optimise performance on a boat with limited input power available, such as a solar boat.
    I have two 300W panels, so I am trying to establish the voltage at which the motor's peak efficiency corresponds to an input power of 600W.

    Once I know the torque and rpm at that condition I can select an appropriate propeller for the estimated cruising speed.
     
  10. gonzo
    Joined: Aug 2002
    Posts: 13,537
    Likes: 368, Points: 93, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    An increase in torque can not generate an increase in current and torque. Expecting it to do something convenient is wishful thinking, not measured values.
    You again claim that rpm decreases with increased torque, which is the opposite to what an increase in torque will cause.
    Claiming that a system is incapable of stalling is saying it can output an infinite amount of power, which if true will void all known science and technology. However, I dare put my reputation on the line by claiming it is complete nonsense.
     
  11. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    Perhaps try and put in your own words how you see this system stalling so I cannot figure out where your misunderstanding lies.
    This is bog standard motor basics.

    With the generator terminals open the rpm is around 2660 and with the motor completely uncoupled it goes up to 2800. Actual KV is therefore 216.

    I went ahead and shorted out the generator. The scale maxed out so I could not calculate torque, shaft power or efficiency.
    Rpm dropped to 1450, 52% of no-load rpm so the motor should be just a fraction shy of maximum power-out and roughly at 50% efficiency.
    Power-in was 570W. I am very lucky that my ESC did not get fried!

    With two more bulbs to load the generator, power to the motor was up to 170W.
    Rpm 2480 (88% of n-max),
    Scale 590g, so with a lever of 0.1m that gives a torque of 0.6Nm.
    Power-out 156W so efficiency is up to 92%.
     
  12. Dejay
    Joined: Mar 2018
    Posts: 348
    Likes: 48, Points: 28
    Location: Europe

    Dejay Senior Newbie

    I think this is just confusion about the torque and current of the motor vs the generator in the test setup. The generator just acts as a variable brake to test the efficiency of the motor coupled to it. The less electrical resistance attached to the generator the more current will flow at a certain RPM and the more power is consumed and the more brake torque it will transfer to the brushless motor that is coupled to it.

    And in turn the brushless motor has to work harder so the back emf goes down and the more current and power it consumes. Hope I got that right. I still need to learn more about electrical motors.

    BTW there is a good youtube channel by Jeremy Fielding about motors and other cool stuff. Not saying any of you need it, but for me and maybe others interested.
     
  13. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    Results so far for 13V.

    I have added the shorted generator load using a torque estimate that would yield 50% efficiency. It gives a better picture of the curve trends beyond 200W.

    At 184W output the efficiency had already started decreasing (down to 88%) so it looks like my earlier measurement of 156W (170W input) was in fact at or very near to max efficiency.
    Rpm was at 88% of n-max which is also consistent with max efficiency rpm for brushed motors in this power range.

    Keda dyno graph 13V.JPG
     
    Dejay likes this.
  14. gonzo
    Joined: Aug 2002
    Posts: 13,537
    Likes: 368, Points: 93, Legacy Rep: 2031
    Location: Milwaukee, WI

    gonzo Senior Member

    Motors stall by lacking enough power to overcome the resistance applied. In short, they stop turning. You claim that regardless of the resistance applied the motor will never stall.
     

  15. Will Fraser
    Joined: Feb 2014
    Posts: 158
    Likes: 14, Points: 18, Legacy Rep: 11
    Location: South Africa

    Will Fraser Senior Member

    No, motors stall because they do not have enough torque to overcome a load with a finite static torque.
    A motor operating at its maximum power can still double its torque before it stalls.

    What I am saying is that a generator cannot provide a torque load when stationary, so how can it ever stall the motor? Regardless of the resistance (Ohms) across its terminals, the generator's torque will approach zero as it stops turning. As a motor's rpm drops, its torque increases, so there will always be some finite rpm at which the torque is in equilibrium.
     
Loading...
Forum posts represent the experience, opinion, and view of individual users. Boat Design Net does not necessarily endorse nor share the view of each individual post.
When making potentially dangerous or financial decisions, always employ and consult appropriate professionals. Your circumstances or experience may be different.