Electric Paragliding Winch Design Questions

[C_Marvin]

Hi all!

New here and new to BLDC vehicles...Hopefully I'm posting in the right place!

I'm working on designing a "pay-out" paragliding winch that will use regenerative braking to control tension.

Main setup for a paragliding payout winch:

• Lots of line is wrapped around a drum, attached to the back of a car
  
• The end of the line is attached to the pilot
  
• the car starts driving, then increases tension on the line to a pre-set tension, towing the glider into the sky (the tow up takes about 6-7 minutes)
  
• at the end of the tow, the pilot disconnects and a drogue chute inflates at the end of the line
  
• the motor on the winch then rewinds the line
  
• video link:https://www.youtube.com/watch?v=qMrnE9xFzgw
  

Winch design:

• I'm planning on using a qs138 70H paired with a Kelly KLS7230H
  
• The winch drum will be an average of ~15" in diameter, so we're using about a 9:1 gear reduction to spin the motor at ~1600RPM during the payout, placing the motor in the "efficient" range of RPMs per the performance curve
  
• during payout, the maximum line tension that will be created by regen braking is ~60kg, translating to ~12Nm of torque at the motor with the gear reduction
  
• We plan to dump most of the regenerative braking energy into a resistive heating element with air cooling to avoid needing a huge battery that can handle the regen currents. Then we're planning on just using a couple 80v 4Ah power tool batteries in parallel for rewind.
  

My main questions:

• If the pilot screws up and starts flying the wrong way on tow, the tow operator will need to remove all tension from the line, and the motor may end up spinning quite fast as the pilot pulls the line out (faster than the rated no-load speed of ~6000RPM of the qs138) ...Since motor RPM is tied to voltage via the kV rating, what will happen in this scenario? Will the motor controller burn out? will the motor overheat? will the motor start applying force to the line even if we drop the regen braking to zero?
  
• Anything I'm missing here with this design? Any design points I'm missing?
  
• I would love to do a belt drive, but I can't find a belt pulley with enough teeth to give me at least a 8:1 gear ratio...Any hints on where to find one, or do people just diy them?
  

Thanks so much in advance! I've already learned so much just browsing through all the information and knowledge posted here on the forum

 

[amberwolf]

If you haven't already seen them, there are some posts about making paragliding winches:
https://endless-sphere.com/forums/search.php?keywords=paraglid*+winch*&terms=all&author=&sc=1&sf=all&sr=posts&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

We plan to dump most of the regenerative braking energy into a resistive heating element with air cooling to avoid needing a huge battery that can handle the regen currents.

Does this mean you will have a 3-pole switch (or heavy-duty connector to plug/unplug) to connect the motor phases to the resistors (three of them, one for each phase?) or to the controller phase outputs, as needed?

Or are you going to switch the controller's battery input between the batteries and the resistor(s)?

(you can't connect the resistors across the battery input without disconnecting the batteries or you will drain them).

If the latter, keep in mind that unless the resistors are such a low resistance that they really load down the output the way a battery would, the voltage developed across the resistors will become higher than the controller can handle, and the FETs will blow up.

If the resistance is too low, and current isn't sufficient from the motor thru the controller, it won't develop enough voltage across it to keep the controller on, and if the controller shuts off during operation, it may not restart in a usable mode without stopping the motor first, and stabilizing the voltage at the battery input, and removing all control inputs (no braking, no throttle). Depends on the controller design, but it's a safety feature to "ignore" all inputs and/or halt reset after a power failure, to prevent runaway operation.

Regen may still continue if the motor is spinning fast enough to generate a voltage higher than the battery voltage: in this case you'll get regen current because the FETs will act like a 3-phase rectifier. (even if the controller is not powered on at all).

If you are switching between battery and resistors on the input of the controller, and the motor is spinning at this time, the generated voltage may spike with no load on the controller's battery inputs, and this could blow up the FETs too.


Motors behave differently when above their normal no-load RPM. I've forgotten what typically happens. You may be able to experiment with this here: http://ebikes.ca/tools/simulator.html with various motors and systems, but I can't remember if it allows this situation (you may be able to "force" it; instructions are below the graph.)


There was something else I was going to note, but I forgot what it was.

 

[liveforphysics]

If it's life safety-critical that the winch motor maintains operational in use, it might be best to run a better motor controller, I've had bad luck with sustained high torque operation with Kellys.

 

[C_Marvin]

If you haven't already seen them, there are some posts about making paragliding winches:
https://endless-sphere.com/forums/search.php?keywords=paraglid*+winch*&terms=all&author=&sc=1&sf=all&sr=posts&sk=t&sd=d&st=0&ch=300&t=0&submit=Search

Yep, read through those! Definitely some very useful stuff in there. one of the users "KrisH" is a buddy of mine. He's had decent luck with his winch, but since the commercial electric winch he bought uses a hub motor with a large spool, it is spinning in a really inefficient range of RPMs during payout, and the motor gets quite hot on long tows (resulting in low waiting times between tows for the motor to cool off). Hence why I've decided on a mid-drive motor with proper gearing for efficient RPMs to hopefully avoid the heat issue. He also never implemented the whole resistive heating thing, and just went ahead and bought a ~$1400 battery to simplify the solution.

If the resistance is too low, and current isn't sufficient from the motor thru the controller, it won't develop enough voltage across it to keep the controller on

I should have clarified this in my post above...Slipped my mind! It looks to me like the Kelly controllers have a separate pin from the B+ input for powering just the controller:

My hope is that I can use a completely separate battery for powering the controller to keep it from shutting off. Per the manual the PWR input can be 40-90V at the PWR pin. That being said I don't know if the controller will be happy with different voltages between the B+ and PWR pins. I'll have to contact Kelly, unless somewhere here knows if that will work...

That way I can just use a contactor/switch to switch from the resistive load and batteries so that I'm not draining batteries through the load constantly.

keep in mind that unless the resistors are such a low resistance that they really load down the output the way a battery would, the voltage developed across the resistors will become higher than the controller can handle, and the FETs will blow up.

For a cheap solution, I was planning on using the heating elements from a few used radiant 1500W space heaters in parallel. I'll have to measure the resistance of one to see if they'll work (napkin math says they should be around 10 ohms here in the USA). If that won't work I guess I can just make a big bank of purpose-built power resistors in parallel?

If it's life safety-critical that the winch motor maintains operational in use, it might be best to run a better motor controller, I've had bad luck with sustained high torque operation with Kellys.

Interesting...Can you expand on that? I've seen one other (commercially available) winch using a kelly controller, hence why I chose it.

• During the towing phase, it would be unpleasant (but not immediately deadly) if the controller suddenly dropped the torque created by regen braking
  
• During the rewind phase, if the motor totally cut out, it would zero danger, just annoying to retrieve all the line that fell all over the place when the motor shut off.

 

[amberwolf]

It looks to me like the Kelly controllers have a separate pin from the B+ input for powering just the controller

AFAIK, that PWR is only powering the low voltage power supply for the brain; it doesn't provide power to the drive section. (presumably the LVPS does provide power for the gate drivers, but not for the FETs, etc).

It is intended to be used to turn the controller on or off while leaving the main B+ connected (such as for a reset, or just to leave the system on but in a low-power-usage "off" mode).

It's a pretty common thing in controllers; often called KSI (key switch input), or "door lock" (no idea where they get that from), etc.

My hope is that I can use a completely separate battery for powering the controller to keep it from shutting off. Per the manual the PWR input can be 40-90V at the PWR pin. That being said I don't know if the controller will be happy with different voltages between the B+ and PWR pins. I'll have to contact Kelly, unless somewhere here knows if that will work...

That way I can just use a contactor/switch to switch from the resistive load and batteries so that I'm not draining batteries through the load constantly.

HOpefully it will work the way you want...it depends on the "smarts" of the controller (the smarter they are, the less likely they are to "like" situations like this because of safety cutouts, etc.; dumb controllers don't ask questions but instead just do what you tell them...to destruction if you ask that :lol: ).

But you shouldn't really need a separate battery for the PWR input; just don't turn that one off when you turn off the FET supply (B+).

For a cheap solution, I was planning on using the heating elements from a few used radiant 1500W space heaters in parallel. I'll have to measure the resistance of one to see if they'll work (napkin math says they should be around 10 ohms here in the USA). If that won't work I guess I can just make a big bank of purpose-built power resistors in parallel?

You'll probably have to experiment with it to find out how much drag you get vs how much resistance is used. You could have several banks, and switch them into parallel with the first one(s) to increase drag if it's not enough. The lower the resistance, the higher the drag, because it allows (forces) more current flow out of the controller/motor. I'm sure there's a way to calculate it, but that kind of math is beyond me figuring it out (I have enough trouble with long division).

Remember that is 1500w at the AC wall voltage they're used for. Meaning, if your output voltage is not that high, you won't draw as much power out of the motor (and not create as much drag).