• Howdy! we're looking for donations to finish custom knowledgebase software for this forum. Please see our Funding drive thread

retractable propulsion system for gliders

highlander42

10 µW
Joined
Nov 23, 2023
Messages
5
Location
Europe
Hey,

I testing at the moment a Sevcon Gen 4 Controller for an e-propulsion system to convert old gliders.

Now I had one point missing and no idea how I could solve it with the Sevcon.
To retract the motor in the fuselage I had to reposition the propeller.
in most combustion systems it was made with a mechanical stop while the propeller was windmiling.
with the electric motor, I thought there must be a better way.
I already had a fourth hall sensor to check if the prop is in the right position.
Perhaps with this sensor and an MCU what controls breaking depends on the RPM so it could stop the motor at the exact position.

perhaps somebody has a good idea or I do overlook some functions the Secvon can do?
thanks.
 
If you can apply just enough throttle input (using the external MCU, etc) to counter the windmilling and very slowly move the prop until the position sensor triggers, then (using the MCU) send the brake signal to the sevcon, with the sevcon setup to use battery current to hold and lock the motor in position instead of driving it around.

You could build a mechanical setup with passive electronics to do the same job as the MCU, but it's easier to adjust the MCU's programming than reworking a mechanical setup.


BTW, this thread
also discusses a stowing function, if it's helpful.
 
Thanks. The project with the Ac5 I saw. Realy cool project.
with my electric motor i do not have windmilling because the cogging torque of my motor is so high.
and because of this I do not get really low speeds. Below 500 RPM it’s a really rough and inconsistent run.

So I think I have to brake down the motor from about this speed with a time variable and brake variable depends on the last RPM before I send the stop signal.

before the sevcon it tested Kelly controller.
with them I send short throttle impulse so it was possible to jump step by step with the motor till the prop was on the right position.
but with the closed-loop control of the sevcon it’s not possible.
With the kelly controllers I was no happy because I had a lot of emc problems. Radio communication was not possible.
 
You should be able to "pulse" the motor current (torque) short enough to make it move just one pole at a time just like the kelly; I don't know exactly how to do it, but it should be possible; I'd be surprised if it isn't. (but the Sevcon has a lot of stuff in it that might possibly prevent it from being easy).

If it isn't possible you might have to use a mechanical rotate/locking system. Or use a contactor setup to disconnect the Sevcon's phase output from the motor, then use a secondary electrical control to rotate it to line up and lock (but this has the risk of contactor failure that could let the secondary system that shorts the Sevcon phases.
 
With pulses would be the easiest way. For this I have already a program.
So I would have to do some try and error and test if I find pulses with length and intensity they will work.
I think would be hard with the soft start and closed-loop control of the sevcon.
Build a mechanic system is one of the last options.
Also a secondary electric control would made the system more complex. And for me the goal of and electric propulsion against a combustion system is to made it a simpler system.
However here you put more complexity in the software.
 
Does the sevcon *have* to use soft start? Or does it have an option for non-soft start?

Similarly, does the sevcon have the option to use one of it's I/O lines to change from closed loop to open loop and back?
 
After trying a lot of different pulses I found no solution till yet.
either it doesn't even jump a pole and only wobbles a little, or it starts to turn fully after a while.

i have seen that i can switch off some safety functions in the DVT. now I just have to find out if I can do this via an i/o.

or does anyone here have experience with sevcon controllers and can give me a hint?
 
With pulses would be the easiest way. For this I have already a program.
So I would have to do some try and error and test if I find pulses with length and intensity they will work.
I think would be hard with the soft start and closed-loop control of the sevcon.
Build a mechanic system is one of the last options.
Also a secondary electric control would made the system more complex. And for me the goal of and electric propulsion against a combustion system is to made it a simpler system.
However here you put more complexity in the software.
You should resort to a mechanical brake and stop. Even only a brake will do if you can operate it quickly enough (e.g. with a servo). It can be fixed to the engine mount similar to the brakes at the motorgliders with combustion engine.
The advantage of a mechanical solution is that it will work even if the electrical drive fails. In that case the prop will start to windmill and you have only 2 choices: land with extended and windmilling prop (bad glide angle) or break the prop when trying to retract.
 
How about a bicycle hydraulic disc brake and a small mirror in the top of the canopy so you can see it? If retraction is done manually with a lever maybe mount the brake lever to that. Stick a 1/8" pinstripe onto the canopy length-wise above your head as an aid to aligning the prop.
 
Last edited:
60-2 variable reluctance ring...... trigger on the rising edge of the gap? Or maybe a LED window and a photosensor. How timing is done in modern automobile chassis. many commonly available off shelf parts. Or a small servo driver i would say with an encoder. feedback loop closed system with a hysteresis adjustable. to 'tune ' it. Tons of very accurate servo motor systems, linear actuators, glass scale based even..or a capacitance strip... let alone available in all sizes... simple inputs.

IDK. Interesting design challenge. how many RPm are we talking? i didnt happen to see that.
 
If you can apply just enough throttle input (using the external MCU, etc) to counter the windmilling and very slowly move the prop until the position sensor triggers,


i dont think propeller efficiency, and load, is linear with altitude so it would need a compensation. For velocity and altitude. real time. or dataAQ for all that jazz to be heard.

IDK how how airplanes fly though. im more of a helicopter guy. lol.

Stepper motor on a resistance braking circuit with a 1.8 deg/step and a 1:30 reduction. lol. Stop that thing right where you want it toto +/- 10 arc min.... H.... ell use G-Code and a 3$ GRBl on an Uno.

( Ok so here is a JP edit: this is where Im getting all this brainstorming idea from.. Actively stopping the spindle is an important part of CNC programming when your threading holes. "The "G84" cycle is intended for tapping. This cycle rotates the spindle clockwise to tap a pre-drilled hole; when the bottom of the hole is reached, the spindle rotates in the reverse direction and exits the hole." ..... the High RPm spindle MUST Stop when it gets to the bottom of the hole.. the absolute correct place... no room for error in the RPm halt OR the depth on the axis... etc... and they do it with servos and encoders on teh spindles. All in milliseconds at megahertze update rates.

Ok Jp edit done. ).


bahahahah.
 
Last edited:
I'm sure you're right; I know a fair bit about aircraft / flight stuff in general, but not much detail, specifics. I was really just throwing out a general idea that the OP would have to work out the details of. ;)
 
Back
Top