Help on my project : need more rpm/speed

[FirePT]

Hi @EmanuelFeru , since I now have everything working i realized that SPEED mode is not enough for my project (a tennis ball throwing machine). Looking into the firmware I see all the warnings around the field weakening LO/HI values and I think you have that warnings because of the normal usage in "vehicle" mode but in my project I will not use it to locomotion but only to drive the ball forward. Can you help me try to archive more rpm from the motors ?

I already changed the default FIELD_WEAK_MAX from 5 to 8 but I think it didn't made any change...

Thanks!

[EFeru]

If you just need fast speed, then I recommend Sinusoidal control type with Phase advance, both selectable in config.h.

[FirePT]

Thanks! I'll try it.

[FirePT]

@EmanuelFeru thank you! It worked, I'm still tweaking the settings but the difference is enormous in rpm.
Currently I have :

// Control selections
#define CTRL_TYP_SEL SIN_CTRL
#define CTRL_MOD_REQ VLT_MODE

// Field Weakening / Phase Advance
#define FIELD_WEAK_ENA 1
#define PHASE_ADV_MAX 45 //from default 25

I have one question regarding :

#define N_MOT_MAX 1000 // [rpm] Maximum motor speed limit

do you think it's safe to increase this value in terms of integrity of the motors/board ?
With the settings I defined above I already concerned by the amount of heat and how quickly it rises in both motor, board and battery... I don't think for now that I can maintain it for more than 1 minute safely. I'm testing a start and stop ball sensor so it just maxes velocity an instant before it have to throw the ball and slows down until a new one triggers the sensor again.

Thanks for your help (and patience) !

[benjaf]

@FirePT If you need more speed without using excessive amounts of power, you might consider modifying the motor to a 'delta' configuration (see issue #26). This trades some of the torque for speed, which seems like a good deal in this case.

[EFeru]

@FirePT

do you think it's safe to increase this value in terms of integrity of the motors/board ?

You can increase that number to 1500 to avoid fixed point overflows. Personally, the maximum I went was around 1200 rpm. It was for a short period, but it felt quite dangerous, and I didn't do more testing to check robustness or temperature, etc. So, I have no idea what will happens on long runs at high speed. If you have a more proper setup, then such tests are possible and interesting to do.

[FirePT]

@benjaf

@FirePT If you need more speed without using excessive amounts of power, you might consider modifying the motor to a 'delta' configuration (see issue #26). This trades some of the torque for speed, which seems like a good deal in this case.

I'm currently reading about the delta configuration and I may try it thank you. Just saw this twitter post and got excited with the possibilities 😄

@EmanuelFeru

@FirePT

do you think it's safe to increase this value in terms of integrity of the motors/board ?

You can increase that number to 1500 to avoid fixed point overflows. Personally, the maximum I went was around 1200 rpm. It was for a short period, but it felt quite dangerous, and I didn't do more testing to check robustness or temperature, etc. So, I have no idea what will happens on long runs at high speed. If you have a more proper setup, then such tests are possible and interesting to do.

Thank you. I'll continue testing and share the results I got.

[FirePT]

Hi,

in the last days I managed to have some time for my project. Changed both motors to delta variant (thanks @benjaf) and the change in rpm is great. Also changed firmware to SIN control type and made a lot of different test with the phase advance 'degrees' value.

Regarding phase advance :
Tested a lot of values (25º, 30º, 55º, 90º) but only the 55º give me almost the same rpm on both motors, other values change either left of right motor velocity almost like when we use direction value different from "0" !! Couldn't find any reasoning for this...

Regarding RPM :
I have ordered one RPM meter but it's taking its time... form china... so I'm left with observations. I manage to throw a tennis ball with 25º inclination resulting in 12 meters long travel before it hits the floor. I got this results with 55º Phase Advance, 2000 Max RPM.

I do not know what else I can try to increase RPM and my goal it's still short of the ~20 meters throw I need. In the twitter post even if it's just one motor the RPM amount is way higher than what I'm getting right now and even if I may not need that amount of velocity I still need more than what I'm getting. I tried to reach out to the person that made the twitter post but didn't got an answer yet.

I'm currently out of ideas : )

[Candas1]

I never understood how they got that speed. I also tried delta, it wasn't as fast as what they got.
I would be interested if you get more details. Jan is in the telegram group, but I never asked him.
They even used the original 10s2p battery.

There are wheels looking similar but with more copper, you could get different speed from different wheels probably.
You could also try a 12s battery you would get even more speed.

[FirePT]

Thanks @Candas1 , I was already set to use a XT60 Y connector to put another 36v battery in parallel so that the strain in the battery is reduced but didn't considered upping the voltage, I may try to use a Bench power supply to fine control voltage and make some tests.

[Candas1]

Just one thought, have you used the maximum speed coefficient?

[FirePT]

Hi @Candas1 I don't believe so... What do you mean by that? How does it map to settings file ?
I've been away since our last talks and wasn't able to further advance my project, that's the reason for all my silence, I think that I will have some free time during my Christmas vacations.

[Candas1]

This is the default speed coefficient:
https://github.com/EmanuelFeru/hoverboard-firmware-hack-FOC/blob/4fbf2549c979eeebf393ff067568cf7e58425570/Inc/config.h#L181

The maximum value would be 32767 ( Higher value would be negative ).
This is also influencing final speed ( if no other bottleneck I guess ).

I am wondering if you would reach higher speed.

[kris86k]

Hi.
Any newses? I need ~3000RPM for small grass cutter.

[Candas1]

3000rpm is not possible with this firmware

[Candas1]

https://github.com/EFeru/hoverboard-firmware-hack-FOC/wiki/Setup-current-and-speed-limits#o-speed-limit

[kris86k]

Can you give me the optimal settings for the max RPM?
It's hard for me to understand.

[Candas1]

What is the point you don't understand ?