You are here

VESC6 and CableCam

15 posts / 0 new
Last post
wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65
VESC6 and CableCam

 

To give you an impression, this is how my CableCam behaves with the VESC6.

https://youtu.be/1NV-bPIsbX0

 

Some bullet points

  • VESC is configured in speed mode and gets from a CC3D revo the speed commands via UART
  • This controller has a bluetooth module on one of its UARTs and with that you can control it, the VESC and read the VESC status
  • One area the VESC performs beyond expectation is the speed control. I have speed ramps programed and with a normal ESC this controls the thrust/energy. Result is an uneven acceleration. Thanks to the speed mode of the VESC however the acceleration/braking is linear and very smooth.
  • Problem I have is at speed 0 and on steep hills. How can we make the VESC to hold its position? Not an issue when the rope is level as in this video.
  • Ronin-M Gimbal is quite heavy and the VESC accelerates more than enough with that. Quite scary in fact and that with 3s LiPos only. 6s would be more appropriate. 
  • Very silent. Inaudible in fact. The T-belt gear makes most of the noise.
  • Slow speed movement beautiful with the brushless motor hall sensors. 

Summary: I love it. Absolutely worth it.

 

 

Davideo
Offline
Last seen: 1 year 8 months ago
Joined: 2017-09-19 15:38
Posts: 21

Cool! I'm working at something with similar properties (even if it's not quite a cable cam). Battling some of the code for the ramping with known acceleration. When you say 'speedmode' do you mean you're working with the RPM of the VESC or is this a new setting on the VESC6 (i'm using an older VESC at the moment, so not too familiar).

 

Keep up the great work!

 

D

 

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

I was talking about the closed loop (PID) mode where you specify the eRPMs, yes. Nothing VESC6 specific afaik.

frank
Offline
Last seen: 6 days 3 hours ago
VESC BronzeVESC FreeVESC GoldVESC OriginalVESC PlatinumVESC Silver
Joined: 2016-12-27 20:19
Posts: 846

Handbrake is on the list. It is also a feature many other applications require. I will chat about it with Benjamin.

Frank

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

Thanks, greatly appreciated!

benjamin
Offline
Last seen: 1 week 1 day ago
VESC FreeVESC OriginalVESC Platinum
Joined: 2016-12-26 15:20
Posts: 490

Very nice project, thanks for sharing!

As you noticed, the speed controller will not hold a position at 0 speed when there is load torque. To hold a position there are a few ways:

1. Use position control mode. This works best with and encoder, but with some PID parameter tweaks it could work with hall sensors as well.

2. Use the open loop handbrake function when the speed is low enough. This function will apply a current vector with a magnitude you set in open loop, which can hold the motor still when there is some load. If the current is high the motor will get a bit warm though. You can test this one by using the handbrake function in VESC Tool, or by sending a handbrake command over UART.

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

Thanks Benjamin, as I wrote in the other thread (https://vesc-project.com/comment/1364#comment-1364) the handbrake does not have enough momentum. And even if I would use a value as low as 10A, the 5000mAh battery would be drained within 30 minutes.

Hence I am waiting for what you can come up with to support a true PID=0rpm feature.

josh
josh's picture
Offline
Last seen: 2 years 3 months ago
VESC Original
Joined: 2017-05-24 12:15
Posts: 92

This is way cool! I've always wanted to build one of these. Thanks for sharing!

frank
Offline
Last seen: 6 days 3 hours ago
VESC BronzeVESC FreeVESC GoldVESC OriginalVESC PlatinumVESC Silver
Joined: 2016-12-27 20:19
Posts: 846

Probably it would be best to have a mechanical brake at the cable, which doesn't need a lot of power, servo driven. Something that clamps against the cable and locks in, so no power is needed at all, once locked down.

Frank

 

benjamin
Offline
Last seen: 1 week 1 day ago
VESC FreeVESC OriginalVESC Platinum
Joined: 2016-12-26 15:20
Posts: 490

And even if I would use a value as low as 10A, the 5000mAh battery would be drained within 30 minutes.

That is by far not true. To achieve 10A at the motor with 0 speed, you probably need less than 5% modulation on the motor (depending on some parameters, I can calculate the exact value if I have the motor resistance and battery voltage). That means that the battery lasts more than 10 hours in that case, not less than 30 minutes.

What you have to keep in mind is that holding torque always requires power, even in closed loop. The closed loop control will just end up using the lowest power that is just enough. If you, for example, experiment with the hand brake function an by trial and error determine how much current you need to have just enough to hold the motor, this current will be very close to what you would get with closed loop control. This is why many applications where holding torque is required for a long time use mechanical brakes, just as Frank said.

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

Okay, then I am more positive and will look into the handbrake more closely again.

I know a mechanical brake is by far the best. It would also serve as a backup in case battery dies, motor went up in flames or whatever. But it is an awful complicated thing to do, increasing the costs.

On a side note, if you could investigate low speed movement with FOC, Hall and PID once you implement the PID=0 feature, that would be appreciated. I assume it is the same logic and currently it does not behave well due to the large duration between Hall signals. I am thinking about giving up PID logic at speeds < min-eRPM = 120 and rather calculate the amount of time between hall sensor signals and us that as vector speed.

 

I have implemented in the controller a logic to change the handbrake amps dynamically and to set ranges, hence I will put the cablecam on a steep hill and try it out. Keep you posted!

willyfan
Offline
Last seen: 4 years 5 months ago
Joined: 2018-01-04 14:14
Posts: 7

Hi Werner, a question: iI have at home a old VESC (not VESC 6, maybe 4.12) do you think that it work or are you using some VESC 6 spercific settings?

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

One of my questions as well. Will it work? Yes, it is the same firmware, has the same capabilities. I have a HW4.10 myself and it does work.

Reasons I went through the expenses and bought the VESC6 are

  • Nice casing
  • Higher Amps
  • unknown better behavior in FOC mode

The latter I keep hearing about but actually don't know what it is meant to be precisely. Maybe somebody else can chime in?

Personal Summary: If you have the old hardware already, give it a try. Nothing wrong with it.

 

wdaehn
Offline
Last seen: 2 years 5 months ago
Joined: 2017-09-12 17:26
Posts: 65

Speaking of which: I cranked up the power supply from 3s to 6s, now the handbrake is strong enough. Braking works beautifully from high speed downhill to a standstill by reducing the speed via the ramp down to zero, then wait until the tacho_abs shown almost no movement and then engage the handbrake.

There is some unevenness in the last step as the FOV vector can point in any direction and the handbrake has specific vectors only but that is not so important. Transition from zero to drive is nice.

Summary: Looking forward to the speed=0 support in the VESC. On my list the most important feature by far.

willyfan
Offline
Last seen: 4 years 5 months ago
Joined: 2018-01-04 14:14
Posts: 7

If every VESC work, I will test my 4.12 in the next days (or just I can test outside, now is snowing a lot...). Cheaper that VESC 6 but in a good case there is the FocBox, suitable for install on a cablecam (and half price). VESC 6 hava the little advantage of amperes (80 instad 60, anyway better that usual 4.12 hardware that is 50 A) but I think that this is not a big problem.