Questions about VESC for FFB application

General topics and discussions about the VESC and its development.
pf26
Posts: 166
Joined: 28 Mar 2016, 14:37
Location: FR Valence

Re: Questions about VESC for FFB application

Postby pf26 » 18 Dec 2016, 10:17

"The RMS value is the effective value of a varying voltage or current. It is the equivalent steady DC (constant) value which gives the same effect. For
example, a lamp connected to a 6V RMS AC supply will shine with the same brightness when connected to a steady 6V DC supply." (from http://www.raeng.org.uk/publications/other/8-rms

Ok for the belt (i guess something like timing belt T5 or so). Then you could even go for a higher ratio (up to 4 or 5) and use a smaller motor.

Summing up the motor phases currents would result in an oversized power supply, because the motor current is used at a lower voltage than the power supply (unless the motor turns at full speed). Maybe you can calculate the power supply requirement from the mechanical power you need: For instance 30N at max 2m/s -> this is 60W, say with an efficiency of 80% -> 75W need , so less than 2Amps at 48Volts.
To be on the safe side, and because PS are not that expensive, you could select a 150W/48V power supply for single motor, or 250W/48V for 2 motors.

Oconnor
Posts: 11
Joined: 09 Dec 2016, 13:34
Location: France

Re: Questions about VESC for FFB application

Postby Oconnor » 20 Dec 2016, 23:46

To be precise about the belt, i'll use HTD3m Belt, they are way better than T and AT belt for "high" torque application with less backlash.
The issue with ratio like 4 or 5 is the pulley size for the larger one and the low count of engaged teeth on the smaller one (risk for the belt to jump teeth).


Since i'm looking for max torque capability, why not choose the power supply based on the max current ? For exemple if my motor need 5.8A rms at holding torque, i'll need 2x5.8 =~12A so around 600W for the power supply (without counting margin).

rew
Posts: 576
Joined: 25 Mar 2016, 12:29
Location: Delft, Netherlands.

Re: Questions about VESC for FFB application

Postby rew » 21 Dec 2016, 09:12

You only need that amount of power when you are moving your motor at "max speed". I (almost?) calculated the max speed for someones motor last week. But with a "force feedback" application you are probably not going to need that max speed.

Note that you can easily get 12A in your motor with a 3A powersupply. As long as you're not going faster than a quarter of the max speed of the motor at 48V.

Oconnor
Posts: 11
Joined: 09 Dec 2016, 13:34
Location: France

Re: Questions about VESC for FFB application

Postby Oconnor » 22 Dec 2016, 12:02

Ok, i've got it from a power point of view. But i what i don't understand, is how a power supply can provide 4 times it's rated current ? :oops: even if the motor spin at 1/4 speed (=1/4 of the rated V), how it works ?

As a mechanical engineer sometimes electrical and electronis things are a bit strange or abstract. :mrgreen:

rew
Posts: 576
Joined: 25 Mar 2016, 12:29
Location: Delft, Netherlands.

Re: Questions about VESC for FFB application

Postby rew » 25 Dec 2016, 09:44

To put it in mechanical terms... it's like a gearbox. You can convert high voltage low amps to lower voltage, more amps, just like you can convert high RPM, low torque to lower RPM more torque. Some losses in both cases, but otherwise the same thing.

In more technical terms, the motor acts as a coil, an inductor. The current changes depending on the voltage. Put a higher voltage on a resistor and the current becomes higher immediately. With an inductor, the current starts RISING faster.

So assuming some numbers... we assume the battery has 40V, and we are at 10V effective. Then we are PWM-ing 1/4 on 3/4 off. During the on-time current through the inductor rises with 30V (= 40V - 10V) times 1 time-period. During the off-time the current goes back down again 10V times 3 time-periods. The product is the same, the current, stays mostly the same. (it goes up and down a bit, but it doesn't grow more every cycle).

Anyway, you end up taking the full current only 1/4th of the time. so that averages out to 1/4 the motor current. In practice you need some capacitors to charge using that 1/4th the current during 3/4th of the time and then they discharge at 3x the current 1/4th of the time (the last 1x is still being provide by the battery).

Oconnor
Posts: 11
Joined: 09 Dec 2016, 13:34
Location: France

Re: Questions about VESC for FFB application

Postby Oconnor » 11 Jan 2017, 15:54

Thanks for the explanation rew.

I finally found a motor for my use, in the size i want and with an encoder (1000ppr).
It's this one but with an encoder at the back : Nema 34 3phase stepper (4.5N.m).

I thought about the regenerative current flowing back to the power supply. Can i dissipate it in a resistor with a couple diodes to protect the power supply and force the current into the resistor ?

rew wrote:In practice you need some capacitors to charge using that 1/4th the current during 3/4th of the time and then they discharge at 3x the current 1/4th of the time (the last 1x is still being provide by the battery).

Ok, so i'll need some capacitor between the power supply and the VESC, that's why most VESC pcb sellers sells them with a capacitor board ? But how to choose those, are they included in the BOM of the VESC ?

rew
Posts: 576
Joined: 25 Mar 2016, 12:29
Location: Delft, Netherlands.

Re: Questions about VESC for FFB application

Postby rew » 13 Jan 2017, 11:20

Genrally people seem to use about 2000uF (3*680).

I stumbled upon:
https://www.aliexpress.com/item/Leadshi ... 7083d01f99
which is rumored to be even stronger than yours....

Dissipating the extra energy when there is regneeration is not trivial if you want to be smart about it.

If you expect max 10W of regeneration you can calculate a resistor for 10W and put it across your powersupply. You'll be wasting 10W when there is no regeneration.

If your powersupply is 40V nominal you might be able to connect the power supply to a zener and a resistor (in series). The other side goes to the base of a transistor that switches on the dissipating resistor. With 48V nominal I'm not sure if this will properly limit things below the 60V where VESCs start smoking.

Oconnor
Posts: 11
Joined: 09 Dec 2016, 13:34
Location: France

Re: Questions about VESC for FFB application

Postby Oconnor » 24 Jan 2017, 12:25

I don't why your is better than mine, it have about 4.5 times less torque ;) but maybe i missed something.

I don't mind losing few Watts if i avoid to blow my power supply :), so i think the easiest solution is to have a diode to protect the power supply form reverse current and between the diode and the VESC capacitor a resistor in parrallel.

In my design i have a dilemma, is a better to have VESC very close to the motor (few cm) with a remote power supply (like 1/1.5m away from the VESC) or to have the VESC far from the motor with the power supply close to the VESC ?

rew
Posts: 576
Joined: 25 Mar 2016, 12:29
Location: Delft, Netherlands.

Re: Questions about VESC for FFB application

Postby rew » 24 Jan 2017, 22:28

Ooops. It seems I miscalculated the torque on one of the motors. Indeed impressive!

If you prevent current into the power supply, you will only be charging the capacitors near the VESC when regenerating. When you hit 60V, the VESC goes up in smoke. Not sure if that's a good plan.

In general, the inductance between the battery and the ESC is difficult to deal with. And the inductance between the ESC and the motor is good. There is a needed inductance in the motor coils already. So I would prefer short battery leads and long wires to the motor.

The downside is that you're emitting some high frequency EMI from the motor wires...

Oconnor
Posts: 11
Joined: 09 Dec 2016, 13:34
Location: France

Re: Questions about VESC for FFB application

Postby Oconnor » 01 Feb 2017, 12:41

I thought about cable length and VESC position. It will be way easier to have the VESC close to the motor. This way i only need 2 or 4 wires from the "command/power box" to the joystick instead of 2x6 + 2x4 + 2 (Motors + encoders + encoder power).
And with the VESC into the joystick base i can use the 5v output to power sensors and other thing i might need to make it work.

To protect the PSU form the regenerating current i thought to this, i don't know if this is stupid or not :

Schema1.JPG
Schema1.JPG (16.94 KiB) Viewed 151 times


I don't know if this can work. Can i reduce the regenerative current from the VESC ? May be there is a way to totally suppress the regeneration ? it will be great


Return to “General”

Who is online

Users browsing this forum: No registered users and 1 guest