Motor braking.

[BrianG]

OK, thanks! That makes sense. I guess I just didn't understand the statement:

Originally Posted by GriffinRU

If controller doesn't open FET's then BackEMF from motor will be rectified by built-in diodes and will be charging battery, if battery voltage will be less then backEMF+voltage drop. You can optimize this process if you open FETs in sync with diodes, then you will cut off diodes voltage drop and losses.
If you noticed braking can be done with only one side, while recuperating involved both.

My confusion stems from the diode aspect. I know the diode is built into the FETs. So, if the FET simply shorts the phases, how can you eliminate the v drop of the diodes?

[GriffinRU]

Originally Posted by BrianG

OK, thanks! That makes sense. I guess I just didn't understand the statement:



My confusion stems from the diode aspect. I know the diode is built into the FETs. So, if the FET simply shorts the phases, how can you eliminate the v drop of the diodes?

Diode conducts in one way only, if you close corresponding FET then you eliminate voltage drop associated with diode. ESC knows exact rotor position and then polarity of the BackEMF, this allows to use FET's instead of diodes to do the rectification.

[zeropointbug]

Simple and short, my opinion on braking.

BL motor braking: (most likely mechanics)

first: a shorted motor has incredible resistance from turning, and, the faster it spins, the more resistance it has.

- Using full motor shorting for braking would undoubtedly flip an R/C car over it's lid with ease.

- So, you need to control this shorting in order to control the motor shaft resistance, to create an average resistance or torque.

-Along came PWM... the controller sets up the FET's to unify all motor phases to create a short. So the controller will change PWM duty cycle depending on throttle position... so anywhere between say 1% to 95% (for ABS, guessing) duty cycle for proportional braking.

-These latest controllers (like the Quark) most likely have some good algorithms for braking.

This is just my simple theory on motor braking, not going into detail what happens in FET's.

:025:

[GriffinRU]

Originally Posted by zeropointbug

Simple and short, my opinion on braking.

BL motor braking: (most likely mechanics)

first: a shorted motor has incredible resistance from turning, and, the faster it spins, the more resistance it has.

- Using full motor shorting for braking would undoubtedly flip an R/C car over it's lid with ease.

- So, you need to control this shorting in order to control the motor shaft resistance, to create an average resistance or torque.

-Along came PWM... the controller sets up the FET's to unify all motor phases to create a short. So the controller will change PWM duty cycle depending on throttle position... so anywhere between say 1% to 95% (for ABS, guessing) duty cycle for proportional braking.

-These latest controllers (like the Quark) most likely have some good algorithms for braking.

This is just my simple theory on motor braking, not going into detail what happens in FET's.

:025:

I like it :)

[zeropointbug]

thanks, i tried :)

I agree with zpb. I'd also like to raise another question on this talk of ABS...

ABS is an abbreviation for Antilock Braking System. By saying antilock one assumes that there is a mechanism in place to detect when locking occurs so that action can be taken to avoid it.

Do ESCs even do this detection?

I have a suspicion they do not, and I also think that they don't really need to.

As ZPB already pointed out, a motor's braking force increases with RPM assuming a constant shorting resistance across its poles. This leads me to think that ABS would be unnecessary in a brushless braking system as the motor itself naturally applies less braking force as it slows down.

The only time I can see ABS being of value is if there were devices monitoring the rotation of the wheels themself so that braking force could be released if only one wheel locked (while the diff action kept the motor spinning).

What do you all think?

[GriffinRU]

Originally Posted by Aragon

I agree with zpb. I'd also like to raise another question on this talk of ABS...

ABS is an abbreviation for Antilock Braking System. By saying antilock one assumes that there is a mechanism in place to detect when locking occurs so that action can be taken to avoid it.

Do ESCs even do this detection?

I have a suspicion they do not, and I also think that they don't really need to.

As ZPB already pointed out, a motor's braking force increases with RPM assuming a constant shorting resistance across its poles. This leads me to think that ABS would be unnecessary in a brushless braking system as the motor itself naturally applies less braking force as it slows down.

The only time I can see ABS being of value is if there were devices monitoring the rotation of the wheels themself so that braking force could be released if only one wheel locked (while the diff action kept the motor spinning).

What do you all think?

The key element of ABS is to keep track on traction (maintain traction), have nothing to do with RPM

[zeropointbug]

Yeah, it's not true ABS (like the Quark), it's just the controllers algorithm controlling it. It knows how fast the rotor is spinning, etc.... there is a minimum rpm the motor can spin when shorted anyways.

It is quite simple how it would 'detect' the rotor rpm, it's just a matter of making a good algorithm to do some braking magic. I imagine that controllers only use a fraction of the motor braking resistance.

[GriffinRU]

Originally Posted by zeropointbug

Yeah, it's not true ABS (like the Quark), it's just the controllers algorithm controlling it. It knows how fast the rotor is spinning, etc.... there is a minimum rpm the motor can spin when shorted anyways.

It is quite simple how it would 'detect' the rotor rpm, it's just a matter of making a good algorithm to do some braking magic. I imagine that controllers only use a fraction of the motor braking resistance.

Of course controllers do simple things, we do not have individual sensors and brakes per wheel, but idea is the same.
And check your last statement, because it is the motor which stops the vehicle not the ESC or Battery

Originally Posted by GriffinRU

The key element of ABS is to keep track on traction (maintain traction), have nothing to do with RPM

Yes, but avoiding a wheel lock (or skid) is a key part of maintaining traction. Static frictional force is always greater than kinetic frictional force given the same two surfaces...

Originally Posted by GriffinRU

And even with the last one it would be impossible with eagletree's resolution to see anything useful...

What resolution do you think would be necessary? My DPR can sample at upto 20 times per second (I think), and it is capable of measuring current flow in both directions...

[zeropointbug]

Originally Posted by GriffinRU

Of course controllers do simple things, we do not have individual sensors and brakes per wheel, but idea is the same.
And check your last statement, because it is the motor which stops the vehicle not the ESC or Battery

I know that, I was stating that it's up to the controller to do good braking, as many ppl here know that different controllers have much different braking characteristics.

[zeropointbug]

The software engineers of the controllers are putting 'assumptions' into the algorithms, if you catch what I mean? It's hard to explain this. But also, maybe it's monitoring current levels, and there is a max?

A simple braking alg. would prob do some pretty decent braking effect. But who knows how sophisticated they are, or maybe they are simpler than one might think?

[GriffinRU]

Originally Posted by Aragon

What resolution do you think would be necessary? My DPR can sample at upto 20 times per second (I think), and it is capable of measuring current flow in both directions...

Much faster, at least 2 times faster then number of PWM steps times switching frequency.

[GriffinRU]

Originally Posted by Aragon

Yes, but avoiding a wheel lock (or skid) is a key part of maintaining traction. Static frictional force is always greater than kinetic frictional force given the same two surfaces...

You lost me here, or smth else