Regen braking on cheap ESC

[akio123008]

Hello everyone, 

I've built an electric bicycle which is powered by a brushless motor esc setup on the front wheel. A schematic looks like this:

 

The ESC I use is a fairly cheap, though quite beefy "Red Brick OPTO" type that I got from a Chinese website. The setup works just fine, and the motor can drive the bike without any problems.

However, there's something I'm quite curious about: When the ESC is turned on, and the throttle (potmeter) is closed, meaning the motor isn't being powered, it's quite hard to turn the motor manually.

Of course that's some kind of "motor braking" going on. What I want to know is is this braking regenerative? That is, does the ESC charge the batteries when an external force turns the motor? 

 

As I've mentioned, the ESC is fairly cheap and came with some extremely minimal documentation, and I couldn't find the answer online, so I'm hoping that someone out there has more experience with this and knows what's going on here.

[TechyBen]

I would assume not. Is it just resistive current on the wires when the circuit is closed?

 

I'm not sure if you can do regenerative breaking easily on a digital/brushless motor.

 

https://www.evelo.com/blog/why-dont-more-bikes-use-regenerative-braking/

 

I wonder if you could put a seperate dyno on. One that usually only powers lights etc, and apply it as the breaking arm to a front or back wheel. XD Would not brake as well though, but could use it to get power back when coasting (slowly) down hill if you live somewhere with lots of hills. (You would slow the cycle down, but get some power back only when going down hill)

[akio123008]

26 minutes ago, TechyBen said:

I would assume not. Is it just resistive current on the wires when the circuit is closed?

Well yes that's what I thought too, but it only does it when the ESC is powered, and in a certain mode; I've been screwing with the settings of the ESC, and in some modes it doesn't brake at all. 

 

26 minutes ago, TechyBen said:

I wonder if you could put a seperate dyno on. One that usually only powers lights etc, and apply it as the breaking arm to a front or back wheel. XD Would not brake as well though, but could use it to get power back when coasting (slowly) down hill if you live somewhere with lots of hills. (You would slow the cycle down, but get some power back only when going down hill)

Yes that's perfectly possible, but it'd be cool if it could be done through the existing motor. I actually never intended to make the thing regen capable, but I just kind of stumbled upon the possibility when I noticed the motor was hard to turn when the ESC was powered.

 

Also fun fact, the motor actually drives the wheel by rubbing against it, just like a bike alternator, it's an outrunner motor with the outside covered in rubber, which is then pushed onto the tyre using a spring mechanism. The reason I want to know if it regen-brakes, is because if it doesn't, I will make some kind of "clutch"  mechanism to lift the motor from the wheel, to allow for smooth coasting. 

[TechyBen]

Yeah. I don't think it's entirely "impossible" to get regen on a brushless motor, but you would be looking at more complexity in the controller and wiring setup than is probably possible on a bicycle.

 

Basically, you'd have to alternate the (3) phases correctly, in reverse, convert it back to DC and dump the waste heat from your converter. Not worth it for the size/weight/complexity. One way battery -> controller -> motor is fine, as it's driving the motor, and deciding the speed etc. But doing it in reverse is harder. Plus depending on the type of motor, it may not be designed to be efficient in reverse or for generation.

 

The resistance depends on if the circuit is open or closed (I don't remember which off the top of my head.  ).

[akio123008]

34 minutes ago, TechyBen said:

Yeah. I don't think it's entirely "impossible" to get regen on a brushless motor, but you would be looking at more complexity in the controller and wiring setup than is probably possible on a bicycle.

 

Basically, you'd have to alternate the (3) phases correctly, in reverse, convert it back to DC and dump the waste heat from your converter. Not worth it for the size/weight/complexity. One way battery -> controller -> motor is fine, as it's driving the motor, and deciding the speed etc. But doing it in reverse is harder. Plus depending on the type of motor, it may not be designed to be efficient in reverse or for generation.

 

The resistance depends on if the circuit is open or closed (I don't remember which off the top of my head.  ).

True, but there are certainly ESCs that do it, for example, a guy named Tom Stanton posted a whole series of videos about an electric bike he made, and that thing does regenerative braking. Now he does use an expensive ESC with more options than mine, but I thought that since it is clearly possible to have regen braking support in a tiny ESC, this may also be the case on mine. I don't think so, because it's a fairly cheap model, but I guess I'm just going find out by experimenting. Not sure what's the best way of measuring this though, battery voltage could be an indicator, but that would mean a lot of charging needs to be done, as the voltage goes up by only small amount on these li-ion cells.

[TechyBen]

AFAIK it needs physically different circuitry. So doing it on a cheap on may not be possible. You need the physical outlets and at least the chip programming options too.

 

If you just experiment, it risks blowing a cap/diode/fuse etc in the ESC.

 

As in. Why are you worrying about measuring the battery voltage? You have to first confirm you can at least tap a single phase (and thus it's output voltage/current), let alone all 3 if using a rectifier/converter.

[akio123008]

16 minutes ago, TechyBen said:

As in. Why are you worrying about measuring the battery voltage? You have to first confirm you can at least tap a single phase (and thus it's output voltage/current), let alone all 3 if using a rectifier/converter.

If the battery voltage rises that's proof that the ESC does regen braking. Tapping a phase could be done using external wiring, and actually, regen braking would work with an external circuit, but before I build something like that, and buy a bunch of capacitors and a rectifier, I thought I might just try it on the ESC first. You're right about blowing things in the ESC though, that's risky indeed.

[TechyBen]

Don't even need to measure the battery. Just the output. Like, does the wire produce any voltage/current with a small resistive load. (same way you'd test a solar panel).

*IF* the motor produces power (resistive load?) when in certain modes (from the ESC) then a bypass DIODE strait to the battery might work. But the battery is gonna get hot without proper charge control... and that again, is the second problem in the cycle, if using LiPos.

[akio123008]

Just now, TechyBen said:

Don't even need to measure the battery. Just the output. Like, does the wire produce any voltage/current with a small resistive load. (same way you'd test a solar panel).

Yes, but the ESC only does the braking when it's powered, so that is, when the battery is connected, meaning that when I measure the voltage across the input that's actually the battery voltage. 

[TechyBen]

34 minutes ago, akio123008 said:

Yes, but the ESC only does the braking when it's powered, so that is, when the battery is connected, meaning that when I measure the voltage across the input that's actually the battery voltage. 

Spin the motor with no battery connected. Does it power up the ESC?

 

[akio123008]

38 minutes ago, TechyBen said:

Spin the motor with no battery connected. Does it power up the ESC?

 

No it doesn't. Not too shure it works that way though.

[TechyBen]

That's the way it works.

If there is no power going back over the ESC, then it's not getting power back. It would only have an rectifier or similar to pass power back to it. Granted, there might be a triac or something... that needs switching on or off. But if it's over a single wire for charging, I don't know how the controller would power itself and charge the battery if not a custom/specific setup.

 

So if the motor cannot power the ESC, I doubt the ESC can draw any power/divert it to the battery either!

 

https://www.pandaebikes.com/regen-braking-ebikes-worth/

 

As said. Using a bike dyno and tapping that into the battery is the cheapest/easiest way for you to do some regenerative braking (could engage/disengage the wiring loop with a switch). You'd still not make much power, but would be some fun and learning (like putting it through a charge controller so you don't explodedeaded yourself!).

[akio123008]

1 minute ago, TechyBen said:

That's the way it works.

If there is no power going back over the ESC, then it's not getting power back. It would only have an rectifier or similar to pass power back to it. Granted, there might be a triac or something... that needs switching on or off. But if it's over a single wire for charging, I don't know how the controller would power itself and charge the battery if not a custom/specific setup.

 

So if the motor cannot power the ESC, I doubt the ESC can draw any power/divert it to the battery either!

That makes sense, but turning the motor and listening if the ESC powers on doesn't reveal if the motor can power the ESC; there's some digital electronics involved (the ESC is also programmable) meaning that there may be some sort of "boot sequence" that only works when powering up through the battery terminals.

[akio123008]

Also I think I'd have to turn the motor quite fast, eg attach it to the bike wheel continuously to get the right voltage to power up the ESC, which I haven't tried.

But indeed most likely it doesn't do regen braking.

[TechyBen]

You need a PCB diagram of the ESC to see if it even has the control hardware, before even getting into if it has the control chip/options on the chip or the control software. That's 3 levels of rabbits holes there.

 

If you get it wrong, you blow the diodes!

[akio123008]

Just now, TechyBen said:

You need a PCB diagram of the ESC to see if it even has the control hardware, before even getting into if it has the control chip/options on the chip or the control software. That's 3 levels of rabbits holes there.

 

If you get it wrong, you blow the diodes!

Most likely the braking done is just a short across the motor, perhaps through a resistor, this appears to be used a lot in model aircraft, which often use these ESCs too. Braking is done to stop the propellers to prevent damage on impact when the user loses control over the plane/drone. I guess this is possible to do using a short/resistor because propellers have a very small inertia and can stop quickly. However, my bike (with me on it) will without a doubt keep going for a while, and so the motor is braking for quite a long period of time. That could definitely destroy some of the ESCs electronics. We can never be sure without a proper diagram though, which I don't have.

[TechyBen]

Heres somewhere to start with examining the difference in the ESCs... These ones do regen, though I've not looked at any diagrams.

 

https://www.ebikes.ca/product-info/grinfineon.html

 

Hope it works out. I'm hoping to do a solar powered motorboat one day... but that's a much bigger project/motor and will take me a longer time.

Quote

3.3.1 Ebrake Input
In the first method, you connect an ebrake cutoff lever or other momentary switch
between pins 2 and 4 of the 4-pin ebrake plug. When the ebrake signal is pulled
to ground, then the controller switches to a baseline level of regenerative braking,
putting drag on the wheel and a small amount of current in the battery pack.
If you then press the throttle while the regen lever is activated, the amount of
braking force and regen current will increase, with full throttle giving maximum
regen. This way you can modulate the intensity of the braking with the standard
throttle already on your bike. The only thing to be careful about is releasing the
throttle before you let go of the brake lever once you come to a stop.

 

Quote

3.3.2 0-0.8V Throttle Signal
The second method to engage regen is with a throttle signal that is less than 0.8
Volts. Normally a hall effect throttle will only swing from 0.9V to about 3.6V,
leaving the range of 0.0-0.9V unused. In the Grinfineon controllers, this unused
range is mapped to regenerative braking, with regen starting at 0.8V and then
increasing to a maximum as the throttle signal goes down to 0.0V.
This throttle mapped regen is handy since it allows for variable braking without
any additional wires to the controller, and in principle a small bidirectional throttle
would allow you to both accelerate and slow down without the need for any kind
of brake sensors. It is also how regenerative braking is activated with the V3
Cycle Analyst, which can send any signal to the throttle line.
If you have a geared motor or mid-drive motor, then the regen won't have much
effect. It will cause the motor to quickly stop spinning when you squeeze the
brakes, but it won't be able to help slow the bike down due to the intrinsic
freewheel of these systems.

I assume yours is not freewheeled on the motor. But even things like gearing may mean the results are not always as expected.

 

[akio123008]

3 minutes ago, TechyBen said:

Heres somewhere to start with examining the difference in the ESCs... These ones do regen, though I've not looked at any diagrams.

 

https://www.ebikes.ca/product-info/grinfineon.html

Interesting, I'll take a look at that. If it doesn't work out I can always opt for the bike dyno solution

6 minutes ago, TechyBen said:

 

Hope it works out. I'm hoping to do a solar powered motorboat one day... but that's a much bigger project/motor and will take me a longer time.

That's going to need a whole lot of solar panels isn't it? It does sound like a fun project.

[TechyBen]

1 minute ago, akio123008 said:

Interesting, I'll take a look at that. If it doesn't work out I can always opt for the bike dyno solution

That's going to need a whole lot of solar panels isn't it? It does sound like a fun project.

Lol, yeah kinda. IIRC 1kw output = roughtly 4-5 250w or 300w panels on top... for cruising on near 100% solar (unless I've put the decimal point in the wrong place XD )... but main thing is it would charge a large battery bank between trips. So cruise, charge, cruise. Batteries would be the main power dump.