To brake or not to brake - that is the question

[Frode]

I had this event to day - no no, nothing happened, no new faceplant, no near-death experiences - but it could have been serious if I had not been as cautious, or prepared, or alert, or whatever - from my earlier experience as I was.

Description: My wheel (KingSong KS-16) was fully charged (left it charging overnight) when I started on my  ca. 6 km trip to my work. First there is about 500 meters flat, then there is a long uphill, not particularly steep, for about 500 meters, then about 500 meters flat and then this quite short (30 meters?), but somewhat steep downhill (mye guess is 15% / 8 degrees, about the same as my previous face plant slope, so steep that it is quite hard to pedal up on an ordinary bicycle, but not undoable even if most people will step off and walk). Since I don't know all the inns and outs of my wheel I lowered the speed to about walking speed (3-5 km/h) before starting on the downhill, trying to keep a constant speed. After a few meters the wheel started to beep and the pedals tilted back. I should probably have stepped off immediately (and probably will next time), but since the downhill was so short I decided to just give it a go downhill keeping the wheel at the edge of the beeping/tiltback with the result that the wheel accelerated, not much, but distinct. Since the downhill was so short I did not reach the maximum speed before the slope ended (and what do you think would have happened if the max speed had been reached ! ).

I believe that most people that know what actually happened by now see the problem, and that this is quite a serious problem. There is actually more than one problem. First of all my explanation of what happened (or at least what I think happened - there is no description of how the wheel behaves under different situations , so it is up to the rider to experience what happens and hypothesize (guess) what is the cause and how she should cope with it, hopefully without any serious outcome.

My hypotheses: The reason for the beeping/tilt back is that the batteries was too full to take the current needed for the motor to brake enough to keep a constant and rather slow speed down that slope. The wheel warned me about this breaking creating a (sort of) overpower situation and started to beep and tilt back the pedals. Note that this back tilting the wheel does with the pedals will not generate any breaking in itself. It just offsets the vertical of the wheel. You will have to force the pedals even more back on top of that tiltback to brake in order to lower the speed of the wheel !!!. Since speed was low (being precautious) and I therefore did not feel insecure, I decided to let go of some of the breaking with the result that the beeping/tiltback stopped, then trying to lean back again to brake, which of course brought back the beeping/tiltback. This is how I went down the whole slope: lean back - beep/tiltback - lean a little forward - beep/tiltback stopped - lean back - beep/tiltback - lean forward againt - and so on, resulting in the speed slowly increasing downhill.

Now, I was precautious and lowered the speed to very slow before the downhill, because I know in general how today's wheels work, and how they brake and that there is a limit to how much it can break, especially when the batteries are full. BUT: what would have happened if I didn't know anything about this and entered the slope whit a speed of say 20 km/h? Or the downhill had been longer, the speed increasing and it becomes more and more scary to just jump off? You can't brake since the battery will not take the generated current. You can't even keep the speed constant. Accelerating downhill you will either reach the bottom of the slope - or, much more serious: reach the upper speed limit where you either have to jump off or be thrown off.

So the moral must be:

1) Start slow when going downwards on steep hills, and jump off immediately if the wheel cannot brake enough to keep the speed constant (and it is not just a very short downhill, <30 meters?). Or else you might hit a runaway situation, ending in being thrown off the wheel because the maximum speed limit is reached.

2) The wheel-makers should ASAP inform the riders of how exactly the individual wheel will behave in different situation (the user guide that followed my wheel is of limited value - for example it does not mention anything about my above experienced/hypothesized situation). Give the rider a good mental model of the system in order for her to understand the limits of the locus of control - and what exactly will happen if those limits are exceeded. The other option is (as it is today with at least my wheel) to let the rider experience the limits herself and let her find a solution on the fly - if she can and have enough time). General advice as "keep speed low", "there are alarms and pedal-tilting for everything" or "be cautious" is of limited value (read "no value") when you are standing in the midst of a situation that you did not understand would come and how to handle it best.

3) The wheel-makers should start looking into emergency braking (power resistor? mechanical? or maybe even simple hydraulic?), that will be applied when the user is trying to break, but the motor cannot generate enough breaking force. This will of course generate heat, but since it is to be used only when the motor itself cannot break any more - and the user is alarmed to stop (through beeping, and also she is well informed that this is what she must do - stop - when hearing the alarm - or end up in trouble) - this should be of no problem.

 

Now, that was my 2 cents, or "16.56 øre" as it will be in my currency...Any comments?

 

[Cloud]

I think you " overpowered" the wheel. The slope was so steep that the wheel struggled to keep you going slow resisting all that notorious gravity. The wheel alarmed because it was approaching the limit of its torque capabilities. Yes, i guess one way is to increase the speed, thus reducing the load on the motor, but a more wise cours of action, especially since its dangerous to keep accelerating downhill and because you are not familiar with the wheels alarms and what they mean, is to step off the wheel and take a moment to reflect on what might be happening.

another alternative to reduce load is to go downhill sideways ( sort of like a skiing zig zag style) to lower the amount of decline the wheel has to fight. Still, if you are not aware of how the wheel behaves, first thing is to slow down and in this case to stop.

[Frode]

1 hour ago, Cloud said:

Still, if you are not aware of how the wheel behaves, first thing is to slow down and in this case to stop.

As I tried to explain, this was not an option (impossible). The minimum speed I could have had when stepping of the wheel was the entrance speed to the slope. The longer you hesitate, the higher the speed (until you are either at the bottom of the hill or thrown off by the wheel).

BTW: I have been down that and similar slopes several times before (10+) without any trouble, even slowing the speed down to a stop. The only reason I can figure out that I overpowered the wheel this time is that the batteries were full (alternatively that there is some technical fault)

1 hour ago, Cloud said:

another alternative to reduce load is to go downhill sideways ( sort of like a skiing zig zag style) to lower the amount of decline the wheel has to fight. Still, if you are not aware of how the wheel behaves, first thing is to slow down and in this case to stop.

Good idea. I'll try it out. Remember though that in every turn the alarm will go off, the pedals will start tilting back and the wheel will accelerate.

Turning (even so far as to go uphill again) and then breaking to stop might be a viable option for emergency breaking downhill (when there is room for turning). I think I'll start practicing that.

[The Fat Unicyclist]

1 hour ago, Frode said:

First there is about 500 meters flat, then there is a long uphill, not particularly steep, for about 500 meters, then about 500 meters flat and then this quite short (30 meters?), but somewhat steep downhill (mye guess is 15% / 8 degrees, about the same as my previous face plant slope, so steep that it is quite hard to pedal up on an ordinary bicycle, but not undoable even if most people will step off and walk).

Hang on though - back in school when I learnt physics (and other things that I have never needed to remember again - until getting an EUC), we were taught that you can't make or destroy energy - only convert it. 

In general, riding an EUC along a flat converts stored (battery) energy into kinetic energy. Add in the uphill and you convert even more stored energy, as you need to lift the mass (you) vertically as well. 

On the downhill you then regenerate, converting the energy regained from the descent back into stored energy. 

BUT, unless your final altitude is significantly lower than your original altitude, surely with the 1500m of horizontal travel thrown in to the mix there is no way you can overcharge the battery? 

I live in a "mountainous" area too, but I start partway up a large hill. So I have similar questions...

Any actual physicists or there, please feel free to point out the probable flaws in my logic (in either discussion). 

[Frode]

27 minutes ago, The Fat Unicyclist said:

In general, riding an EUC along a flat converts stored (battery) energy into kinetic energy. Add in the uphill and you convert even more stored energy, as you need to lift the mass (you) vertically as well.

Of course you cannot break the law of conservation of energy. But this is not about conservation of energy. It is about charge current to the batteries.

What happens is that as long as the charging system/batteries can take the needed current you will have a current through the windings in the motor. As soon as the demanded current is higher than the maximum that can flow, some (or even all if completely shut off) energy will be lost converted to accelerating the speed and eventually to heat generated when you plant your face in the ground

[The Fat Unicyclist]

46 minutes ago, Frode said:

Of course you cannot break the law of conservation of energy. But this is not about conservation of energy. It is about charge current to the batteries.

What happens is that as long as the charging system/batteries can take the needed current you will have a current through the windings in the motor. As soon as the demanded current is higher than the maximum that can flow, some (or even all if completely shut off) energy will be lost converted to accelerating the speed and eventually to heat generated when you plant your face in the ground

So you're saying it isn't that the batteries are (necessarily) fully charged, they just can't take the charge as quickly as required? 

And if that is the case, would the battery temperature factor into how well the battery can accept the charge? 

[EricGhost]

I've a ks16 680wh and on july I went to work facing the following down in the morning up in the afternoon at least 15 times in the month, at something around 10-20 max kmh but never batteries near 100% more like 60-85%, I do not know if It is enough steep, but by bike many prefer to walk

 

 

[Keith]

1 hour ago, The Fat Unicyclist said:

BUT, unless your final altitude is significantly lower than your original altitude, surely with the 1500m of horizontal travel thrown in to the mix there is no way you can overcharge the battery? 

Any actual physicists out there

I had thought that if you have taken so much out then surely there is no problem until you put more than that in........ right up to the point where I accelerated to full speed on the flat with a recently charged battery and then braked, only to find my peddles completely floppy before taking a short flying lesson.

My theory, which as Monty Python once said, is mine and mine alone, is that with a moderately close to full battery there is a real tendency for that regenerative voltage to become to high - I.e. Significantly above 67.2V which, even though the battery isn't actually full yet is too high. I think this is where we need Eagletree's data acquisition equipment to see exactly what is happening to current and voltage at those times.

As an aside, I decided to investigate my wheel's low voltage behaviour recently and set off for a 4 mile run with 4 of 10 LED's showing. Going up a paved path over a man made  hill was painfully slow beeping away at 8MPH with zero to 2 LED's. As I got to the top I accelerated with some relief down the other side, hoping to get some charge into the wheel and it started beeping at 8MPH and then tilted back on me as well, The LED's did some strange things including the top 2 and bottom 2 flashing alternately. Today I took my wheel over the same hill fully charged 1.5 miles before reaching the hill I was able to maintain 16MPH both up it and down it. The LED's dropped to 8 on the way up and all 10 on the way down.

So yes, looks like I could do with a good physicist too!

What I do know is there are significant problems with using any sort of braking mechanism instead of, or as well as, regenerative braking. A mechanical brake would be very difficult to operate with the finesse of control to also keep the rider balanced and shunting the charge through a resistor of some form once the battery has had enough would create a VERY large amount of heat. Cruising on the flat consumes something like 250-350 Watts of power, it is reasonable to assume that maintaining a constant speed down a relatively steep hill would generate somewhere between 100-400W, that's one very big resistor and heat sink.

we do actually do exactly that on our model aircraft literally shorting the motor coils via MOSFETs. it is done on motor gliders with a folding propellor, when the motor is turned off the prop would keep windmilling and will not fold. So the controller applies an electronic brake to stop the prop and allow it to fold. In this case it works fine because there isn't a lot of energy in the prop and it's stopped within a few seconds.

[Frode]

39 minutes ago, The Fat Unicyclist said:

So you're saying it isn't that the batteries are (necessarily) fully charged, they just can't take the charge as quickly as required? 

And if that is the case, would the battery temperature factor into how well the battery can accept the charge? 

I'm not into battery chemistry, but I assume that there might be a correlation between how much current pr time unit you can put into a cell and how much charge it has received/voltage (putting more into it maybe rices the required voltage to put it in there and/or the temperature, but this is my assumption). For lithium ion batteries there always is a battery maintenance system that protects the batteries from exploding/taking fire. What exactly limits the current into the battery  (and how) I'm not knowing. It might be the BMS   in a way dependent on the battery voltage/temperature or whatever to protect the battery's - and rider. Li-ion batteries are dangerous if not handled correctly - look 2 minutes into this video:

 

 

[Frode]

27 minutes ago, Keith said:

Cruising on the flat consumes something like 250-350 Watts of power, it is reasonable to assume that maintaining a constant speed down a relatively steep hill would generate somewhere between 100-400W, that's one very big resistor and heat sink.

You don't need a resistor able to consume 400 W continuously, only for the short time it takes to stop. Consumed energy (what is accumulated as heat in the resistor) is the wattage times the time (Ws or J), not the wattage alone (just as it takes a while before your electric kitchen stove gets hot even if you turn it on 3 kW).

A resistor can be a viable alternative for the wheel to allow you to stop instead of just throwing you off the wheel after accelerating to maximum speed. If you sound an alarm as long as there is power going through the resistor (and even an extra alarm when it gets to 200 Celcius and starts burning?) and the rider is well informed of what this alarm indicates and what to do (stop immediately), I can see no problem, especially if the alternative is speeding up to maximum speed and then throw her to the ground.

[HEC]

5 hours ago, Frode said:

My wheel (KingSong KS-16) was fully charged (left it charging overnight)

I don't know what battery size do you have, where did you get your KS from and if you have the fast 5A charger but it's a good practise to not charge the batteries fully all the time. Either using fast charger @Jason McNeil provides with his 680 and 840 Wh wheels or Charging Doctor limit the charging to 80 - 87% of the capacity on "regular" charges to both prolong your battery life as well as avoid issues with breaking or riding down hill while close to full battery charge and only perform a full charge once every 15 - 20 charges to keep the cells balanced or when you're planning extra long trip.

[Frode]

40 minutes ago, Keith said:

A mechanical brake would be very difficult to operate with the finesse of control to also keep the rider balanced

Not necessarily if it sits in a closed loop (a regulated system). You could say the same for the motor. If you were to control the motor moment by say a accelerator handle (like an equivalent of an accelerator pedal in a car) it would be very difficult for you to stay balanced on the wheel. But because it is controlled by a closed loop it works fine (measuring the angle of the wheel and trying to maintain  it vertical with input to the motor depending on this angle). You could feed a mechanical or hydraulic break in the same manner. If you fall to far back, break more, if the wheel breaks so much that you are moving forward, loosen the break (it should of course be free from lugging etc). Systems like that is in use every day, for example in trailer brakes (you cannot let a heavy trailer behind a car brake more or less than the car itself, it must adjust itself to break exactly as much as the car, the brake regulator can even be made quite simple and fully mechanical)

[Frode]

6 minutes ago, HEC said:

I don't know what battery size do you have, where did you get your KS from and if you have the fast 5A charger but it's a good practise to not charge the batteries fully all the time. Either using fast charger @Jason McNeil provides with his 680 and 840 Wh wheels or Charging Doctor limit the charging to 80 - 87% of the capacity on "regular" charges to both prolong your battery life as well as avoid issues with breaking or riding down hill while close to full battery charge and only perform a full charge once every 15 - 20 charges to keep the cells balanced or when you're planning extra long trip.

I bought the KS-16 here: http://e-wheels.no/ and they told me they imported it from China. It is the Sanyo 840 Wh battery (printed on a label on the side of one pedal), and I use the original charger that comes with the wheel (from China I guess). Its output is specified as being 67.2 V, 2 A and charge time aout 5.5 h. The user manual also tells me the following (in exact words):

"When the red lights turn into green, indicating the charging is completed.
If not in urgent need, please wait until charging complete before unplugging, because after charging, the protection circuit will perform flow equilibrium."

And just to clarify: I don't think the KS-16 misbehaves. I just think that the user, possibly a newbie with allergy against technology (that one that have a car where the most technical information he is willing to accept is how to put gas on the tank and how to get it into gear and move) should be informed about the behavior of the newly acquired wheel. I also think that the wheel makers should put more effort into avoiding potentially dangerous runnaway situations as the described situation is (by providing information on when to expect it, how the wheel will react and what you should do, for example for the described situation: "step of the wheel immediately, otherwise the speed will increase and you will eventually be thrown off at maximum speed"). There is nothing - and I really mean m\nothing in the manual saying anything about this. How should you know if you are not technically oriented and understand the concept buried in the technology that itself is buried inside the wheel's shell?

[HEC]

@Frode here is a bit of readout about battery charging from @Jason McNeil: https://www.wheelgo.com/ewheels-fast-chargers-rock/

As you have just standard KS 2A charger without option to limit the charging directly on the charger you shall consider getting the Charge Doctor instead: http://hobby16.neowp.fr/buy/ (read the rest of the blog there for more info)

If you'd like to have an option of fast charging you can also order the V2 CD with two inputs GX-16 sockets for extra 2 EUR which will allow you to attach 2 "slow" chargers to double the speed. There is a plenty of places online where you can order additional cheap 67.2V / 2A charger with standard GX-16 plug instead of purchasing much more expensive 4 or 5A chargers (although those are still pretty affordable on aliexpress for example and you'll avoid hassle of shuffling with two chargers all time and can then even take the original slow charger to work, summer house etc.).

More info on CD also directly here in forums (it's quite old thread so a lot has changed since it has started and also Fred now have even newer version of CD with both current as well as voltage limit cut-off options):

In regards of the "insufficient" instructions - on one hand - yes I do agree that provided information in user manual are less than sparse though on other hand you do not expect user manual with a bicycle or car to teach you how to ride one or how to observe the traffic rules and regulations either ... It's still emerging market for EUCs and we as early adopters need to treat them as such and do our research. That indeed doesn't prevent "ignorant" users to smash their face on ground (or worse) by not doing the research or at least using the common sense sadly.

[Frode]

50 minutes ago, HEC said:

If you'd like to have an option of fast charging ...

Thanks for info!

I'm not very well into charging Li-ion batteries, but I have learned that not all Li-ion-designs deal well with fast charging. I'm not sure if the KS-16 Sanyo batteries are made for fast charging. Are they?

Anyway, at the time I'm not in need for fast charging. 840 Wh is plenty for my feet needing a good nights sleep before I mount on the wheel again After all, standing with my 48 size feet (european, US 13?) with all my near 120 kg (265 lbs?) on those small KS-16 pedals is not as comfortable as sitting on my Vulcan 900 with big floorboards.

 

50 minutes ago, HEC said:

though on other hand you do not expect user manual with a bicycle or car to teach you how to ride one or how to observe the traffic rules and regulations either ...

 

Yeah, I agree. However, before getting a license for driving a car, you need hours of theory courses and tens of hours of practical training (at least in my country). And also in school they teach children about how to behave in traffic (still in my country).

And neither cars nor bicycles have built inn potentially dangerous runaway "features" like the one described. What would you tell the car/bicycle maker if you drove your car/bicycle downhill and suddenly your bicycle or car started to accelerate giving you no other option than either sit and wait for the crash or jumping off? And if he said that this is how we designed it, but we thought that you should find it out by experience? Cars and bicycles have two independent (more or less) braking systems. Fore some reason or another they are required to.

 

[HEC]

39 minutes ago, Frode said:

I'm not very well into charging Li-ion batteries, but I have learned that not all Li-ion-designs deal well with fast charging. I'm not sure if the KS-16 Sanyo batteries are made for fast charging. Are they?

Yes they're - the very same @Jason McNeil provides his fast charger for  And indeed  if you're OK with normal charging speed its all good though I'd still look into limiting the charging in some way. If you don't want to invest into CD you can use a cheap AC socket timer to end the charging before the full charge.

[Frode]

7 minutes ago, HEC said:

Yes they're - the very same @Jason McNeil provides his fast charger for  And indeed  if you're OK with normal charging speed its all good though I'd still look into limiting the charging in some way. If you don't want to invest into CD you can use a cheep AC socket timer to end the charging before the full charge.

 

 

OK. Thanks!

 

But with respect to the timer, what about KingSong's advise in the user guide:

1 hour ago, Frode said:

"please wait until charging complete before unplugging, because after charging, the protection circuit will perform flow equilibrium."

 

[HEC]

8 minutes ago, Frode said:

But with respect to the timer, what about KingSong's advise in the user guide:

1 hour ago, Frode said:

"please wait until charging complete before unplugging, because after charging, the protection circuit will perform flow equilibrium."

Which is referring to cell balancing which is though not needed on every charge. If you want to be really conservative you can perform a full charge every 10 chargers though every 15 to 20 will do just fine with high quality cells. As the battery pack will age after couple of years then indeed the frequency of the full charges shall increase.

[Blechi]

Wait a second.

Is regenerative braking the only way an EUC slows down?

I can't believe that.

What about the other ways to slow down an electric motor without using a mechanical brake:

 - shorting the coils (either with a resistor or a PWM controlled short via the MOSFETs)

 - controlled acceleration in the opposite direction (drains battery instead of charging it)

 - applying constant DC to the coils (also drains battery)

This works with any electric motor (except the third one which doesn't work with brushed motors), so why shouldn't it be used in an EUC?

The MOSFETs should be able to handle the current and the thermal mass of the motor should be sufficient for that.

Are the manufacturers / designers really that dumb to rely just on regenerative braking?

Or am i on the wrong path here?

[HEC]

12 minutes ago, Blechi said:

This works with any electric motor (except the third one which doesn't work with collector universal motors), so why shouldn't it be used in an EUC?

Because you'll eat tarmac the moment any of those would be used - the wheel must keep rider upright at all times (braking or not) hence why "alternative" motor breaking methods will be no use here.

[Blechi]

12 minutes ago, HEC said:

Because you'll eat tarmac the moment any of those would be used - the wheel must keep rider upright at all times (braking or not) hence why "alternative" motor breaking methods will be no use here.

That's why i wrote 'controlled'.

A full short for example would take the rider down immediately, that's right.

But think of it:

These all are controllable ways of braking. Balancing can still be achieved by automatically braking more or less.

The same way it's done by regenerative braking.

[HEC]

Most of the current EUC electronic designs is down to "legacy" from e-bikes. Additional methods from your list would either include extra heat generation in some already heat loaded parts (motor coils, MOSFETs etc.) or complicate the circuits enough to most likely lead to even more dangerous faults rather than beneficial improvements. There is no discussions that it's far from perfect but in this case I'd say the strength is in simplicity. As far as I'm aware all (serious) manufacturers are warning the users about breaking or riding downhill with full batteries when this problem can demonstrate.

[Blechi]

4 minutes ago, HEC said:

Most of the current EUC electronic designs is down to "legacy" from e-bikes. Additional methods from your list would either include extra heat generation in some already heat loaded parts (motor coils, MOSFETs etc.) or complicate the circuits enough to most likely lead to even more dangerous faults rather than beneficial improvements. There is no discussions that it's far from perfect but in this case I'd say the strength is in simplicity. As far as I'm aware all (serious) manufacturers are warning the users about breaking or riding downhill with full batteries when this problem can demonstrate.

Ok, lets remove the resistors from the list as too complicated and additional parts.

The other methods of braking only need minor modifications of the motor / electronics design. The rest is software.

For example: using wire that is a little thicker for the coils would improve the thermal properties (without changing the magnetic part).

The manufacturers already start to use more MOSFETs in parallel, just use some more and improve the thermal design.

As it is now the motors and mainboards are built down to a price.

Add a nice housing, some cheap blinky blinky LEDs and / or some cheap bluetooth squeekers and you have today's EUC.

Ah yes i forgot: and make it faster, no matter what.

I think it's time for in depth R&D and solutions, not just 'what do you want? It moves.'

To achieve 80% of the goal you have to do 20% work. The last 20% require 80% work, and IMHO that's what today's manufacturers are not willing to do (the big ones, not the cheap copycats).

I'm thinking of the Tesla way: build a flagship to gain experience, then reduce details to get to a version for the masses.

Now have a look at the EUC 'flagships' and compare.

 

[HEC]

2 minutes ago, Blechi said:

For example: using wire that is a little thicker for the coils would improve the thermal properties (without changing the magnetic part).

The manufacturers already start to use more MOSFETs in parallel, just use some more and improve the thermal design

Changing the thickness of the coil wires will though have impact on torque / speed ratios ...

While adding multiply MOSFETs for each channel could improve the thermal load / dispersion (or not if implemented incorrectly) it can also increase failure rate due to higher amount of active electronic parts.

I do agree that EUCs needs larger overhaul in design though that will take some more time and without manufacturers being pushed be legislative / certification process it's not going to happen as they're way to small to bear the risk of increased cost without any obvious benefits (to manufacturers) as long as they can get away with it  

[MaxLinux]

3 hours ago, Frode said:

neither cars nor bicycles have built inn potentially dangerous runaway "features" like the one described

How common is this "runaway" problem on hills? I live in a very hilly area, and I have never had this problem.