Jump to content

Overspeed Cutoffs


Recommended Posts

Can anyone please explain what exactly happens during an overspeed cutoff, and how to best react in that situation?

Why do they happen? Why can't these things be engineered to simply not go above a certain speed, so that the motor doesn't cut out?
(If you unlocked it to allow overspeeding on purpose, that's a different matter)

 

 

Link to comment
Share on other sites

My understanding of this topic is not comprehensive but this is the basic rundown. The wheel basically has 2 functions that draw power - acceleration (forward or backwards) and stabilization - attempting to stay upright. We will assume for the moment that lights, bluetooth, etc draw effectively zero power. As you draw a higher and higher percentage of available power to acceleration, less and less is available for stabilization. The power draw of stabilization is dynamic. If you lean hard suddenly or change the orientation of the wheel suddenly it will spike power to try to compensate and keep itself in a vertical orientation. The faster the attempted change, the more power it draws per second to stay vertical.

As you push the edges of your wheel's performance limits you run into a problem - if 95% of available power is used to keep your current velocity you do not have enough reserve power ceiling to handle a sudden change of orientation. Something like a bump in the road, a sudden change in slope, or balance by the user and the wheel is suddenly pushed over its power limit. When the wheel exceeds its power envelope it does an emergency shutdown which we perceive as a "cutoff." Presumably, if it did not shut down it would suffer a catastrophic failure soon after (burned out mosfet, melted wires, etc).

When unpowered, the wheel has no stabilization which EUC riders rely on. Without it we are now leaning, probably relatively strongly if we were going fast, and the wheel is no longer keeping itself vertical so we just fall straight onto our faces as if someone just pulled the chair from under us.

This often happens very suddenly because many riders eliminate or move the warning alarms so there isn't much warning before it happens. Even if not, sometimes you just hit something or do something that draws an unexpected amount of power. Also the power envelope your wheel has to work with changes as the battery discharges. You can lose 20% or more of your power output between full charge and low battery so sometimes something you are used to your wheel being able to handle is now just outside its capability due to the lower battery charge.

Sometimes people don't realize that going fast isn't the only thing that can draw a lot of power. Going slow can draw just as much power since the wheels don't use traditional disc brakes, they simply provide power in reverse to slow down/stop. This means that maintaining a slower speed on a steep hill can be just as taxing as going fast, and we often do not intuitively understand exactly how close we are to the edge of the wheel's ability like we do when we are going fast.

On top of this, the wheel itself could have faulty firmware or other issues that cause a cutoff to happen unnecessarily.

I am not an electrical engineer so I am not certain what could be done to improve how this situation is handled. If this were a major industry with regulated safety standards the wheels would probably be locked at a much lower speed than they are theoretically capable of reaching. Kingsong actually locks their wheels at 20kph even though some are capable of nearly double that speed. But pretty much everyone unlocks it (me included). If you stayed under 20kph it probably would never cutoff due to lack of available power. You would also be slower than a ninebot =). Humans are greedy I guess.

As for what you can do to handle it? Don't turn off the warning alarms, set tiltback below the max speed to leave some safety margin, be aware to ride more carefully as the battery runs down below 50%. When it actually happens it is usually too fast for you to react to but if you sense your wheel acting oddly you should stop and try to figure out what is wrong. Other than that... wear protective gear?

 

Hope this helps.

Link to comment
Share on other sites

12 minutes ago, sanman said:

Can anyone please explain what exactly happens during an overspeed cutoff, and how to best react in that situation?

This question comes up at regular intervals. It isn't usually a cut-off, unless you have a BMS (Battery Management System) that shuts down due to overload - most good quality wheels do not.

What it usually is, is simply a matter of running out of torque. Maximum Torque is inversely proportional to speed. At a standstill torque is maximum at the maximum RPM of the wheel it is zero. At some point before that maximum RPM the torque will be just about enough to power the wheel and hold the rider in balance, the smallest bump, or further lean forward by the rider will be too much for the motor and the rider will fall forward.

13 minutes ago, sanman said:

Why do they happen? Why can't these things be engineered to simply not go above a certain speed, so that the motor doesn't cut out?

This question relies on not understanding the fundamental way the self balancing device works. The wheel does not actually drive forwards when you lean forward, it balances you - THAT IS ALL IT DOES!

If you lean forward the wheel has to counter your lean with an opposite torque reaction the byproduct of that reaction is that the wheel moves forward and accelerates. All the time you lean forward the wheel HAS TO COUNTER YOUR LEAN WITH TORQUE. I.e. It cannot "be engineered to not go above a certain speed" - the only way to do that is to make you fall off of it. 

To summerise: if you lean forward it HAS TO GO FASTER. The only real way of stopping you from leaning forward is to discourage you either by tilting back (which requires even more torque to achieve) or by beeping irritatingly at you. If you ignore those warnings there can be nothing that will stop you going even faster until you faceplant.

The other problems are things like weight of the rider, gradient you are riding up, battery voltage level, etc, etc. All of these will impact the speed at which the wheel will run out of sufficient torque to hold you up. These are easily fixed by having VERY conservative tiltback speeds, which will upset many riders (speed, speed gimme more speed!) or very clever algorithms that adjust tiltback on the fly. Some wheels, Kingsong for example, will lower max speed tiltback as the battery gets lower (and people then complain!) but I do not know any yet that are clever enough to adjust tiltback speed for the weight of the rider.

Link to comment
Share on other sites

1 minute ago, Keith said:

To summerise: if you lean forward it HAS TO GO FASTER.

What do you mean "it HAS to go faster"? There's no physical law that says leaning should magically draw more power out of the battery to make the motor run faster. This behavior is engineered into the device - and that engineering could be modified so that if you're over a certain speed, or drawing too much power, then further leaning should not result in any further current draw. If you're saying that autostabilization forces more motor rpm to offset your leaning, then what needs to happen instead should be tiltback.

Link to comment
Share on other sites

19 minutes ago, electricpen said:

Presumably, if it did not shut down it would suffer a catastrophic failure soon after (burned out mosfet, melted wires, etc).

 

11 minutes ago, sanman said:

There's no physical law that says leaning should magically draw more power out of the battery to make the motor run faster.

Neither of you are understanding this at all! The faster the wheel goes the more back e.m.f. there is from the motor. That back e.m.f. (The motor acting as a dynamo) reduces the apparent voltage at the motor and as a result the current reduces as well. I.e. going too fast does not result in excessive power being drawn. the highest torque is available at zero speed and highest power at 50% of the absolute maximum speed beyond that the power drops due to the back e.m.f.  Key thing: it is torque not power that balances the rider.

Yes there is a physical law - I've already explained it. You lean forward and the wheel has to produce enough torque to counter your lean - that torque results in acceleration. If it did not do this you would faceplant, end of story. And,Yes, the only option available to the designer is tiltback which HAS TO happen whilst there is still enough excess torque to do it.

This has been discussed in a lot of detail before.

Link to comment
Share on other sites

Okay, that's what the last sentence in my reply said - the autostabilization compelling a counter-response from the motor - but fine, I should have said torque, so point taken. Yeah, I recognized that you need spare torque in order to do the tiltback - you can't just be doing it while you're crossing the breakdown point. But so then it sounds like what's needed is higher performance envelope to allow for appropriate safety margins.

 

Link to comment
Share on other sites

1 minute ago, sanman said:

But so then it sounds like what's needed is higher performance envelope to allow for appropriate safety margins.

Again, physics is against you. For a given physical size of wheel, you can only get so much copper into the motor and, unless we invent even more powerful magnets and/ or room temperature superconductors, only so much magnetic flux. To increase the torque you need to lower the kV (RPM per Volt) which means lowering the no load RPM - I.e. Maximum speed. To go faster a higher kV is needed which lowers the maximum torque. Higher battery voltage would help (which is why we are now seeing 20 cell wheels ) but there are cost, safety and electronics limits to that. The manufacturer has to make a set of informed compromises. 

Or you can use a physically bigger and heavier motor (and therefore entire EUC) to get the increased performance you desire at a weight, portability and cost penalty.

Link to comment
Share on other sites

1 hour ago, sanman said:

Why do they happen? Why can't these things be engineered to simply not go above a certain speed, so that the motor doesn't cut out?
(If you unlocked it to allow overspeeding on purpose, that's a different matter)

The concept of "self-balancing vehicle" and "simply not going above a certain speed" are mutually exclusive.

If you lean, the wheel has to compensate (do something to regain balance) or you fall (loss of balance). Technically (principle of "self-balancing"), it cannot 100% force you to go slower. It certainly can try to do that (beeps, tiltback) but technically a rider could overcome this (keep leaning, imagine tiptoeing on heavily tiltbacking pedals on a beeping wheel) and the wheel would either have to accelerate to keep the balance, or not do that and lose balance (crash). It can accelerate only so far, so if that barrier is broken, crash too.

Now you can ask related questions:

  • Why does the wheel allow you to go "too fast" without stopping you?
    Answer: It cannot technically stop you, just try to get the rider to reduce the lean (beeps, tiltback) by him-/herself.
  • Why does the wheel not warn you (beeps, tiltback) when you're going too fast/too close to the dangerous zone?
    Answer: Wheels do, just some wheels allow the warning to be switched off, then there's nothing they can do except accelerate until they can accelerate no more.
  • Why does the motor switch off instantly in such a situation, instead of just keeping giving it's doable maximum, and the rider would "slowly" fall forward and could maybe catch balance by reducing lean, instead of instantly losing the ground (pedals) beneath the feet?
    Answer: This is complicated. Some say to protect the electronics. You could argue, if a fall is guaranteed anyways (after all, the wheel has not enough ability to balance), no point in pushing any more energy into the situation (which might just increase injuries) by having a running powerful motor piling on it. Maybe it's just technically easier to build it that way.
  • Aren't there any other clever ways to get around this problem and not allow people to stress their wheels too much?
    Answer: Certainly. If you have a good idea...;)
25 minutes ago, sanman said:

But so then it sounds like what's needed is higher performance envelope to allow for appropriate safety margins.

Exactly. Reaching the maximum power/torque/whatever means a guaranteed crash (principle of "self-balancing vehicle"), so this must never happen and the higher the performance envelope, the less likely that is.
But in theory, you could overlean any wheel (hold on to the shell and keep leaning it, might need some acrobatics while it's tiltbacking and giving you electric shocks and whatever) until it can no longer balance and then you forced a crash. In the end, the wheel can only react (balance), the rider has to do the rest.

Link to comment
Share on other sites

@electricpen Great post!:thumbup:@Keith Nice, always good to hear the technological details!:thumbup:

46 minutes ago, electricpen said:

The wheel basically has 2 functions that draw power - acceleration (forward or backwards) and stabilization - attempting to stay upright.

Those two are exactly the same thing. That's the very principle of "self-balancing vehicle". It's literally just a motor that reacts to a sensor's tilt. Can't have one without the other (aka the wheel can't brake on its own, the rider has to lean back first).

Link to comment
Share on other sites

I realize that it is one action but I feel that it is easier to understand why it cuts out suddenly if you conceptually break the power draw into categories. To the motor, it is the same thing but to the rider, it might be easier to understand why power spikes under certain circumstances if they are considered separately. I could also be totally wrong but that's how I had thought of it.

Link to comment
Share on other sites

You can look at this however it helps you think about something conceptually:) There's no right and wrong, just (ultimately subjective) stuff like "misleading", "suggesting the right or wrong conclusions", etc.

For the old question "Why does this thing not just brake?", you could argue that the it's-all-the-same-viewpoint might be the most helpful, to arrive at the "There's only one thing for speed control and balance control, so you can't change one without changing the other." answer.

Link to comment
Share on other sites

15 hours ago, electricpen said:

My understanding of this topic is not comprehensive but this is the basic rundown. The wheel basically has 2 functions that draw power - acceleration (forward or backwards) and stabilization - attempting to stay upright. We will assume for the moment that lights, bluetooth, etc draw effectively zero power. As you draw a higher and higher percentage of available power to acceleration, less and less is available for stabilization. The power draw of stabilization is dynamic. If you lean hard suddenly or change the orientation of the wheel suddenly it will spike power to try to compensate and keep itself in a vertical orientation. The faster the attempted change, the more power it draws per second to stay vertical.

As you push the edges of your wheel's performance limits you run into a problem - if 95% of available power is used to keep your current velocity you do not have enough reserve power ceiling to handle a sudden change of orientation. Something like a bump in the road, a sudden change in slope, or balance by the user and the wheel is suddenly pushed over its power limit. When the wheel exceeds its power envelope it does an emergency shutdown which we perceive as a "cutoff." Presumably, if it did not shut down it would suffer a catastrophic failure soon after (burned out mosfet, melted wires, etc).

The wheel does not perform an emergency shutdown in such situations - there is not cutoff! And it would not suffer a catastrophic failure soon after.

If one reaches the power/torque limits (power is direct proportional to torque times speed) the wheel just cannot produce more torque/power in because of the back emf generated from the motor. Once this back emf (together with the voltage drop over the coils) reaches the battery voltage no more further acceleration is possible -> no more (just limited) balancing -> overlean. This can happen (especially at higher speeds) without any real high power needed - so no danger for the controller/wiring and nowhere near the limit of what the battery could provide.

How fast and severe this "no/limited balancing" happens depends upon how "fast" one accelerates into this limit. With "slow" approaches there should be a chance to break and handle the situation - with "faster" approaches it just feels like a cut-off...

Cutoffs by now only happen in overcurrent situations (also many are with nowadays boards already "prevented"), overheating, battery cell over/undervoltage. And of course some major malfunctions...

http://forum.electricunicycle.org/topic/7855-anatomy-of-an-overlean/ shows wheellog data of an overlean together with the limits...

At the point he reached the torque limit just something around 1-1.2kW where used - just a bit above the nominal power for the KS16C. Nowhere near some peak performance... (Absolute maximum possible output power should be around 2.2kW)

Link to comment
Share on other sites

2 hours ago, Chriull said:

Cutoffs by now only happen in overcurrent situations (also many are with nowadays boards already "prevented"), overheating, battery cell over/undervoltage. And of course some major malfunctions...

Thank you @Chriull, I think you have explained it better than I did.

@electricpen I think you may be confusing the cut off that happens if you lift the wheel off of the ground. This IS programmed into the firmware to stop the wheel from becoming dangerous if you fall off of it, similarly there are cut offs if the wheels tilt in any direction exceeds a certain number of degrees for safety I.e. It will cut off if it falls over.

The high speed cut-off that happens if the wheel is lifted off of the ground is close to the no load speed - I.e. Absolute Maximum Speed the motor is capable of. At that point there is virtually zero torque and therefore virtually zero power and zero current - back e.m.f. is very close to the battery voltage - it is much faster than the wheel can achieve with a rider on it so there should be no danger of that accidentally happening with a rider.

Theoretically, at least, it might be possible that a rider is travelling fast, goes over a bump or kerb, becomes airborne and the wheel then speeds up to the point where the safety cut off happens. By the time the rider hits the ground again the wheel is powered down. Should that occur in any make of wheel it ought to be seen as a firmware issue, I.e. If an EUC experiences that behaviour the programme needs to be adjusted to increase the delay before the wheel shuts down, ensuring the rider is back on the ground before the firmware can cut power. I'm not at all sure I've seen any conclusive evidence this is happening?

Don't get me wrong, EUC's do suffer sudden power cut offs for a whole variety of reasons from blown fuses to bugs in the firmware, to simple mechanical or electric faults such as a burnt MOSFET - since these usually fail short circuit they cause a substantial braking action on the motor which will ruin anyone's day. It is just not usually the reason a rider comes off if going too fast. When cut-offs are identified as repeatable problems with a particular model of EUC there is usually a vociferous lobby about it and manufacturers have often been quite quick to respond. A really good example is below: I think @Marty Backe is rapidly becoming Gotway's chief test pilot, along with @EUC Extreme who REALLY knows how to push the envelope on EUC's.

 

Link to comment
Share on other sites

21 minutes ago, Keith said:

...

The high speed cut-off that happens if the wheel is lifted off of the ground is close to the no load speed - I.e. Absolute Maximum Speed the motor is capable of. At that point there is virtually zero torque and therefore virtually zero power and zero current - back e.m.f. is very close to the battery voltage - it is much faster than the wheel can achieve with a rider on it so there should be no danger of that accidentally happening with a rider.

Could maybe happen on a decline where one could drive down almost without load for the motor to accelerate and no need for the motor to break. But imho just a theoretical probability.

21 minutes ago, Keith said:

Theoretically, at least, it might be possible that a rider is travelling fast, goes over a bump or kerb, becomes airborne and the wheel then speeds up to the point where the safety cut off happens. By the time the rider hits the ground again the wheel is powered down. Should that occur in any make of wheel it ought to be seen as a firmware issue, I.e. If an EUC experiences that behaviour the programme needs to be adjusted to increase the delay before the wheel shuts down, ensuring the rider is back on the ground before the firmware can cut power. I'm not at all sure I've seen any conclusive evidence this is happening?

Imho there where long time ago reports from airwheels shutting of in the air from people jumping down from from quite some height... If i remember right this was also on a youtube video?

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...