To brake or not to brake - that is the question

[The Fat Unicyclist]

15 minutes ago, MaxLinux said:

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

I second that - where I live I am changing elevation by 100-150m on a daily basis, and while the downhill is the scariest bit, I have never had an actual problem.

[Keith]

5 hours ago, Blechi said:

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)

I mentioned the first in my previous post: works on model aircraft, not on a wheel where all that energy has to go somewhere.

The second is a description of regenerative braking, the motor generates a higher voltage than the battery is putting out so current flows the opposite way. If you lean back until you stop and keep leaning back you accelerate backwards, that isn't some clever electronic switch, it's the wheel continuing to do what it had been doing since you started braking.

[who_the]

Sorry to hear about your issues; obviously, even the slightest issue with our wheels can get us unnerved, since we depend on them for our lives.

As you've discovered, the beeping indicates an "overcharge" situation. It's a feature, not a bug, of King Song wheels. The same thing happened to me once on my KS-14C while riding on a mild downhill slope on an absolutely full charge.

Be thankful for the beeping, lesser wheels might have just failed. Or worse.

Here in San Francisco, I've successfully ridden down (and up) hills steeper than most will ever see. To be clear, my KS-14C has never had so much has a hiccup other than my aforementioned — and obvious — "overcharge" situation. Over several hundreds of miles on hilly thoroughfares with reckless vehicle traffic, I have more confidence in my King Song than any other wheel I've ever ridden.

Solution: Charge your wheel to 80-90 percent at most, or give it a good ride before riding downhill. Really that simple. Unless you're riding 30-40 km/day, you won't need anywhere near a full charge. As you've stated, in a simple system, the juice has to go somewhere. And King Song chooses to warn us (probably conservatively, based on my experience with their other warnings) rather than just wipe us out.

 

[johrhoj]

I think that it is well established in this topic that the problem is not overcharging the battery, but either overloading the motherboard, heating problem, or max torque issue. Would it not be nice if you had different warnings, so you can distinguish between them and act accordingly?

I can imagine that an emergency break could be built into the wheels, a mechanical break. The wheel would need to detect impending hardware failure and the runaway situation. However, it will invariably throw you off the wheel. Before mechanical breaking, the wheel could tip you over backwards, thus sitting you down on the street in stead of faceplanting you. A distinct warning before doing would be much appreciated. 

The mechanical break would add to the weight of the wheel considerably. 

[Frode]

6 hours ago, MaxLinux said:

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

The less common, the larger the surprise when it happens

I don't know how common it is. I have only experienced it this single time. If you want to try your wheel, charge your wheel over night to ensure 100% charged, find a steep hill and start from the top with the newly charged battery. I have 3 advices though: First of all: Start slow! Secondly: Remember to start slow and last but not least: Start slow!

I have read about it in this forum (don't remember where). That is why I lowered my speed before going down, even if I had driven the wheel about 1,5 km before the donwhill which must have used some charge (~0,5 km was uphill).

I don't want to over empasize this tough, and make a big problem out of it, but I do tend to get more carefull the more I use the wheel and learn how it behaves...

[Frode]

2 hours ago, who_the said:

Be thankful for the beeping,

I am!

2 hours ago, who_the said:

or give it a good ride before riding downhill. Really that simple.

That is my strategy also! In this situation I actually thought it would have been enough to ride 1,5 km including 0,5 km uphill, but it wasn't quite enough. Maybe it has to do because I am "slightly" heavy (but still below the specified maximum).

I have been to San Fransisco once (back in the 90's) and seen the hills. Would have loved to test my wheel there.

2 hours ago, who_the said:

And King Song chooses to warn us (probably conservatively, based on my experience with their other warnings) rather than just wipe us out.

Where are the warnings? There is nothing in my user manual as far as I can see (other than very general ones). Maybe there was a booklet I overlooked and threw out with the box in my eagerness to get started...

[Frode]

Thank you everybody for your contributions/clarifications/questions in the discussion. I think I gained some insight and also some riding techniques to try out. The question is no longer wether to break or not to break, but "to jump off or not to jump off" and how to help the wheel to break in other ways if you didn't jump! 

[Keith]

4 hours ago, johrhoj said:

I think that it is well established in this topic that the problem is not overcharging the battery, 

Well, I for one don't think any such conclusion has been reached.

4 hours ago, Frode said:

If you want to try your wheel, charge your wheel over night to ensure 100% charged, find a steep hill and start from the top with the newly charged battery.

And @Frode doesn't think so either.

3 hours ago, Frode said:

 and how to help the wheel to brake in other ways if you didn't jump! 

Slaloming or even turning side on to the hill, if there is room, are going to be best and slaloming will help reduce the load on the electronics as well even if not too steep a hill

There is a true irony in a long discussion on something that risks injury, and passive acceptance of a statement that risks death. Even though the KingSong batteries and charger are good quality and the risks may be very small, please take a look at the start of the video on the below link where the fire is a laptop, on charge in a departure lounge, a device that should have significantly higher safeguards and is made in the millions. I personally would ABSOLUTELY NEVER leave a large lithium battery charging in the house whilst I am asleep. I'm way, way more willing to accept the current designs going down a steep hill than I will ever be to do that.

It is another really good argument for the fast charger. An 840Wh battery is more than 14 amp hours so even a 5amp charger is well under the recommended 0.5C charge rate. A 2 amp charger takes so long it almost forces users with large batteries to leave it on overnight.

 

[johrhoj]

4 minutes ago, Keith said:

 

4 hours ago, johrhoj said:

I think that it is well established in this topic that the problem is not overcharging the battery, 

Well, I for one don't think any such conclusion has been reached.

 

At the start of the topic it is clearly said that some distance and some height already was covered, before encountering the short steep downhill. I did do the math in the topic EUC Math - question #1. The amount of recharge you get from an occasional downhill is not as big as you might think, and going uphill-downhill all the time reduces your range by roughly one third (e.g. from 30km to 20km).

Of course it is possible to overcharge the battery when starting to go downhill with a full battery, but that was clearly not the case. My first wheel was a Ninebot One E+ and in the manual was explicitely stated that one should not do such a thing. So, please do not try this. It may harm you, but also your batteries. 

I said that overcharging the battery was not the problem. When braking long and hard you could still have a charging problem in the sense that the regenerated power is more than the battery can absorb in the same amont of time. This will lead to overheating, possibly frying components on the motherboard (or blowing a fuse). 

[HEC]

4 hours ago, Frode said:

Where are the warnings? There is nothing in my user manual as far as I can see (other than very general ones). Maybe there was a booklet I overlooked and threw out with the box in my eagerness to get started...

I think @who_the is referring to actual warnings during ride (beeps / tilt backs) rather than printed warnings in manual

[HEC]

39 minutes ago, johrhoj said:

I said that overcharging the battery was not the problem. When braking long and hard you could still have a charging problem in the sense that the regenerated power is more than the battery can absorb in the same amont of time. This will lead to overheating, possibly frying components on the motherboard (or blowing a fuse). 

I disagree - the voltage drop during the initial discharge when battery is topped to the rim is minimal so even after some short ride the controller board might still detect the battery as full resulting in "overcharge" warnings (beeps and tilt backs) while riding down the hill despite battery actually having a bit of "space" left there to consume the charge. It's even more pronounced with larger battery capacity packs in 4P configuration which is the case with OP (840Wh pack). It's all down to "simplified" firmware running on minimalistic HW (nothing wrong with that in general = less chances of failures) not taking in consideration actual Wh / Ah consumptions or re-charging or the battery pack capacity (as it's not as simple to detect by software just by battery voltage). Basically the FW behaves the same way / treats the various battery pack sizes or even different states of their aging equally resulting thus in putting in "disadvantage" or "over-protecting" users of higher capacity and / or new battery packs to accommodate the worst case scenarios safely (e.g. aged 1P packs). This has been already discussed in multiply FW threads mainly relating to all various speed limitations based on battery capacity and so on ...

Bottom line - don't charge you battery to the top unless really needed for a planed extra long ride (especially if you have a high capacity models) or unless you need to perform a balancing charge once a while (every 10 - 15 max 20 charges).

[johrhoj]

9 minutes ago, HEC said:

I disagree - the voltage drop during the initial discharge when battery is topped to the rim is minimal so even after some short ride the controller board might still detect the battery as full resulting in "overcharge" warnings (beeps and tilt backs) while riding down the hill despite battery actually having a bit of "space" left there to consume the charge.

That sounds pretty plausible  . If this is the case, it is my lack of understanding how the firmware and hardware reacts to the situation.

Firmware: I would think that it is not the voltage of the battery that solely indicates how full the battery is. Surely the system would know the traveled distance (or even, more accurate, the consumed energy) since charging stopped? 

Hardware: Even when the above is false, the hardware would be wired in such a way that the energy surplus is offered to the battery, full or not?

As you see, my conclusions were based on assumptions that may not be accurate. The bottomline is that I do not know for sure. It may also differ from wheel to wheel. I anyone knows whether my assumptions are true or false, please let me know. 

[HEC]

21 minutes ago, johrhoj said:

Firmware: I would think that it is not the voltage of the battery that solely indicates how full the battery is. Surely the system would know the traveled distance (or even, more accurate, the consumed energy) since charging stopped? 

Hardware: Even when the above is false, the hardware would be wired in such a way that the energy surplus is offered to the battery, full or not?

Sadly it seems that waste majority (if not all) of current EUC firmwares is using only the battery voltage at any given moment to decide what to do. So no any smart averaging or calculating based on the consumption, speed, distance etc. It's both down to already mentioned simplified SW as well as HW and also due to the legacy / inheritance from e-bikes where this was not really an issue as the health or even life of rider was not in danger there.

While hardware will indeed still try to push the current back to battery pack the wheel's firmware and controller board will still initiate the warning / safety protocols based on voltage as those needs to kick in before the battery is full and unable to take any more charge. If the wheel / motor will keep pushing the excessive voltage / current into the battery pack either the BMS will eventually cut it off or pack will get damaged (worst case catch fire) with both scenarios resulting in rider's faceplant.

[US69]

58 minutes ago, HEC said:

Sadly it seems that waste majority (if not all) of current EUC firmwares is using only the battery voltage at any given moment to decide what to do. So no any smart averaging or calculating based on the consumption, speed, distance etc. It's both down to already mentioned simplified SW as well as HW and also due to the legacy / inheritance from e-bikes where this was not really an issue as the health or even life of rider was not in danger there.

 

As (all?) EUC have only Powercable going from BMS to board, there is no way of more communication, thats correct....

I think only Inmotion is going a "new" way with at least the V8? As can be seen in the Video it is one of the first wheels, where the Batterie pack has on more cableset than power/charging...there is a set of cables from the pack to the Motherboard. I have no clue what it is doing...but at least there is something happening :-)

[dmethvin]

8 minutes ago, KingSong69 said:

As (all?) EUC have only Powercable going from BMS to board, there is no way of more communication, thats correct....

I think only Inmotion is going a "new" way with at least the V8? As can be seen in the Video it is one of the first wheels, where the Batterie pack has on more cableset than power/charging...there is a set of cables from the pack to the Motherboard. I have no clue what it is doing...but at least there is something happening :-)

The early Firewheels had a line going from the battery pack to a pin on the control board marked "BMS". However, in later models (including the one I have) they just tied that BMS pin to ground. I don't know whether it was a data line or a simple high/low signal of some sort, but knowing the Firewheel it probably just shut off the wheel to save the components. On those early models I think their slogan was "Wounds heal, electronics don't!"

[HEC]

1 hour ago, KingSong69 said:

As (all?) EUC have only Powercable going from BMS to board, there is no way of more communication, thats correct....

I think only Inmotion is going a "new" way with at least the V8? As can be seen in the Video it is one of the first wheels, where the Batterie pack has on more cableset than power/charging...there is a set of cables from the pack to the Motherboard. I have no clue what it is doing...but at least there is something happening :-)

Yes - but controller board could (if programmed accordingly and HW would be able to run advanced program like that) make some calculations based on at least voltage and current (as those are both measured by board already anyway) over certain period of time or make it even more advanced by including the distance and so on. This would though require some more calculation "power" and data storage for which the current controller boards / "CPUs" might not be sufficient.

In regards of V8 - the additional 5 pin lead from battery pack is also present in V5F models and (as already theorised in @Jason McNeil's thread) it might serve some so far unknown / unclear purpose including the better firmware awareness of the battery status. After posting my "elaboration" on that there I've actually came with one more possible explanation which would be even quite easy to verify: if I'm not mistaken V5F shows the battery charging status even when the wheel is turned off so maybe this 5 pin cable is used for that? By unplugging it we'll see if that's the case or not

[Jason McNeil]

1 hour ago, HEC said:

I'm not mistaken V5F shows the battery charging status even when the wheel is turned off so maybe this 5 pin cable is used for that?

That's possible, although arguably the CB could just take a reading of the voltage from the XT60, no need for additional wires/connectors. Also these wires are connected to the control-board rather than four-bar power meter.

I'll put the question to IM, see if they can enlighten us. 

[HEC]

41 minutes ago, Jason McNeil said:

That's possible, although arguably the CB could just take a reading of the voltage from the XT60, no need for additional wires/connectors. Also these wires are connected to the control-board rather than four-bar power meter.

I'll put the question to IM, see they can enlighten us. 

That was indeed just a wild guess on my side based on assumption that it might be needed / more safe or easier to feed the power meter separately while the wheel is turned off / in sleep mode. Hopefully IM will shed some light on this (unless it's some deep corporate secret ).

[Frode]

5 hours ago, HEC said:

I think @who_the is referring to actual warnings during ride (beeps / tilt backs) rather than printed warnings in manual

Ahhh - of course! Thank you for the clarification!

[esaj]

On 12.8.2016 at 7:10 PM, dmethvin said:

The early Firewheels had a line going from the battery pack to a pin on the control board marked "BMS". However, in later models (including the one I have) they just tied that BMS pin to ground. I don't know whether it was a data line or a simple high/low signal of some sort, but knowing the Firewheel it probably just shut off the wheel to save the components. On those early models I think their slogan was "Wounds heal, electronics don't!"

I had one of the older boards with the BMS data(?) wire. If left floating, the wheel would start repeating something like "Low battery or bad cell in battery" and wouldn't engage the motor.