Speed vs Torque -- Margin of Safety

[dmethvin]

There was some discussion in the "I plan to test" thread about how to make electric unicycles safer. It has been many years since I took engineering but I do remember something like this graph, taken from an MIT course on DC motors:

From that graph you can see that once the motor is at its top speed, there is no more torque to be had. You can't just apply more battery power to the motor to give it a little extra speed or torque, because the torque-speed-power curve was decided when the motor was designed and manufactured. At or near the top speed, there is also no torque remaining to create warning measures like inclining the pedals.

 

So as far as I know, the only way to create a margin of safety is to "hold back" some of the top speed in order to use the remaining torque for maintaining balance. The challenge is that a 50kg rider requires much less torque for the same balance safety margin as a 120kg rider. Perhaps it would be really useful to have a unicycle that let you set that with an app, since it seems pretty difficult for the device to determine that by itself.

 

That graph also has bad news regarding the tradeoff between speed, torque, and battery life. You can't have it all! Right now a 500Wh battery is really heavy, so we all want smaller and lighter batteries. But we can't get that without sacrificing either speed or torque. Plus, if the manufacturer gives up torque to get more speed they will still need to hold back from maximum speed in order to have enough torque left for a safety margin.

 

When traveling on my Firewheel at speed with wind noise, or especially on noisy streets, the audio warnings are useless. Using power to incline the wheel way in advance of the danger zone seems like a waste of precious torque. Maybe there could be some kind of Bluetooth device that clipped on to a visor or glasses that just had a color LED to tell you how close you were to the edge?

[Gimlet]

This is the exact point I've been trying to get over to vee73. Gotway now have the very useful speed setable vibration alert to your phone over blue tooth that helps a lot but I'm sure it could be improved on.

Basically as it stands you need to find your own limits by taking it to the limit and beyond and then decide your own safety margin and set the speed for the vibration alert.

[CampGareth]

On your final point, getting power monitoring into a wheel wouldn't be that tricky, just have a microcontroller reading one of the motor's phases and smoothing it out, then reporting that over bluetooth. After that I'm out of my comfort zone, report it to an app on your phone, have that speak via bluetooth headset/wired headphones and update your smartwatch? If Google Glass were still available that'd be pretty great for a speed readout along with directions. Not sure how speed limiting would work or what else we want to monitor, thoughts?

[Gimlet]

Speed limiting by its very nature needs to act early to be sure of being reliable. That's why pedal tilt starts so early.

It would be nice if someone who has more skills than I do could use the online app maker to port the Gotway vibration warning across to a Pebble watch.

[hobby16]

From that graph you can see that once the motor is at its top speed, there is no more torque to be had. You can't just apply more battery power to the motor to give it a little extra speed or torque, because the torque-speed-power curve was decided when the motor was designed and manufactured. At or near the top speed, there is also no torque remaining to create warning measures like inclining the pedals.

True, but since we don't know the motor's curve, we don't know where the wheel's top speed is compared to the the motor's theorical top speed.

I suspect that even at 30km/h, a direct-drive (gearless) motor is still far from the zero torque area : the fact of life is it's much easier to make high speed-low torque motors than low speed-high torque motors, so on monowheels, motor's max speed is rather high compared to wheel's max speed (where software limitations has been imposed, for dubious reasons IMHO).

 

The question is : why trying to warn user at near-top speed (by pedals incline, sound, and even... cut off !) instead of simply keeping speed steady ? Why not just let it be, like on any other vehicles ?

If I'm right, it's because the pedals will end up inclining (without the rider aware of it) until they reach ground. If that's the only reason, I persist to say it's still much safer to keep speed steady at near-top speed than cutting-off power.

[hobby16]

On your final point, getting power monitoring into a wheel wouldn't be that tricky, just have a microcontroller reading one of the motor's phases and smoothing it out, then reporting that over bluetooth. After that I'm out of my comfort zone, report it to an app on your phone, have that speak via bluetooth headset/wired headphones and update your smartwatch? If Google Glass were still available that'd be pretty great for a speed readout along with directions. Not sure how speed limiting would work or what else we want to monitor, thoughts?

Wheel's already have a current sense circuitry, not on one motor's phase but on the power lines. I have a Gotway's motherboard right under my eyes, it uses the Allegro™ ACS709 current sensor IC.

So power is monitored indeed (you can even see it real-time on a Ninebot One android app).

[Gimlet]

I suspect that even at 30km/h, a direct-drive (gearless) motor is still far from the zero torque area : the fact of life is it's much easier to make high speed-low torque motors than low speed-high torque motors, so on monowheels, motor's max speed is rather high compared to wheel's max speed (where software limitations has been imposed, for dubious reasons IMHO).

 

I am aware of the greater problems of producing torque with a direct drive wheel but if they had all the extra speed available as you suspect why would my high torque version of the M18 with its extra windings on the coils have a top speed of 28kph or there abouts as oposed to the higher top speeds of both the medium and high speed models with less windings on the motor coils?

As far as I am aware the control boards are the same and we know for sure that Gotway are not speed averse or overly worried about rider safety, which is great as far as I am concerned. It's a case of let the buyer beware!

[EUC Extreme]

I follow with great interest the debate. But my information is not sufficient for this participation. In addition, there is a small language barrier, which is due to me. Keep up .. 

[Gimlet]

Your command of the English language vee73 is far better than mine is of Finnish/Suome. That's despite having a Fin exchange student stay with us for a month or two.

[dmethvin]

Perhaps the problem is that the electric unicycle makers are mostly taking off-the-shelf parts rather than designing their own custom components. When that happens there will be times when we don't get the best match out of the EU controller, BMS, and motor together. The issue that hobby16 mentions with the shutdown at top speed seems like one example of that.

 

I feel like Gimlet does, if we have the choice between the super-safe Solowheel and the rider-beware Firewheel or Gotway, I am okay with that. It would just be nicer if those wheels did not try to actively kill you. 

[hobby16]

I am aware of the greater problems of producing torque with a direct drive wheel but if they had all the extra speed available as you suspect why would my high torque version of the M18 with its extra windings on the coils have a top speed of 28kph or there abouts as oposed to the higher top speeds of both the medium and high speed models with less windings on the motor coils?

I don't know (and I'm proud of it ) and I certainly would like to know.

According to what you said, I suspect (again, mere speculation) that at "high" speed, the emf generated by the motor is  high enough compared to the battery voltage that there is too low voltage differential to inject power into the motor anymore. But again, at 30km/h, the rotational speed for a 14" wheel is only 448 rpm (348 rpm for a 18" inch), I don't consider that very high speeds, even for a 24 poles motor. Another possible explanation is the power stage of the controller board, which is not so oversized for obvious cost reasons. In fact there are many culprits out there, the motor is only one suspect.

 

BTW, I don't know where you get that a "high torque GW has extra windings on its coils" but if your information source is technically reliable, he IS the one for our questions to be asked.

[hobby16]

I feel like Gimlet does, if we have the choice between the super-safe Solowheel and the rider-beware Firewheel or Gotway, I am okay with that. It would just be nicer if those wheels did not try to actively kill you. 

The neat thing on Firewheels is the level indicator. When riding at high speeds or accelerating on steep slopes, you can see the huge voltage drop (you can shortly get from 50% level down to 0% level then back to 50% again!), a definitive proof the speed limitation comes from the battery internal resistance and wirings (cables' diameter is IMHO rather small).

So I'm not even sure that with a shunted BMS, there would be a top-speed cut-off (what is interpreted by users as a cut-off is in fact a cut-off by the BMS because of undervoltage, not by the controller because of overspeed). I have shunted my BMS but I have never pushed over 35 km/h to tell, but someday, I'll try.

For the moment, unless proved otherwise, my interim conclusion is that on Firewheels, there is no overspeed cut-off by the controller (and "nearly" no pedal incline warning). Which is a really nice thing. So bad, you can't enjoy your FW, it's really a nasty little beast   

[Gimlet]

Gilles, there are three models in total.  Fast, balanced, and high-torque.  The fast has a top speed of 40km/h but a top sustained hill climb of 27 degrees.  The balanced model has a top speed of 34km/h but has a top sustained hill climb of 33 degrees, and the high-torque model has a top speed of 28km/h and has a sustained hill climb of 39 degrees.  These are measurements in "degrees" not "percent slope" so is very impressive.  Each wheel has a different motor, or rather each version's stator has a different number of wire wounds around it to determine the torque vs top speed issue.  It is hardware and software difference.

  

The neat thing on Firewheels is the level indicator. When riding at high speeds or accelerating on steep slopes, you can see the huge voltage drop (you can shortly get from 50% level down to 0% level then back to 50% again!), a definitive proof the speed limitation comes from the battery internal resistance and wirings (cables' diameter is IMHO rather small).

So I'm not even sure that with a shunted BMS, there would be a top-speed cut-off (what is interpreted by users as a cut-off is in fact a cut-off by the BMS because of undervoltage, not by the controller because of overspeed). I have shunted my BMS but I have never pushed over 35 km/h to tell, but someday, I'll try.

For the moment, unless proved otherwise, my interim conclusion is that on Firewheels, there is no overspeed cut-off by the controller (and "nearly" no pedal incline warning). Which is a really nice thing. So bad, you can't enjoy your FW, it's really a nasty little beast 

The Information came from the earlier thread about the M18.

[CampGareth]

The neat thing on Firewheels is the level indicator. When riding at high speeds or accelerating on steep slopes, you can see the huge voltage drop (you can shortly get from 50% level down to 0% level then back to 50% again!), a definitive proof the speed limitation comes from the battery internal resistance and wirings (cables' diameter is IMHO rather small).

So I'm not even sure that with a shunted BMS, there would be a top-speed cut-off (what is interpreted by users as a cut-off is in fact a cut-off by the BMS because of undervoltage, not by the controller because of overspeed). I have shunted my BMS but I have never pushed over 35 km/h to tell, but someday, I'll try.

For the moment, unless proved otherwise, my interim conclusion is that on Firewheels, there is no overspeed cut-off by the controller (and "nearly" no pedal incline warning). Which is a really nice thing. So bad, you can't enjoy your FW, it's really a nasty little beast 

Damn, I'm just about to buy an F528 after seeing a guy riding around effortlessly on an Airwheel today (waiting for supplier to turn an email quote into an alibaba quote). Is the cable width printed on the sheathing? A quick trip to online calculators says it should be 24AWG or lower (lower being thicker). I'd also like to see if the controller board has that same power sensing chip on it.

[hobby16]

Damn, I'm just about to buy an F528 after seeing a guy riding around effortlessly on an Airwheel today (waiting for supplier to turn an email quote into an alibaba quote). Is the cable width printed on the sheathing? A quick trip to online calculators says it should be 24AWG or lower (lower being thicker). I'd also like to see if the controller board has that same power sensing chip on it.

Most of boards I've seen have shunt resistors to sense current (power is calculated = current x voltage). The Firewheel's board too. GW is rather unique in having a dedicated current sense IC (but the PCB has a double layout for a shunt resistor based sensing scheme).

 

I didn't pay attention to the AWG marking on the cables but I felt they are thin, an impression confirmed by the voltage sag when accelerating. Another Firewheeler even had some torque problem (he could not go as fast as me), then he dismounted the wheel and realised the power cable was unadvertently pinched between the two half enclosures! (clean manufacturing and attention to details are not strong points of Firewheel). He replaced the cable and the problem was solved, but that concurs corners have been cut as to wire diameters.

 

The F528 is heavy but you'll really enjoy it ... as a body building toy. No, just kidding!

Firewheels rock as long as you don't forget to insulate, caulk, and shunt the BMS.

[MarkoMarjamaa]

On your final point, getting power monitoring into a wheel wouldn't be that tricky, just have a microcontroller reading one of the motor's phases and smoothing it out, then reporting that over bluetooth. After that I'm out of my comfort zone, report it to an app on your phone, have that speak via bluetooth headset/wired headphones and update your smartwatch? If Google Glass were still available that'd be pretty great for a speed readout along with directions. Not sure how speed limiting would work or what else we want to monitor, thoughts?

I own a Airwheel myself.

I wrote a small app for reading GPS and reading out loud the speed. Actually it does not read speed but difference to target speed.

Problem with low Airheel speeds is that the GPS reading is not accurate enough ( it's not using Doppler data ).

 

https://play.google.com/store/apps/details?id=appinventor.ai_marko_marjamaa_2.RecordBreaker

Sorry for price but I need my coffee money....

[Shady Tools]

Ad section?

[MarkoMarjamaa]

Ad section?

Well if I'm the only one that knows about that app (I just released it and tested it with Airwheel) and someone asks that kind of app, should I wait months or year until someone else here knows about that app and reads this thread and is kind enough to mention it?

Or to put it  another way, is 1 euro worth fighting for?

[ironfox]

Well if I'm the only one that knows about that app (I just released it and tested it with Airwheel) and someone asks that kind of app, should I wait months or year until someone else here knows about that app and reads this thread and is kind enough to mention it?

Or to put it  another way, is 1 euro worth fighting for?

Why would you need such an app? There are already such apps... And apart from that, this is more than off-topic. It has nothing to do with "speed vs. torque"

[Gimlet]

It might have been more acceptable if it wasn't your very first post on the forum!

[MarkoMarjamaa]

It might have been more acceptable if it wasn't your very first post on the forum!

Sorry

I've had my Airwheel from November. Found this forum a week ago.

[antberz]

Hobby 16: I have a Firewheel F260- loving it for commuting been riding it about 15-20km per day for the last few months - way faster than other wheels I have ridden..

Please could you tell me more re your comment about "insulate, caulk, and shunt the BMS" - should I be concerned about this? It all looks pretty well sealed up in the inside drum that houses the components.. Also, is it easy to disconnect the speed warning? I comfortably ride it more in this zone than below it..

[antberz]

I don't know (and I'm proud of it ) and I certainly would like to know.

According to what you said, I suspect (again, mere speculation) that at "high" speed, the emf generated by the motor is  high enough compared to the battery voltage that there is too low voltage differential to inject power into the motor anymore. But again, at 30km/h, the rotational speed for a 14" wheel is only 448 rpm (348 rpm for a 18" inch), I don't consider that very high speeds, even for a 24 poles motor. Another possible explanation is the power stage of the controller board, which is not so oversized for obvious cost reasons. In fact there are many culprits out there, the motor is only one suspect.

 

BTW, I don't know where you get that a "high torque GW has extra windings on its coils" but if your information source is technically reliable, he IS the one for our questions to be asked.

 

Most of boards I've seen have shunt resistors to sense current (power is calculated = current x voltage). The Firewheel's board too. GW is rather unique in having a dedicated current sense IC (but the PCB has a double layout for a shunt resistor based sensing scheme).

 

I didn't pay attention to the AWG marking on the cables but I felt they are thin, an impression confirmed by the voltage sag when accelerating. Another Firewheeler even had some torque problem (he could not go as fast as me), then he dismounted the wheel and realised the power cable was unadvertently pinched between the two half enclosures! (clean manufacturing and attention to details are not strong points of Firewheel). He replaced the cable and the problem was solved, but that concurs corners have been cut as to wire diameters.

 

The F528 is heavy but you'll really enjoy it ... as a body building toy. No, just kidding!

Firewheels rock as long as you don't forget to insulate, caulk, and shunt the BMS.

I have a Firewheel F260- loving it for commuting been riding it about 15-20km per day for the last few months - way faster than other wheels I have ridden..

Please could you tell me more re your comment about "insulate, caulk, and shunt the BMS" - should I be concerned about this? It all looks pretty well sealed up in the inside drum that houses the components.. Also, is it easy to disconnect the speed warning? I comfortably ride it more in this zone than below it..

[dmethvin]

antberz, I added some more detail about sealing the Firewheel in 

[GQS]

BTW, high speed has LESS coils in the stator, high torque has MORE coils.  More coils = more torque = more heat at higher speed = lower top speed limit. The current/power required to spin the motor at the very high speeds will cause more heat for a motor with more windings such that it may cause it to burn up.  So higher torque = lower top speed.