Electric conversion Ciao no power

[Bogdan_128]

Hello all and sorry if I posted this in the wrong place.

I am extremely close to finalizing my Piaggio Ciao electric conversion and during testing i encountered an issue. The old petrol engine was replaced with a 160kv drone sensorless bldc motor, a tmotor flame 60a esc and a custom 36v10s6p battery (cells picked to be able to output more current than required) with a Chinese Daly 60A bms. A few weeks ago i finished a "just to test the motor" wiring with an arduino generating the pwm for the controller based on analog potentiometer imput and the motor spun up to life right up. I then bought a pulley, made the transmission, mounted the motor and decided to go on a first test ride (forgetting that i used 16a rated wires between the esc and the motor). It was going great, pulling very hard once you gained about 5km/h by pedals as the motor as espected has no power from a standstill. I got too excited and managed to heat the motor to about 90C (what i measured with my IR temp gauge a few minutes after the power cut off) and melt the 16a rated wires I used in my "just to test" setup between the esc and the motor. The controller was a bit warm to the touch but not dramatic (40-45C). Today i finished the final wiring and with all pieces into place i wanted to test ride again but surprise surprise it barely ran. If i managed to match by pedals the lowest speed setting on the motor and then EXTTREMELY slowly increase throttle i would get to about 50% throttle before it would cut off and require a standstill start again. What could be the problem? Could I have melted the windings on the motor ? Is the esc fried? A small prop is on its way to be mounted to the motor to cool it off. Also worth mentioning is that the motor overheated when mounted to a plastic 3d printed test part. Today with the metal bracket i managed to go about 2km with the motor being barely warm. Another weird thing is that with the wheel in the air the motor manages to go to full speed in just a second or two no problem, but as soon as the wheel touches the ground, it cuts off.

Any ideas? Thanks for reading the big block of text

 

[99t4]

Moving this thread to EScooters/EMotorcycles.

 

[amberwolf]

If the temperature was still 90C a few minutes later, it's possible the magnets were heated much more than that and have changed properties. If they were weakened the motor may now spin faster than it used to when unloaded, so you can test for this to see if it's kV (rpm/volt) has changed.

If this did happen, the motor will have proportionally lower torque and not be able to push as hard.


If hte windings are damaged, you will be able to see darkening of them; most of the drone type motors I've seen have open frames you can see them thru.


It takes very little current / power to spin a motor up when it's unloaded and properly driven by the controller, so it's not surprising that it will spin up correctly offground. If it's a torque-based controller (throttle modulates current, not speed), then it only takes a tiny bit of throttle to cause full speed when offground, as well.


FWIW, most of these types of motors ahve a high kV, so they require a huge reduction ratio to a wheel to drive them. For instance, 36v x 160kV = 5760RPM. Wheel RPMs vary depending on the speed you want to go and the wheelsize, but are usually in the range of 1/20th of that, which would mean a reduction ratio of 20:1 to do this.

What is your reduction ratio? Is it correct ratio to drive your wheel offground at approximately the same speed you want to ride at? (typically loaded motor speed will end up around 80% of the offground speed, but this is still in the ballpark of "being the same").


Also, they are designed to spin up very quickly to their full commanded speed, and be cooled by the massive amounts of air flwoing past them from their props. This isn't going to happen even with a prop on them when used on a ground vehcile to drive a wheel, so depending on your use case you might need a forced-air cooling system that keeps a fan moving air thru it even when it is at low speeds or even stopped, as those are the highest-load conditions in many cases, and cause more waste heat in the motor that a prop on the motor itself cannot help get rid of (because it's not spinning fast enough to mvoe the air).

 

[neptronix]

Do you have a realtime view of the amps and watts your controller is pulling?

If so do you notice the voltage dropping massively before it cuts off?

What AW said about these kind of motors is totally true, whatever this peak power rating they sell RC motors at is usually considering you have 100mph of air coming into the motor and the motor is only doing that for like.. a second.

So a 5kw motor is more like a 2kw or so continuous motor.
Can you provide a link to the motor you bought?

I also suspect you might have too tall of a gearing ratio..

 

[Bogdan_128]

If the temperature was still 90C a few minutes later, it's possible the magnets were heated much more than that and have changed properties. If they were weakened the motor may now spin faster than it used to when unloaded, so you can test for this to see if it's kV (rpm/volt) has changed.

If this did happen, the motor will have proportionally lower torque and not be able to push as hard.


If hte windings are damaged, you will be able to see darkening of them; most of the drone type motors I've seen have open frames you can see them thru.


It takes very little current / power to spin a motor up when it's unloaded and properly driven by the controller, so it's not surprising that it will spin up correctly offground. If it's a torque-based controller (throttle modulates current, not speed), then it only takes a tiny bit of throttle to cause full speed when offground, as well.


FWIW, most of these types of motors ahve a high kV, so they require a huge reduction ratio to a wheel to drive them. For instance, 36v x 160kV = 5760RPM. Wheel RPMs vary depending on the speed you want to go and the wheelsize, but are usually in the range of 1/20th of that, which would mean a reduction ratio of 20:1 to do this.

What is your reduction ratio? Is it correct ratio to drive your wheel offground at approximately the same speed you want to ride at? (typically loaded motor speed will end up around 80% of the offground speed, but this is still in the ballpark of "being the same").


Also, they are designed to spin up very quickly to their full commanded speed, and be cooled by the massive amounts of air flwoing past them from their props. This isn't going to happen even with a prop on them when used on a ground vehcile to drive a wheel, so depending on your use case you might need a forced-air cooling system that keeps a fan moving air thru it even when it is at low speeds or even stopped, as those are the highest-load conditions in many cases, and cause more waste heat in the motor that a prop on the motor itself cannot help get rid of (because it's not spinning fast enough to mvoe the air).

Hi, amberwolf!

Thanks for the answer!

I don't know my reduction ratio exactly, the Piaggio Ciao has a reduction built into the rear wheel and the belt ratio is very similar to the old clutch-pulley ratio. I figured that if a 1hp petrol engine with similar rpm range could move the scooter so can this 2.3kw bldc. What I noticed about the motor was that before the event it was sort of clicky when turning it by hand with no load, nowadays it feels way less clicky which leads me to believe that your theory with demagnetization might be very true. The motor now is mounted to a 2.5mm 200x100mm steel plate which i believe pulls enough heat from the motor and passes it to the chassis as after mounting this component the motor barely warmed up and i rode the scooter for quite a few km in this limp mode state. Aluminium would have been better but I wasn't able to find a plate thick and big enough for this application. I will open up the motor and check for darkening, but i think the magnets are the problem now that you mentioned.

 

[Bogdan_128]

Do you have a realtime view of the amps and watts your controller is pulling?

If so do you notice the voltage dropping massively before it cuts off?

What AW said about these kind of motors is totally true, whatever this peak power rating they sell RC motors at is usually considering you have 100mph of air coming into the motor and the motor is only doing that for like.. a second.

So a 5kw motor is more like a 2kw or so continuous motor.
Can you provide a link to the motor you bought?

I also suspect you might have too tall of a gearing ratio..

The esc doesn't support telemetry afaik, but I will research more online. I will check for the voltage drop before the cutoff, but I don't think i would be able to take an oscilloscope with me on the bike I now checked and the motor is actually 200kv, sorry for the misinfo in the title, I was researching cheap drone motors for too long. I have this one. With the calculations done the motor would have a theoretical maximum of 36x200 = 7200rpm which is close enough to the ~8-10k rpm redline of the 49cc. I will attach photos of the setup to clarify a bit. If the magnets are the issue, as AW mentioned, would it be possible to replace them or is it just better to get a more suitable motor designed for continuous high load and with more adequate cooling.

 

[neptronix]

This motor looks like a maytech ( nice quality ) clone ( maybe not nice quality ).
That motor is probably undersized for the job.
Selecting the motor around the original RPM/ratios is a good idea. But maybe you want to lower the RPM and get a 150kv motor.

If your ESC is "60A", by RC ratings, that probably means 60A phase amps, not 60A battery amps.
60A phase amps = approx 20A battery amps if the controller is not tuned.
So unless you have a super weak battery, i think the battery might not be cutting out.

These are all best guesses because we don't have the tools to measure what's going on.

Cheap watt/amp/volt meter: https://hobbyking.com/en_us/turnigy-180a-watt-meter-and-power-analyzer.html?srsltid=AfmBOoo5zC9IwCqEcL6v5h2ImH44omS3Cy2aHBJrb_0K7h2tuBZarEm2

More fancy & much better meter: https://ebikes.ca/shop/electric-bicycle-parts/cycle-analysts/ca-sa.html

Cheap laser tachometer to help you figure out RPMs at the wheel:
Digital Tachometer Non Contact Laser Photo RPM Tach Meter Motor Speed Gauge AS | eBay

 

[neptronix]

BTW if magnets weakened, you'd see higher RPM at the wheel and less torque output.
But it sounds like you didn't take any baseline measurements of RPM, so telling how much they weakened is hard.

 

[Bogdan_128]

BTW if magnets weakened, you'd see higher RPM at the wheel and less torque output.
But it sounds like you didn't take any baseline measurements of RPM, so telling how much they weakened is hard.

Couldn't I assume that the 200kv rating with however much the battery voltage is while running could give me a baseline rpm? If I were to replace the motor, what should I look into? Should I also replace the ESC with a beefier one? I would like to keep it as it seems to be of quality and I don't think I would mind less power than expected for now at least. I will make myself a rpm measuring tool and will come back with updates. As for the amp measurement, I will use a shunt and measure the voltage drop on that to calculate the current. Here, in Romania the items you mentioned are not cheap or easy to find. Thanks a lot!

 

[amberwolf]

The esc doesn't support telemetry afaik,

Don't need anything from teh controller (though its' nice when it's available since it's often able to measure both battery and phase properties). Just use a wattmeter in between batteryyadf and controelrle, there;s losfs of cheap ones out there that work ok, some that do logging, some save data even after poweroff, some just do realitme dispaly so you have to watch it, soem have peak / min readings as well as realtime, etc.

but I will research more online. I will check for the voltage drop before the cutoff, but I don't think i would be able to take an oscilloscope with me on the bike

not sure what an oscilloscope would do for you there. just need a voltmeter and watch the screen during the test. or multimeter on volts setting. or wattemter.

I now checked and the motor is actually 200kv, sorry for the misinfo in the title, I was researching cheap drone motors for too long. I have this one. With the calculations done the motor would have a theoretical maximum of 36x200 = 7200rpm which is close enough to the ~8-10k rpm redline of the 49cc.

if you're driving the original crankshaft of the ice with the outupt shaft of the drone motor, then you get whatever reduction the ice had.

is that enough to allow the drone motor to spin very near it's theoretical speed at whatever voltage / throttle percentage it is getting? If not, it'll overheat because it doesn't have enough thermal mass to deal with buildup of heat--it requires huge airflow thru it for normal operation in it's intended use, and will need even more airflwo for your usage.

however loud the dronemotor prop would've been in it's original operation, it will probably be at least twice that loud, all the time (because you'll have to run it on a separate motor so it always keeps cooling the drone motor especially when the drone motor is slowed down by a load), keeping enough air moving over the dronemotor (and the motor that's driving the prop).

if you use a separate fan system that's just designed to mvoe that much air thru a duct, it might be quieter, or you could enclose it in baffling ductwork to reduce the soudn level but thats tought to do with a prop.

I will attach photos of the setup to clarify a bit. If the magnets are the issue, as AW mentioned, would it be possible to replace them or is it just better to get a more suitable motor designed for continuous high load and with more adequate cooling.

if you can find suitable replacmeents, but unless you know what strentght/etc the originals are, you can't match what it had except by accident.

Couldn't I assume that the 200kv rating with however much the battery voltage is while running could give me a baseline rpm?

yes, you've already done that math.

If I were to replace the motor, what should I look into?

depends on your drive system / reduction setup, your available space, budget, noise limitations, weight limit, power limit, and what the moped's usage is going to be under what specific riding conditions / terrain / etc, what kind of acceleration you're after, and so on. (the same applies to battery and controller needs)

if you take that info to the ebikes.ca motor simulator, you can customize it to help yuou find out what power level you need to do what you want. then you can find a motor that is mroe capablea than that, and find out if your contorller a nd battery can handle that too.

it'll also tell you about how many wh/mile you will use for various terrains you put in, and you can use that plus the actual distances on those terrains to figure out how many wh you need for the range your equire, which then lets you make sure your battery is big enough for that. then you should generally add at least 25% to that to leave you headroom for battery aging, or potential detours, etc.,

Should I also replace the ESC with a beefier one? I would like to keep it as it seems to be of quality and I don't think I would mind less power than expected for now at least.

if it works and does the cjob you need and has the features you want, no reason to change it.

I will make myself a rpm measuring tool

there's free phone apps to help with taht; don';t recall the specific names thoufgh

 

[Bogdan_128]

Don't need anything from teh controller (though its' nice when it's available since it's often able to measure both battery and phase properties). Just use a wattmeter in between batteryyadf and controelrle, there;s losfs of cheap ones out there that work ok, some that do logging, some save data even after poweroff, some just do realitme dispaly so you have to watch it, soem have peak / min readings as well as realtime, etc.





not sure what an oscilloscope would do for you there. just need a voltmeter and watch the screen during the test. or multimeter on volts setting. or wattemter.


if you're driving the original crankshaft of the ice with the outupt shaft of the drone motor, then you get whatever reduction the ice had.

is that enough to allow the drone motor to spin very near it's theoretical speed at whatever voltage / throttle percentage it is getting? If not, it'll overheat because it doesn't have enough thermal mass to deal with buildup of heat--it requires huge airflow thru it for normal operation in it's intended use, and will need even more airflwo for your usage.

however loud the dronemotor prop would've been in it's original operation, it will probably be at least twice that loud, all the time (because you'll have to run it on a separate motor so it always keeps cooling the drone motor especially when the drone motor is slowed down by a load), keeping enough air moving over the dronemotor (and the motor that's driving the prop).

if you use a separate fan system that's just designed to mvoe that much air thru a duct, it might be quieter, or you could enclose it in baffling ductwork to reduce the soudn level but thats tought to do with a prop.



if you can find suitable replacmeents, but unless you know what strentght/etc the originals are, you can't match what it had except by accident.




yes, you've already done that math.




depends on your drive system / reduction setup, your available space, budget, noise limitations, weight limit, power limit, and what the moped's usage is going to be under what specific riding conditions / terrain / etc, what kind of acceleration you're after, and so on. (the same applies to battery and controller needs)

if you take that info to the ebikes.ca motor simulator, you can customize it to help yuou find out what power level you need to do what you want. then you can find a motor that is mroe capablea than that, and find out if your contorller a nd battery can handle that too.

it'll also tell you about how many wh/mile you will use for various terrains you put in, and you can use that plus the actual distances on those terrains to figure out how many wh you need for the range your equire, which then lets you make sure your battery is big enough for that. then you should generally add at least 25% to that to leave you headroom for battery aging, or potential detours, etc.,




if it works and does the cjob you need and has the features you want, no reason to change it.


there's free phone apps to help with taht; don';t recall the specific names thoufgh

The scope would catch the spike. The motor can also reach its design rpm theoretically. And about cooling i was thinking to use the 4 mounting points on the back of the motor to attach a custom reverse prop that would blow into the motor. Or just use a fan with high static pressure (maybe from a server) and a duct to redirect air into the motor, maybe use a thermistor mounted between two windings inside the motor to only activate the fan in a specific temp range. Today I will measure the rpm with no load and calculate the actual kv. Thanks for the answer!

 

[neptronix]

A much better idea is to get a larger motor that doesn't generate that much heat in the first place because it can handle what you're pushing..

You are in a low efficiency situation here and you'll spend even more power turning a fan to lower temps, so that effectively reduces your efficiency further.

A properly sized motor won't rapidly heat up in the first place, thus, no need for a fan.

Measuring what you're doing would be very helpful.

 

[Bogdan_128]

Update: I measured the RPM of the motor with no load and it was almost bang on with the calculated value, maybe ±200rpm. I also took apart the motor and found no burn marks on the coils whatsoever. Without a powermeter, what else can I check? What am I missing? Could my BMS limit the current? Can there be a weak connection somewhere ? I believe that the esc is fine as it manages to turn on the motor and under very specific circumstances even allows me to ride it (as I explained above). The wires used from the battery to the esc are 10AWG iirc and their length is about half a meter each. The connectors i chose are XT-60. The pwm generated by the arduino is stable. Could the esc work but have an internal issue of some sort that only becomes apparent under load? Also worth mentioning is that the motor and esc haven't warmed up at all after replacing the plastic mounting plate used in the first test with a metal one. With the bike in this kind of limp mode i rode for about 3-4km, only reaching a top speed of 30 km/h with very slow acceleration.

 

[amberwolf]

Without measurements, just general information, all we can do is give general ideas of things that can go wrong, that you'll then have to check for.

Could my BMS limit the current?

Only by completley turning power off. It can't modulate antyhing, just turn it totally on or totally off.

Can there be a weak connection somewhere ?

Any poor connection has higher resistance than it should, so it's temperature goes up from the power loss at that point.

The pwm generated by the arduino is stable.

Is this PWM to drive the controller's analog-voltage throttle input? (non-RC controllers usually have a non-PWM'd throttle input) If so, is that filtered to create a smooth output waveform? If not, and if the ESC doesn't have it's own filtering, it might detect some of the pulses as pulses, and be trying to react by switching motor output on and off rapidly. In this case you can use an RC filter on the PWM to smooth it out; if that's not enough you might have to use a DAC.

Or is the PWM driving a controller's "servo" throttle input? (RC controllers usually have a PWM'd throttle input). It wouldn't need smoothing in this case, it just has to be the right frequency and duty cycle range.

Could the esc work but have an internal issue of some sort that only becomes apparent under load?

Sure--so could the motor, or connections, etc. Almost all of these kinds of things create extra heat at the damaged loaded point, but it may not be enough to feel--can require some form of heat sensing (IR gun, FLIR, etc), depending on conditions.

If the FETs ever got too hot, they might operate, but have higher than normal RDSon, or a change in switching rise or fall time, etc. That might only be apparent at a higher current, or temperature, etc.

 

[neptronix]

Update: I measured the RPM of the motor with no load and it was almost bang on with the calculated value, maybe ±200rpm. I also took apart the motor and found no burn marks on the coils whatsoever.

That's good!

Without a powermeter, what else can I check? What am I missing? Could my BMS limit the current? Can there be a weak connection somewhere?

Yeah, it would be useful to know what the controller is doing.. a power meter is the only way to find out.
Weak connection would show itself as heat on the connector after a ride.
BMSes can't limit current, they just cut the power from the battery when they decide you're pulling too much.