Fiido Q1S 5kW dual motor: VESC + modular DIY OpenSource electronics and software

[WarrenC]

I burned this new again on his first 5 minutes!!

How did u burn them? overheated? current surge?

 

[casainho]

How did u burn them? overheated? current surge?

Sine I remember, yes. I did connect to the battery power without disabling the BMS switch.

 

[casainho]

Hi, great project! How do you like the comfort on light offroad, in the forest or on gravel paths with these new tyres? Did you upgrade the suspension?

So yesterday I had to leave the car at the mechanic workshop for a week - the distance from my home is 25kms. So I decided to put the Fiido Q1S on the car and I returned home riding it. And I did easy secondary roads as also some gravel roads - this are places where I grow up. Was a nice sunny day, after various days of raining. I enjoyed a lot this ride!!

On the gravel roads, I felt comfortable to ride between 15 to 20kms/h.

I connected the motor temperature sensor to VESC and I found the motor was getting much more hot than I expected... I need to better understand all this, and I am afraid my VESC configurations are not so good at low speed / high currents, that may rise this issue of the motor getting to much hot.

I also implemented the Cruise Control feature and it works as expected. But I also did understand the throttle signal and a good amount of noise - I wish I can improve it.

I also found that I need a buzzer on the display, so like when Cruise Control starts, I can get a quick confirmation. The buzzer will also help on the turn signals.

 

[casainho]

I have a new issue to solve... I am trying to do longer rides, and sometimes even on more hilly terrain, but the motor heats a lot and stops to work.

The next pictures shows the Fiido Q1S on a train, on a nice early morning sunny Sunday, where I went to a distant place by train and then I did plan to go to a hilly place where I would run on a Trail Run event. But I didn't get there because the motor got very hot and lost power with that... - anyway, was a good experience transporting Fiido Q1S on the train, since it is way smaller than a regular bicycle. Then after, riding between the traditional villages is amazing because they are beautiful!!






So, this is the motor inside, when was new:



After that heating, I took this new picture. As you can see, the enamel coating of the coil wires, got liquid and went to that white part, also there are parts that are now brown instead of white. The bearing now has rust, because of the heating.



The motor still seems ok to me, but is more fragile now to more heating, as the copper wires lost some enamel so, more high heating's and they motor coils will short!!

So, because of this, I connected the temperature sensor to VESC - I didn't this before because I used my hand the feel the motor temperature and I always thought it was ok, but I was wrong...

Why I am having this issue now? well, because the previous motor had a different coils winding, it did run at high speed but had low torque. But this new motor version, the only difference is the rotor with a different coils winding, have 30% more torque and 30% less speed. BUT also, the coils resistance is now 70 ohms while previously were 30 ohms, so an increase of 233%!! and this increase was expected. So with a higher resistance comes and higher power losses in the form of heating....

What should I do next??

For now, I think I will reduce the motor torque / current but only at higher speeds, like I want to keep the torque but mainly for the startup and as soon a hit some speed, maybe like 25 kms/h, I may soft transition to a lower torque of -40% at my defined max speed of 55kms/h.

I may also later try to use a solution as Statorade, to help reduce the motor temperature, with max potential to reduce up to 40%.

 

[SlowCo]

Great little bike! I suspect the StatorAde will help a lot. Next step will be a bigger diameter motor.

 

[casainho]

Great little bike! I suspect the StatorAde will help a lot. Next step will be a bigger diameter motor.

About the motor, if I can find one that is more powerful, with more torque and more power, so I can use on hills... or maybe go with dual motors...

 

[SlowCo]

I don't know what the drop out width of the bike is and what size tire will fit but maybe one of the QS hub motor with rims is a match?

QS moped motor

 

[amberwolf]

After that heating, I took this new picture. As you can see, the enamel coating of the coil wires, got liquid and went to that white part, also there are parts that are now brown instead of white. The bearing now has rust, because of the heating.

The stuff on the white silicone is almost certainly not the enamel, but just rust,etc.

If it was the enamel, the windings would be shorted together without it, and the motor would not operate and your controller would smell funny.

Heating doesn't cause the rust, but water intrusion does, and also would let it be deposited onto the silicone as the motor rotates.

Then heating would dry out the water, leaving the rust deposits attached to whatever they dried on.

Sufficient heating can certain "cook" the enamel, leaving a burned-circuit-board smell in the motor and darkening the enamel color. I don't see a difference in color of the windings between the two images, but the lighting isn't identical nor is the angle, so there could be a difference simply not shown there.

So with a higher resistance comes and higher power losses in the form of heating....

Only if the current thru them is still the same. Do you still get identical phase currents? If not, and the currents are now proportionally lower, then heating should be about the same as well.

How hot does the motor get? It has to be very hot to cook the winding insulation off (burning it); if you have hall sensors then they will probably stop working correctly before those temperatures are reached. If they don't get too hot they may recover when cooled down, but while that hot they are unlikely to operate at all, or at least not as designed.

Long ago I overheated a Fusin "350w" geared hubmotor repeatedly to the point the halls would stop working, have to stop and pour water on the motor casing until it cooled enough to ride again, and the windings didn't cook. At some point during one or more of these events the solder holding SMT caps/resistors on the hall board actually melted and tiny parts came loose...but the windings were still good, though eventually I broke the clutch on the motor at startup from a stop. (didn't melt the gears, either).

 

[casainho]

I don't know what the drop out width of the bike is and what size tire will fit but maybe one of the QS hub motor with rims is a match?

QS moped motor

It is a 12 inches motor. The motor dropouts internal distance is 125mm. The tires can be like 12x3.00 or 12x3.50.

I need to look at that QS motors. For now I am looking at this ones - because they have 40 magnets and mine have 30, so I expect 33% more torque(??). And it is a rim split, meaning it will be way easier to install the strong good quality tires!!

Name: 12 inch Split Hub Motor
Voltage: 48v 60v 72v
Power: 1500w 2000w 3500w 5000w
Speed: 48v1500w: 55-60km/h; 48v60v2000w: 60-80km/h 60v72v3500w: 60-90km/h; 60v72v5000w: 90-120km/h
Max torque: 60N.m
Magnet poles: 20
Magnet steel height: 30H
Open size: 121mm
Axle length: 201mm
Tyre model: 62-203 (12½x2)

 

[casainho]

Only if the current thru them is still the same. Do you still get identical phase currents? If not, and the currents are now proportionally lower, then heating should be about the same as well.

How hot does the motor get? It has to be very hot to cook the winding insulation off (burning it); if you have hall sensors then they will probably stop working correctly before those temperatures are reached. If they don't get too hot they may recover when cooled down, but while that hot they are unlikely to operate at all, or at least not as designed.

Yes, I keep the same current, because I needed more torque comparing to previous motor. I was not aware of the resistance increase.

A search on google:
Does temperature cause rust?


An increase in temperature contributes to an increase in the number of active centers of corrosion on the metal surface and accelerates the development of corrosion processes; therefore, it is necessary to provide the protection of metal products, equipment and structures operating at elevated temperatures.

 

[amberwolf]

Heat can certainly *accelerate* rust...but there has to be something in there to help that happen. Just the oxygen in the small amount of air trapped inside the motor (or even an actively-ventilated one) is not usually sufficient, especially over a short time scale like you've had. High humidity, or actual pooled water, is much more effective, especially if you have salted roads in winter (even if they aren't salted when you ride, it's still in the road surface and still contaminates the rainwater on it).

In hubmotors, the kind of rust you see in your pics there is always (in my experience so far, and that of other posters here) caused by water intrusion, usually thru the bearing seals or motor wiring / wiring channel in the axle, etc.

 

[casainho]

Heat can certainly *accelerate* rust...but there has to be something in there to help that happen. Just the oxygen in the small amount of air trapped inside the motor (or even an actively-ventilated one) is not usually sufficient, especially over a short time scale like you've had. High humidity, or actual pooled water, is much more effective, especially if you have salted roads in winter (even if they aren't salted when you ride, it's still in the road surface and still contaminates the rainwater on it).

In hubmotors, the kind of rust you see in your pics there is always (in my experience so far, and that of other posters here) caused by water intrusion, usually thru the bearing seals or motor wiring / wiring channel in the axle, etc.

I didn't use this scooter / motor in rain... but, I live near the ocean and there is a lot of humidity!!!

 

[amberwolf]

I didn't use this scooter / motor in rain... but, I live near the ocean and there is a lot of humidity!!!

And salt in the humid air, most likely, which will also accelerate corrosion.

So....I'd still call ths one a simple rusting problem, and not a heating issue. (not that it isn't getting hot,, just that ot doesn't seem to be damaged from it at this point)

You may have higher heat, but I don't see damage from heat in the images so far, just the corrosion. (which being metallic, could cause other operational problems if it gets in the wrong places).

Bearing corrosion will make the wheel harder to turn, so take more power to spin, so create more heat...but a serious bearing problem to take serious power to overcome would be easy to tell is happening when trying to spin the motor by hand....


I'd replace the bearings, and pack them with grease, and pack grease between them and the seal rings on the outside. (if there aren't any seal rings outside the bearings, I'd add some).

You can also seal up the motor internally against corrosion with various coatings, once it's completely clean of any salt, humidity, or corrosion. There's a number of threads about that sort of thing, such as Kingfish's old thread about DIY Rust and Restoration (not sure of the exact title).


If you need to replace the motor with one more able to provide and handle the power you need for the riding conditions you have, I'd recommend doing those sealing/greasing mods before using it, to keep the humidity/salt out of it to start with.


2WD will help by spreading the heat generated across two motors, but it's a little less efficient as you have the extra weight of the other motor and controller. But it's also redundant so if you ahve a failure you can still at least limp home on the other one.

That's how I've setup both my successful cargo bikes (well, one bike and one trike), CrazyBike2 and SB Cruiser.

 

[casainho]

I found and confirmed with the seller, that my motor is in fact the 5000W version and not the 2000W version I bought!! But since I asked for a custom winding to have more torque and lower speed, she is saying if I buy another motor but this time for 72V, it will have more torque. However, I found online that the dual motor phase wires version, like mine, is for 60V and 72V motor versions!!

So, probably I have the 72V 5000W motor version. The seller also told me that this motor version should have about 80N.m.

So, to resume, my current motor should be 72V 5000W, has 30 magnets of 60H, no load speed of 1230RPM / 70km/h at 72V battery, and the phase wires are 4mm2 (2mm2 each wires, 2 wires for each phase).

And I was looking to buy this different motor, that everyone says has a strong torque - but I wounder how can I know before I buy it.
The data I have about it:

72V 5000W, 40 magnets of 30H, no load speed of 1600RPM at 72V battery and 6mm2 phase wires.

Anyone knows what should I expect, compared to my actual motor??



What I like on this motor is the is the split rim, since it is very easy to install and remove any thick tire like this one:

 

[casainho]

I am limiting the motor phase current to 135 amps and I see the motor getting hot relatively quick. The seller told me this is a 2000W motor and the max motor phase current should be 80A.

Today I tested the motor for the first time using the VESC foc_openloop 80 1000, that rotates the motor at 1000 ERPM (wheel speed = 4kms/h) and forces a motor current of 80 amps. But if the motor hits near the max temperature, the motor current will be reduced automatically to keep the temperature at a safe value.

So after a few minutes, with 13ºC ambient temperature, the motor stabilized at 90ºC with a motor current of only 27 amps. The battery voltage was 74V and battery current was only 1.46 amps, with a power usage of only 110 Watts!! -- this data is registered on the next graphs recorded from VESC.

So, if riding my motor, it will get hot up to a point that it can only run at 110 watts?? is there anything wrong with my motor or are this regular values??
@amberwolf

 

[amberwolf]

I am limiting the motor phase current to 135 amps and I see the motor getting hot relatively quick. The seller told me this is a 2000W motor and the max motor phase current should be 80A.

Well, if you're running 135A thru something that only takes 80A, I'd expect it to get hot pretty quick. That's almost 1.7x it's design intent. (assuming that this is actually a real manufacturer specification, and is actually for phase amps, vs average battery current (which is a more common spec to be given)).

What's the actual continuous phase amps during an actual ride when it's getting hot? (if you said, I missed it)

If you're only peaking those levels every so often it wouldn't be such a big deal, but doing it a lot or for more than a few seconds at a time is probably going to build up heat pretty quick.






I don't know how the VESC does that test / mode, so I can't say why it is doing what it's doing there.

But...any test that doesn't have a physical load on the motor to create the current demand by it seems that it would not be a real-world type of test, and it's results might not be applicable to actual usage. I don't know enough about how such a test might actually be done without a physical load , by only the controller, to even guess.


The main things I've seen heat up motors that shouldn't for the given power level are running them with poorly-tuned controllers vs the motor design, and running them under loads that the kV they have doesn't support for the speed they're being driven at.

The former is certainly fixable with a VESC, although I have no experience at all with tuning one.

The latter just means that if the motor's loaded RPM at the given battery voltage and controller-output-level is too low, compared to the RPM it would have unloaded in the same inputs, there won't be enough BEMF to counter the current into the motor, and it will heat up more than it should. (you probably already know all this, but if not, play with the ebikes.ca simulator with any of the thermally-modelled motors in various setups to force this to happen, and you can see it in the charted results).

This specific example isn't directly applicable to an FOC contorller as they modulate torque, not speed, but it should get the idea across (I can't presently think how to state an example for an FOC, but there is a way): For instance, if the motor's unloaded speed at say, 50% throttle, is 500RPM, but loaded down it drops to 100RPM at 50% throttle, the current will be say, 5x higher than it "should" be. (realistically loaded is usually around 80% of unloaded RPM, so it *should* be around 400RPM, so current is really only 4x higher than it should be). But that means it heats up proportionally faster than it would if it were able to run at the speed it should.

I'm not sure if that makes any sense...I'm too tired to continue the thought, though.

 

[casainho]

As my motor is heating a lot (at least much more than what I expected, but I may be wrong), I improved the firmware, to keep the highest motor current (135 amps - 0 kms/h speed) at start and then reduce it linearly up to 40 amps and keep that value afterwards (40 amps - after 30 kms/h).
I also added an option for speed control were regen current is 0 amps, so I could coast - as regen also heats the motor!
Then I also updated the the current control mode. In the end, I ended up using the current control mode as I could modulate the current with the throttle and so avoid big spikes in current, to reduce motor heating.
I did on Sunday a ride of 50 kms, mostly flat. I found that motor temperature did stabilize like at 75 degrees, with a 35 or 40 amps motor current. That was is a low torque for me, did work only because was flat roads.

Next step is to try figure out if my motor magnets are a bit bad already - I have the other equal motor but with different winding, let's see if I can compare them.

I need to add a few custom UART commands to VESC firmware, but latter I need to implement them with VESC own scripting language to avoid to have a custom VESC firmware.

The main update on the Python firmware, running on the main board ESP32-S3, to control the scooter:

 

[casainho]

I decided to install a second throttle specifically to control the brake / regen. I installed it on the left side, but I installed a throttle that was made to install on the right side, so on the left side I need to move it forward while the throttle on the right I need to move it backward. While I don't have it yet wired and developed the firmware logic to use it, for what I tested physically, seems it will be intuitive to use to control the brake / regen:



As my motor is overheating, I decided to test the magnets to see if they are ok. And since I have a similar motor, with 60H magnets, I tested ones against the others -- I use a metal bar to get attracted to the magnets and they I tried to lift the bar using the digital scale to measure the force I was doing - I did measure the same force on both motors, so I assume the both magnets are good.

Interesting to see that they installed different size magnets combinations on each motor.
Motor 1:


Motor 2:

 

[casainho]

As my motor were overheating, I couldn't find other way to improve the issue than use a dual motor -- as I had a used second motor in stock (with 30% less torque and 30% higher KV), I installed on the front wheel. I also had in stock a 135mm distance between dropouts fork for Fiido Q1S, exactly to install a front motor.

I am already riding it with dual motor but it is not the final setup and some parts of the new fork were incompatible with original fork. Current state:



I want to go back to the 3.50-8 tire at front wheel, to get improved more stable handling and more comfort, since now there is an extra 5 kgs on the front wheel from the motor!!

This new fork is not compatible with the original stem, so I had to improvise but I will buy a new stem mainly because I want to keep the folding so I can transport Fiido Q1S on my cars.

On this first firmware version, I used 60% of current in rear motor and 40% on front motor to try keep the same used power. It works well, there is only one little issue is that the front wheel always skids at startup, and I don't like it because it wears out the tire.
I guess I will make the firmware to startup only with the rear wheel at max torque and after like 5km/h, I will switch half the power to the front wheel.

Some photos of the dual motor installation process:



Was the first time I connected 2 VESCs together, they are connected by CAN. On the CAN bus I also have the ESP32-C3 running the VESC Express, that let me configure both VESCs wireless.

The additions to the firmware to control the second VESC, was straightforward, as there is a VESC command to forward the data package to a specific VESC on the CAN bus - so it was a matter to send the same command for each VESC, but for the second VESC, that command is encapsulated on the CAN forward package command.



And here the original fork with the 3.50-8 tire and the new fork with the motor, with a cheap not very good 3.00-8 tire:

 

[casainho]

I am using my Fiido Q1S every day, mostly to go make small shopping or to the gym. Sometimes I will need something from the city and is way better with Fiido Q1S then with car, because there is no parking at the city.

I did a few improvements, I hope to post in next days pictures and details.

So I am using a liquid crystal display, with the same 128x64 pixels as the previous OLED display. This liquid crystal display is way better for reading outdoors!!

Currently it is showing this information:
- wheel speed
- time
- battery voltage
- total motor power
- motor temperature



I added an small RTC module based on RTC DS3231. At boot, the ESP32-C3 tries to connect to my home wifi and get automatically the time by NTP. If successfully, updates the DS3231 clock time, if not, no problem as the DS3231 keeps the clock at very accurate value when the scooter is powered off.

I value the time because I am almost always short in time when I go to gym, and then I can adjust my speed based on how short in time I am. If possible, I prefer to ride more slow and relaxed.

This is the next design I am thinking to implement:



The focus are only 4 data types:
- wheel speed with big numbers at right
- on top, the time
- on the left top, the battery SOC
- at bottom the motor power

By the way, I have configured to have total 2500W at startup and then that power reduces gradually to 2000W at 30 km/h.
I have the front motor with low power at startup (to avoid skidding) and then gradually increasing up to the 30 km/h. The rear motor is contrary, high power at startup!!
Max speed is configured to 55 km/h.
With this setup, it is a joy, a lot of torque and fast reaction!!!

 

[casainho]

In August 2024, my Fiido Q1S frame broke!! Luckily I was riding at a slow speed...

I was so sad, because I had a plan to do a long ride on mountains while I went to camping and do a Trail Run race, on the next 2 weeks. And all the efforts I put on the seated scooter, I would not trust anymore on it, even if buying a new one!!

Here a picture on how it broke.


Then I asked on another online forums and found other users had the same issue!! Amazing the amount of sellers selling Fiidos Q1S with high power motors and battery installed but without reinforcing the frame!!!



Luckily, after talking with some friends, I got all the help to send to a local shop for soldering and reinforcing the frame!! I was amazed with such possibility.

Here the reinforcements parts before soldering:


And the final result:


So, all solved in 2 weeks and then I could participate on the nocturnal Trail Run race - here a video of riding Fiido Q1S:

I carried a secondary battery and a charger, and other gear, clothes, etc:


I planned the ride and I used my Garmin GPS I use for my MTB advantures:


I had to do a bit of offroad at the top of the mountain:



Indoor camping was nice, relaxing and interacting with the other runners:


And the night was amazing!!

 

[casainho]

Some months ago, I decided to create lateral panels in acrylic. I used an angle grinder to cut, a heat gun to shape it and finally painted in black.

My little daughter did help me a bit









Sometimes we go to the park (riding very slowly!) and we enjoy a lot riding on the dirt roads

 

[casainho]

The current design of the main screen. Big numbers for the current wheel speed!! A graph bar to represent the current motor power, between 0 and 2000 watts. With smaller numbers, the current time. And I need to work the battery SOC value.

 

[phobos]

This is super cool! Do you think you need more info there like battery or phase amps?

 

[casainho]

This is super cool! Do you think you need more info there like battery or phase amps?

I am thinking in adding a secondary main screen, where I can add technical data like phase amps. Daily, I just need to know if battery is fully charged when I leave home. The current time is also very important!!
Wheel speed to understand if I am riding to much fast and motor power to understand if I am on the limit and try to backup if possible, as to much power will discharge the battery fast as also will heat the motors quick.

For the battery on the main screen, maybe I will add the battery symbol with 5 bars / 20% each bar.

About battery, I wish also to calculate and show actual battery resistance (on secondary main screen), as is something that shows battery degradation over time.