Fiido Q1S Upgrade; 200A, VESC, Copper-Nickel 52v Molicel

[chuyskywalker]

Bought a Fiido Q1S from Voro Motors knowing full well the stock scooter is...anemic...for my tastes. But I love building/upgrading PEVs, and these are great little platforms for doing just that.

View attachment 3

First, I extracted the controller, battery, brakes, and rear motor. (In hindsight, I really should have just bought a frame, lol).

The plan was to use a Trampa VESC 100/250 I got for a great price to make a real monster out of this thing. I wasn't looking for top speed, but really great take off. This led to a few decisions: 1) single motor. It's common to see these built up with two motors, but that's a ton of extra work and, frankly, unnecessary for my purposes. 2) I wanted the biggest, baddest battery I could fit on it.

I checked the frame over and over, did a ton of 3d modeling for fitment and made up a KILLER 14s20p copper-nickel welded pack with Molicel p26a cells...

...only to find that that cells were spaced further than I thought and I hadn't adequately accounted for padding around the battery nor the case it would go in. It would only fit naked, and even then without leaving room for anything else inside the bag compartment -- which I was definitely going to need.

I ended up having to rebuild it as a 14s14p; same cells, same technique, but a significantly better fit that has space for all the wires, the ewheel adapter, and even my fast charger. Only really ended up hurting range, but still sad news and the perks of having a nice space for the charger isn't to be understated.

Other modifications included widening the wheel for a slightly larger motor, drilling holes for VESC mounting and wire pass through, rewiring the cockpit almost entirely (something like 24 individual wires run from the handle bars into the main compartment), and installing a steering damper. Oh, I also modded some EUC pedals to replace the stock pegs -- much chiller stance.

VESC controllers are great for their core competency, but they lack a lot of options many other ebike/escooter controllers come with -- like horn, lights, blinkers, 3way switches, etc, etc. The "Ewheel ADC adapter V2" from spintend does a stellar job addressing all of those issues and the support from them was fantastic.

To fit the VESC into the frame's seat tube I had to modify it quite a bit. I ended up machining an adapter plate to "lift" it from the tube (the rounded corners of the tube made it not fit), and I replaced all of the power and phase wires with much longer wires that came out at a 90 degree angle. I also made several extension wires for the hall, comms, usb, etc so that everything could be wired into it once it was dropped down the tube and the wires fished out of the hold I drilled. Finally, I made a plate that sits under the seat to hold a "metr" which adds ride logging + GPS & LTE tracking to the vesc.

To attach the rear light, I created a 3d printed adapter plate which uses the seat bolts to attach and then has space for the rear and blinker lights.

The handlebars sport dual-hydraulic brakes, 3way speed power selector; horn switch, light switch, left/right blinker switch, two-way half-twist throttle (acceleration and variable regen braking!), and finally a cell phone mount with a built-in 5v usb port.

After that just a ton of wiring and cleanup. Ultimately the specs come out as...

Battery: 14s14p p26a cells; 52v @ 36.4Ah; 1886 watt hours

Battery/Motor Amps Max: 100/200
wh/mi: ~35
Range: ~54mi
Top Speed: ~43mph

Now; how much did it cost? Well, I spent just a bit over $4,000 -- but I also made some purchasing mistakes and had a bunch of batteries left over. I would say to make this again, with the same setup, and especially if I started from a frame instead, this could be much closer to $3,300. But, that's the cost of hobby-figuring things out. I'm not building for profit, but because it's fun for me.

Overall this is a blast to ride and I'm looking forward to putting a bunch of miles on it.

A picture gallery:

View attachment 8
View attachment 14

View attachment 12
View attachment 13
View attachment 10
View attachment 11

View attachment 2

View attachment 1

View attachment 15
View attachment 9
View attachment 7
View attachment 6

View attachment 5
View attachment 4

 

[chuyskywalker]

Ah shoot; this should have been in the builds forum; if a mod sees this please move. Sorry!

Edit: Thanks!

 

[99t4]

I've always felt (possibly mistakenly?) that a steering damper masked a design fault. Does the scooter get speed wobbles?

 

[LewTwo]

What size are those tires (10, 12 inch)?

 

[chuyskywalker]

What size are those tires (10, 12 inch)?

"12" inch, but the actual diameter is 13". Tire sizes are stupid...

The last upgrade I'm looking into is a better (dual crown, air) suspension for the front along with a 14x3 tire. The stock suspension really isn't designed for 15+mph and the spring rebound is pretty jarring at times. The larger/wider tires will aid with cushion as well.

I've always felt (possibly mistakenly?) that a steering damper masked a design fault. Does the scooter get speed wobbles?

Steering dampers are important on steering systems with smaller wheels at higher speeds where small movements are significantly more impactful. They also can help "stiffen up" the steering action so it's less sketchy to take a hand off to itch your nose even at slow speeds.

I wouldn't call it a design fault, per se, more just the reality of smaller wheels and shorter wheel bases.

 

[chuyskywalker]

Well, I pushed the motor a bit hard...

[youtube]eiEgaQnemj4[/youtube]



Oh yeah, that's done for.

Whelp, while I wait for a new motor, I'm working on an extension for the rear swing so I can fit a slightly larger tire on there, but mostly so I can have clamping dropouts.





On both sides I'll use the existing dropout space to put a large M12 bolt in place. Each side also has some M5 holes tapped so I can use those. On the left side is also 2 M6 tapped holes on the frame for the kickstand I can use. I will end up drilling out and replicating those on the right side later.

I used cardboard to get a rough layout. Then I put some paper on it and used a fine point pen to trace all the various features of the original drop outs. Take that outline, scan it in to the computer, bring that into Fusion360 and begin recreating the on-bike dropout. From there I then model up the extension arm and how I want that to work.

Added the clamping dropouts. Also added in a new caliper mount, but "rolled" it forwards so that the motor can be removed out the rear without having to take off the caliper (a flaw in the current design).

This will be 1/2" thick aluminum (on both sides!) when completed, because that's what I can machine on my CNC.

 

[LCLabs]

I'm starting to like these little scooters. What do you think the max wattage on the stock motor is? Is the replacement an upgrade or stock?

 

[chuyskywalker]

I'm starting to like these little scooters. What do you think the max wattage on the stock motor is? Is the replacement an upgrade or stock?

After blowing up the upgraded motor (still waiting on a replacement) I've gone back to the stock motor. I'm currently set to push a max of 500 motor watts. I've no idea what the motor can really handle, but the phasewires get mildly warm so I'm not willing to push much more.

----

While waiting, though, the torque arm has come out great.

 

[CONSIDERABLE SHOUTING]

Did you use straight nickel strip for the battery, or the nickel-plated steel for the sandwich method?

 

[chuyskywalker]

Did you use straight nickel strip for the battery, or the nickel-plated steel for the sandwich method?

Nickel (0.1) on top of copper (0.1). The first battery (which didn't fit, whoops) the copper sheet was 0.2 and required a bit more joules (70) to get it to stick. The 0.1 was around 60 instead, but went faster because I'd upgraded to a dual battery setup and added a second pair of welding cable+tips so I could swap every 10 welds or so.

 

[chuyskywalker]

New (replacement) motor came!

Let's rip it apart




Really wish the company that made this didn't white blob glue all over the PCB. Would have been much cleaner to desolder the existing wire than to have to patch in to the wires themselves.



I did some bench testing with this using the VESC foc_openloop command. From what I can tell, it basically just pushes amps into the motor to heat it up.

Fun note: this command will fail to work if you have your ADC turned on. Which is noted absolutely nowhere I could find.

Before the statorade, I ran foc_openloop 80 3000 to get it to spin and push about 80 amps. However, it quickly heated up to over 100C on the thermistor, which thanks to HAVING that now, the VESC immediately started reducing power to keep the temp at the low threshold (100C). It evened out around 55A.

After adding the statordate, ran the same test and it held steady at 70A instead.

Given this is a 52v system; that means my nominal power went from 2,860 to 3,640 watts. About 780 increase, but I don't really know where that falls on the expectations vs. reality scale for doing the injection. If I bumped up to a 72v battery (which is the more common suggestion for this build) I'd be hitting near just over 5k watts. Don't really need more top-end speed, though, so I think I'll just keep this as is.

 

[chuyskywalker]

Got a new, much better front fork installed. Still working out some kinks with the rotor spacing on that one.


(Also pictured, a temporary 72v battery mounted to experiment with a higher voltage setup.)

And I got my motor extensions back from a proper manufacturing facility, all anodized black and looking hawt.





Been having trouble getting this to climb any hills though, due to overheating. So far one folks I've talked to has blamed the VESC saying they just don't work well compared to Kelly/Sabvoton controllers, while other folks have lamented that you just gotta go for dual motors. I'm still debating each's merits.

 

[chuyskywalker]

I decided, "What the heck, I've already got it there" and bodged up a second controller for the front (stock) motor, allowed it to have 30 phase amps and decided to try and have a rip up the hill I couldn't make it up previously.

...

FLEW UP THE HILL. 20mph easy. (Doesn't sound fast, but this is a 10-15% grade hill.) Mind blown. That little stock geared motor does not mess about; and at 72v it can reach just about 30mph before the rear motor outpaces it and it just spins 0.2A on the freewheel. So, that will be a more permanent setup now.

I also got my hands on a more sturdy sheet metal case and decided to move both controllers into that along with the rest of the electronics.

 

[chuyskywalker]

Got out for a short ride and took a glam shot



Now I'm working on moving up from a 52v to 72v setup. Starting by measuring out, modeling up, and CNC'ing some cell holders for a 20s7p 2170 battery. Should be 29.4ah for 2,176wh total; probably get me just over 50mi of range.






Should get the cells in a week or two and will then begin a copper/nickel sandwich weld up. I specifically went with this pattern so it'll be super easy to weld -- no weird groups or odd shapes to cut out. You could certainly fit more cells here (either by going up and/or to the sides) but the extra hassle to gain another 10-15mi of range just is a headache I don't need. After 50mi of riding, I'm sure a 1 hour break will be more than welcome

 

[PaPaSteve]

Hi
Curious to know what plastic material is being used for the cell holders ?
THX

 

[chuyskywalker]

Curious to know what plastic material is being used for the cell holders ?

HDPE

 

[chuyskywalker]

Drilled a nice hole in the side panel for a charge port. What a trial that was -- went through quite a few bits before hitting up the store for a hole saw to get it done. Tough steel.

Port is 3d printed to hold the xt plug recessed with some CA glue around the connector and a bit of hot glue on the back. Added a cap after these pics.

 

[chuyskywalker]

Mooooar batteries! Upgrading the Fiido to 72v, 29.4Ah (2117wh) 20s7p Molicel P42A pack. Capable of 300+ amp output.

Normally to have that much AMP output you'd need a BEEFCAKE of a BMS, but this will have a teeny, tiny one -- because it will not sit in the discharge path. Instead, the ENNOID XLITE-V2 bms will monitor cell groups and uses a CANBUS connection to transmit cell group voltage levels to the controller(s). This means that the controller will know if any cell group is going to low and run a gradual power rollback (just like it can do with overall pack voltage). All the amp output the cells (wiring/etc) can handle with a tiny bms, yet no lack of protection. The VESC phone app (and metr) can display this data live too!

The cells holders are CNC'd from HDPE sheet and the shape specifically designed to snug into the case. Used 0.1mm copper with 0.15mm nickel. KWelded at about 60-70j. Tabs over the top for the groups, all wired into a DB25 connector (21 leads, 2x2 leads for temp sensors). Lots of kapton, padding, some ABS 1/8" for protection.

Could have gone with 8p or maybe even 9p, either by going up or to the side, but wanted more space to work inside the case for DCDC converters, bms, wires, etc. It was a wise choice as it got quite crowded in there later.

 

[drg007]

beautiful job, must be an expensive batteries pack build

 

[chuyskywalker]

Been a while; made some changes a while back for fun -- specifically, I got a second stock ("250w") motor and hooked them both up. On the 72v, these topped out at 31mph, but my GOODNESS they had some torque. They FLEW up the normal torture test hill at full speed. Nothing else I own had managed that.

Alas, geared motors WHINE, so after having those one for a while, I ran across these bad boys:



That's a direct drive, 2000w, split rim 12" motor.

Split rims are awesome, especially on small tires like this where trying to pry a tire on/off is a major headache. With this you just unbolt the rim and the tire and tube just...slide right out. This is done on the disc side, see here:



The motors perform pretty well too. Obviously not as torquey as the geared stock motors, but the scooter still made it up my torture test hill -- not fast, but did it. I could probably haul a bit more, but I'd definitely need to up the amps. Currently I'm limiting the motors to 40A or 2000w each and it was definitely having a hard time with just the 40A getting up that hill.

These are, according to the seller, the torque version of the motor -- despite that the scooter can do 40+ mph. Inside writing says 16x4 for the wind, magnets/stator are 30mm wide and 40 count. The top speed for the fast version is in the 60's (which is insane). Kind wish they had a 5 turn version as I'd be happy to trade a bit more speed for torque.

Some metr logs from the topspeed/hill climb test:

• Top Speed ~40mph: https://metr.at/r/YcFbe?zoom_start=4568&zoom_end=4816
• Torture Hill: Avg Speed 25km/h, Distance 2.4km, Energy 69Wh

Overall very happy with this little upgrade. Would like to see a version with a temperature sensor though, because I really don't know how hard I can push these without that insight.

 

[CONSIDERABLE SHOUTING]

I forgot to ask- what thickness of plated steel did you use for your battery connections?

 

[chuyskywalker]

I forgot to ask- what thickness of plated steel did you use for your battery connections?

Probably 0.15, I doubt I would have gone thicker since it's just for the resistance to make the weld work with the copper inbetween.

 

[casainho]

Alas, geared motors WHINE, so after having those one for a while, I ran across these bad boys:

View attachment 333302

The motors perform pretty well too. Obviously not as torquey as the geared stock motors, but the scooter still made it up my torture test hill -- not fast, but did it. I could probably haul a bit more, but I'd definitely need to up the amps. Currently I'm limiting the motors to 40A or 2000w each and it was definitely having a hard time with just the 40A getting up that hill.

Hi.

I need probably to go with the dual motor setup and I was looking at that split rim motors - thanks for sharing all this details!!

Did you increase the currents? if so, what are the ones you are using now? and you configure VESC to use the same phase current as the battery current?

I bought a motor with similar shape of your first direct drive motor, and it is: 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).

My question is if I should expect more torque when using that dual split rim motor...

 

[chuyskywalker]

Did you increase the currents? if so, what are the ones you are using now? and you configure VESC to use the same phase current as the battery current?

I don't recall; it's been a while and unfortunately the fiido is in storage for a while so I can't boot it up and check it out for ya.

The split rims I have definitely do not support as much amps as the 60H motors -- those are MUCH beefier, but (the ones I originally used) weren't split rim so pros/cons.