Thank you! I really want to compare the motor construction to the Onewheel+XR motor.
If you have a source for the XR motors I would happily buy one for comparison (or as a backup if the peiscooter is underwhelming)
1 Like
I’m on here too, and yes i did write that 
Your build is one of the illest out there 
4 Likes
Thanks man Although it is the balance app you wrote that’ll be more useful to most people!
1 Like
Good luck on the build!
A rule of thumb in PID tuning is you first zero the P, I and D factors, and start with a very small P. Keep increasing that until it gets rideable, it should be perfectly rideable with only P. A too small P value is easy to see in that there’s no power, you can slam down a footpad easily without it correcting enough. Keep increasing it until you start seeing effects of too high P, meaning a twitchy board that’ll easily go into a death wobble (fast side to side oscillation). If it’s way too high, it’ll just shake and twitch violently.
When tuning I first rigged the Redstar to engage its motor when I push the button without regard to footpads, so I could get the feel for it without standing on it. When I got it to the point where it strongly resisted pushing the footpads down but didn’t oscillate, I hopped on it and it was kinda-sorta ridable already.
After you find the highest P value that doesn’t cause issues, conventionally you’d drop that maybe 10% and start increasing the I factor. That’s what will correct for long term difference from the setpoint, f.ex when you are climbing a hill. With P only you’ll always need to keep a footpad down, with I it’ll stablize so that you can ride up a hill with the board level. I use a strong I term, but if I remember right some balance app users had problems with it. Best is to experiment, but keep in mind I term is the one that’ll make your motor kick to full throttle if the board isn’t reaching the setpoint for whatever reason. Like if you are holding it in your hands 
D term usually has the effect of damping oscillations, it’ll reduce throttle when you are rapidly closing on the setpoint angle before passing it. It doesn’t do much good for me, but apparently works nice in the balance app. This should be tuned last, and too much of it will make the board sensitive to vibrations, as it reacts on fast changes in the gyro signal.
It’s a good idea to read up on PID controllers, what those terms actually do, and then test test test test 
7 Likes
Thanks alot, I know the basics from PID in general (some readup I did for thermistors/thermalcoupls for 3dPrinting on smoothieware and troubleshooting a twitchy thermalcontroler for work since I’m kinda our fix it guy) but the little thinks to minimze death wobble are indeed much appreciated 
1 Like
In that case it’ll be a breeze for you, there was a pretty big range of parameters where it works OK, and after that it’s just fine tuning
Heres a dump of the frame CAD model & drawings just in case.
https://drive.google.com/open?id=1C5Ja17-slUMEVS0x2dM7uQy0UqvWTTQV
2 Likes
Designed a fender for Redstar, got tired of getting water, grime and grimy water thrown all over the front deck and my ankles. Luckily a friend had a printer large enough to print it in one go. Works wonders!
So far this thing has been reliable, and the waterproofing has held through some rainy days. Now that there’s a “800W” version of the Phub-188 out, I’m interested in getting that with twice the Kv of my current one to fix the lack of speed. If only I could get actual specs out of peipeiscooter, last time I tried talking to them about it they answered with a question mark and left it at that. 
I would not be surprised that the new Phub is the exact same as all of the old “options” given the history with those, but you never know…
8 Likes
Not yet, as I’ve been riding it 
But I’m thinking about it…
1 Like
I know. @Narnash said above he was gonna buy one a d check it out… ive been looking forward to that.
1 Like
In which case there’s good news for you, just cracked the Phub-188 open!
I figured the two end caps likely attach to the rotor with those screws in them, and hold the rotor in place with bearings. After removing the screws nothing would budge, but a solution presented itself in the form of a hammer.
By furiously hammering at the non-wire side of the axle I got the stator and the other endcap to slide out somewhat, I could then grab the endcap and pull the stator out with it. Really snug fit between the bearing and the axle.
Tight magnet packing, no retaining rings or anything extra to keep them in place. Also no O-ring between the endcap and rotor, so this motor is not waterproof.
There is an O-ring on the axle which is nice. The hall board is held in place only by the sensor leads and some blobs of putty, but seems to be pretty solid and the sensors themselves are embedded in the laminations.
There’s some knotted string holding the magnet wire in place, and some wooden dowel-things used to wedge it tight in the slots. The magnet wire is wrapped in some sort of insulation in the slots, which I guess would reduce the chance of it rubbing and breaking on the edges.
There’s a fair bit of empty space between the stator and the endcaps on both sides, moreso here where the wires come out. Assuming Peipei’s power specifications actually mean anything, it could be that they simply make the stator with different lengths. There’s certainly room to grow for the “800W” version which is now on Aliexpress.
A zip-tie holds the phase wires in their place.
Someone with more knowledge of motors could say whether this winding is good or bad. It does seem to fill the slots pretty nicely, but there’s the odd strand which is more loose than the others. Handmade?
This here looks like it might be the wye center termination. It’s got three sets of wires going in there. Can someone with more experience with this sort of thing confirm? Appears to have 6 parallel wires in each set of 3, so 6 parallel wires per “turn”?
I guess this is where I’d need to do my work if I’m to convert it to delta, breaking out that center connection and soldering each set of three to the (hopefully) right phase wire.
11 Likes
Good breakdown of the motor, I’ll lookout to see what happens with your delta switch. Your goal is to increase power by switching to delta, right?
2 Likes
This is such an awesome looking Onewheel. A true inspiration!
Your motor will probably have star winding and speed can be increased to over 30km per hour just by soldering it to Delta.
@Remieknaapen already did his Phub motor. I think there are pictures in this or the old forum.
5 Likes
Speed, primarily. The higher Kv would allow me to push more power as well, but I’m limited by the 10s3p INR18650-M29 pack I have. It’s likely I’ll need to make a more powerful pack, looking at some VTC5A’s for that. To keep the same torgue I have now I’ll theoretically need 70% more current, we’ll see if the motoro overheats.
@Bionic24 Thanks! Sadly I think it’s going to be just 24 km/h or less, a 70% increase of what I get now. Ideally, I’d go to a 12S pack if I need to rebuild for that bit of extra, but that would drop the cell count by quite a lot since 12S3P would not fit. Unless I take out the BMS + connectors in the pack 
I need to check @Remieknaapen’s pictures, wasn’t aware of them!
1 Like
Thanks @iKNOWaFATman … real life kinda hit me hard the last months, I actually didn’t touched any e-skate stuff in the last 4 months as I got distracted the whole time 
I just cracked my Phub-141 open (gone the same way as Jopj by just hammer it. The Phub-141 looks as expected very similar, but mine is dirtier and slightly damaged (broken magnet, separated laminations, the stator also seamed to rub a bit on the magnets).
The dimentions of my Phub-141 (@Jopj please validate if the stator is the same on the 188)
My motor has 30 magnets (->15 poles) my calculated ERPM (from my last experiments) is 175 per volt, devided by the 15 poles that gives me a kV of 11.7
It has 2x 6202RS 35x15x11 bearings
The hallsensor board is just glued on top and not coated, there is no o-ring on this motor either but the lits where both also glued with some kind of silicone glue. So lets say somewhat water tight but not water proof, the sensor could die prematurely pretty fast.
My terminations, likely wye I didn’t looked to closely yet but changing the termination looks very doable .
But as you said the stator/motor to hub size ratio is really low, there could at least fit a way longer stator in this thing.
2 Likes
Diameter looks the same…
But there’s a bit more length, mine is exactly 50mm.
I also have 30 magnet sets, but they are made out of two smaller ones, not one large. The hall sensor board is different too. Not sure if these are related to the motors being “different” or just individual variation. Looks like the stator is created by stamping x amount of laminations on a shaft. I must assume this is how they do the different power versions. Atleast that’s how I’d do it in their place, just tell the guy how many to put on this particular motor, rest of assembly remains the same.
5 Likes
It’s now delta-terminated, and nothing seems to be shorted out. Got to test if it works, then wrap everything together more solidly. Which is complicated by the shaft being made of soft metal which expanded while hammering, and no longer fits in the mounts 
7 Likes
ooh looking forward to see how it performs. key factor in deciding between doing a hoverboard diy or using the phub rn.
2 Likes