in that case my statement before should be wrong, I will look it up tommorrow
1 Like
Exaclty as @bevilacqua said
What limits us is thermal losses, not how much current the motor can take, that’s why Kv doesn’t matter for torque production, it only matters when you are building a complete system and have limitations there, in out case maximum voltage and current for battery and esc
(this is all ignoring flux saturation, assuming that we won’t saturate the stator)
I think I posted this on the Feather thread, but lets dive a bit.
Lets use my small 5055 motors since they won’t touch the current limits of the esc, if I use them at 190 Kv and run them at 40 A I will produce 96W jus of ohmic losses at the windings
Now, since I wanted them to be direct drive I go ahead an rewind them to 95 Kv, while keeping the same cooper fill, really important detail or this doesn’t work, this is achieved by doubling the number of turns, but at half the cross sectional area
If I still keeps feeding them the 40 A, my new loss is 192 W, since I know that the original motor at 190Kv @ 40 A already hot, I can’t do this, even thought that if my original gear ratio was 2, by keeping the same current in direct drive I will keep close to the same torque, but no equal since we are wasting a lot of power in heat
So we go ahead and reduce the current to 20 A, now we have half the torque as a 2:1 geared setup, while keeping the losses at the same 96 W
What we can take from here is that the is no magic done by rewinding, what dictates the motor performance is Km, which is proportional to motor torque divided to resistive loses, coincidentally you won’t find this information anywhere since it would make it obviously that a motor is bad, and other at the same specs is much better
So if we want a direct drive that keeps the same performance as a belt/geared drive, we need to produce the same torque at the wheel at the same losses, basically if we would plot and efficiency graph of the motors they should look the same
We can go further and estimate how big a motor would be
Torque for a given motor can be described as T = KD²L, K being a constant for a given motor design, independent of windings, D the diameter and L the length
We will write that as k = T/(D²*L)
I will use my board since I’ve done this calculations a few times already
I run 40 A at my 6355 motors with a 14:53 gearing in 125mm wheels, that gives me a motor torque of roughly 2 N.m at the motor and 7.5 N.m at the wheels
The K for this motor design is 9162
Now lets say we will just get this motor and enlarge it to be capable to produce 7.5 N.m at the wheels, so we do T/k = D²*L , lets limits the motor diameter to 85mm to keep a good ground clearance, and doing the calculations we get that our motor should be 128mm long, so a 85128 motor
I think I went too far, hope it’s understandable
But the main point is that gearing allows us to use a smaller motor, and there is not way to get a small motor to perform like a big one. Sure, you could go some crazy cooling, putting dry ice or whatever, running it somewhere that is -70°C, but it will be crazy inefficient, just look at the Raptor 2 and Ghost motors
If you really want to dive deep in this I suggest this book as a starting point, it’s a complex topic and I’m nowhere near knowing it well, lots of times I think I have everything figured out and pops another details that makes my head spin
8 Likes
Thanks for the detailed explanation @Pedrodemio.
I had stupidly mixed up Kv(with units rpm/volts) and Ke(back emf constant with units volts/rpm) and totally got confused. Ke = 1/Kv. Ke is proportional to number of turns and Kv is inversely proportional to number of turns. Now it makes lot more sense. I just need 1.23times more amps to produce the same torque. That’s not too much of an increase in current(not an issue for ESC), but like pedrodemio said, for same current the ohmic losses would be 2.5 times(190/75, assuming same amount of copper). For the same torque the ohmic losses would be 3.8 times(I^2R loss). So effectively I would see a decrease in wh/km. Currently I am getting 18wh/km, let’s see how much I get with DD.
Case1: 16:50 geared setup with 190Kv(rpm/volts)
Case2: DD with 75Kv(rpm/volts)
Kv2/Kv1 = n1/n2 =75/190 = 0.394 where n is number of turns in the stator
torque on wheel1 = 50/16 T1 where T1 is the torque produced by the motor1
torque on wheel2 = T2 where T2 is the torque produced by the motor2
T1 = K*n1*A1 & T2 = K*n2*A2 where K is some constant and A is current
T1/T2 = (n1/n2)*(A1/A2)=0.394*A1/A2
torque on wheel1 = torque on wheel2 This is what we want
(50/16) * T1 = T2
(50/16) *0.394*A1 = A2 =1.23A1
3 Likes
these motors dont have a “square hole” like some of the other chinese hubs. How does the motor not rotate on the shaft itself? the nut??
1 Like
I appreciate the simplicity in this design. Bring on the Summer 
3 Likes
Maybe hit up @PXSS for adapter designs. He made me a nice set for my SDXL’s and ABEC’s in a group buy.
I’m guessing only the bolt hole pattern would be different on elofty’s
1 Like
The other day the Unity arrived, so I did a test build.
The Elofty’s DKP track is pretty loose. I tightened the kingpin nut, but it doesn’t work very well… @DAddYE, @BillGordon What do you think?
I feel like die at speed wobble. lol
5 Likes
Can you convert it into a single kingpin? It has to be used with a drop down deck or used like a TKP, otherwise it’s just too far from the ground.
1 Like
Yeah, that looks very tippy. I would pull the arm and use it like mishrasubhransu mentioned.
1 Like
Oh, I’ve never thought about that! Thank you.
I’ve always found that DKP + thane works best on drop-through, flush-mount decks like the factory Evolves:
The design calls for the rider to stand between the supercarve base plates.
I think the Evolve Bamboo ONE failed, even on urethane, because it does not feel like the other Evolves, or any drop-through.
You appear to have risers, on top of DKP base plates, and then adding the deck thickness - definitely too tall.
If you want to try a short board, go for the Globe Spearpoint Mini:
For best board feel on DKP, however, one must stand between the baseplates and allow the deck to flex as you pump through carving.
Hope this helps my brother!
11 Likes
Nah! You’ll get used to it! Maybe check my post on the other thread about bushing setup
1 Like
Spearpoint Minis are awesome! I may prototype one with the eLofty
1 Like
Thank you, my brother. Hard to find it’s a Koastal Skillet, and they’ve been out of production for some time. Koastal does have other boards however none of them are quite as dope.
Trampa’s Orrsum works with, literally, everything and I think would make a sick eLofty platform, drop-through DKP.
If you prefer wood, I like Subsonic’s Ruse for DKP:
https://www.subsonicskateboards.com/products/ruse-39-deck-or-complete
It has footwells not unlike the vaunted Colab mountain board.
Good luck!
4 Likes
What is the middle deck? :o
1 Like
4 Likes
I have an Inboard M1, and I know little to nothing about the electronics involved. I’m assuming I have blind spots, so I am looking to get some advice from the ESK8 community. I’d like to attach the Elofty clones to my ‘out of warranty’ M1.
Do you think I would have to do any modifications to the ESC wire connection? Any other reasons I’m not thinking of for why trying this would be a bad idea?
Thanks
You may need to replace the esc in the board. It is not gauranteeded that the esc will run the dd correctly
2 Likes