Leaf / leafmotor / leafbike high efficiency 1500w motor

[cycborg]

The MXUS I've seen referred-to the most is the "mini cro" ("V3"?), which you must not be referring to. MXUS's website only has one DD hub motor listed, it's an 83% 1500W, so that must not be it either. So, curious which you meant.

teslanv got a sample of a thin-lam MXUS XF40 a few months ago. It’s a 30 mm wide stator with 135 mm dropout spacing, so rough equivalent of the 9c or Leaf 1000 W. The sample performed well but unfortunately MXUS only did a limited run with the thin lams, so I don’t know that this version is available currently anywhere. teslanv can special order the less-efficient, thicker-lam version.

 

[spinningmagnets]

One of the specialties around here is to buy a new product and see what the results are in the real world by raising the power you apply to it, and see what happens. I was disappointed to read that MXUS would not stock the thin-lamination motor, and any retailer who wanted them would have to pre-pay for a special order of 100 units (although they might be talked into doing a smaller order, YMMV).

This makes me wonder, how much of a cost difference is there between the cost of 0.50mm lams and the desirable 0.35mm lams? (in a 30mm wide stator, roughly 160mm diameter). Most of the hubs around here do not have a $2 temp sensor from the factory, which is just nuts. With the exchange rate, can thinner lams really be more than $5 per hub? (edit: an industry insider informed me the cost to swap-in thinner 0.35mm lams is just under $10 per hub, thinner lams provide about 3% better efficiency and about 30% less cogging)

Swapping-in a mass of aluminum in the stator support is a way that many motors have recently been able to allow temporary amp-heat spikes, as long as there is a cooling-off phase during cruise mode. The thin stamped-steel centered base models (with same stator width) are lighter and less expensive.

So...if I am comparing DD hubs that are thin enough to have a 7-speed freewheel (and my examples are roughly the same price) do I go for a 30mm wide stator with an aluminum "heat sponge"? or...go for a 35mm wide stator and a stamped steel center support, but it has thinner lams?

Do I accept more hub weight as an aluminum center, or more weight as a wider stator (both weighing the same). One absorbs heat spikes and dissipates them over time, the other makes less heat in the first place (while having roughly 15% more torque so it reaches cruise faster due to 35mm stator width over 30mm).

I'm now thinking that...if you like ventilation or oil-cooling, buy the Leaf with the thinner lams and 35mm wide stator. If you like the motor sealed up with no oil, get the 30mm wide stator with the aluminum center support, even if it has the thicker 0.50mm lams to keep the price down. The MXUS 30mm wide has an aluminum center, but the thin-lam version is currently "unobtanium" so its kind of a moot point, but...worth thinking about?

 

[Kodin]

3mm phase wire freewheel version shaft for 1500W motor: Just shy of 14mm thread major diameter.
3mm phase wire cassette version shaft for 1500W motor: Just shy of 12mm thread major diameter on CASSETTE side, 14mm on BRAKE ROTOR side. This means you'd need a torque plate of each size for this configuration.

Regarding maximum cassette size: you'd be looking at 145mm minimum dropout width, more like 150mm if you use a 9-speed cassette. With a 9-speed, you'll also need some extra washers on the cassette side. KEEP THIS IN MIND.

When cassette unit is mounted to shaft, this is where the inner waser would have to sit to give you clearance for my 9-speed cassette:

Keep in mind the nut actually has to hit the bottom of the thread to lock the freehub cassette into place, so you'll need some washers to make up the difference.

 

[mchlpeel]

So am I right in thinking you can have either a freewheel version or a cassette version?

Too many decisions. Which to pick?

 

[neptronix]

12mm axle on one side is very lame and may be a potential failure point above... i dunno.. 3000w peaks ( just a guess ) :/

 

[SprocketLocket]

Thanks for that insight Neptronix. I guess that's why LeafBike tries to steer you towards the 1000W motor if you ask about having a spline cassette on the 1500W (even though they will do it).

I enjoyed the idea of using my existing cassette from my bike, but maybe getting an old-style freewheel gearset won't be so bad. And I guess you can get a freewheel from them anyway. I don't plan on doing over 3000W, but 3000W spikes could happen occasionally.

I was on Grin's Rev4 torque arm page last night, and he wrote something about how his torque arm is designed to work both on 12mm and 14mm, but that it's stronger or locks more positively onto 14mm, so that's another factor.

http://www.ebikes.ca/product-info/torque-arms.html
[-]Does it matter if I have 12mm or 14mm axles?

No it does not matter, the torque arm plates are sized to accomodate a 14mm axle, but will fit a 12mm fine as well. With a 12mm axle, there is a small gap at the top or bottom of the slotted hole, but it is of no consequence and does not affect the spinout retention strength.

A 14mm axle has a much wider flat area and provides more spinout protection, both inside a dropout but especially inside a torque plate. However, a 14mm axle on front forks often leads to interference issues with the large diameter 14mm nut and washers, and it's a bad situation if the washers and nut don't stack up flat against the dropout.

Most rear dropout plates are flat and don't have this issue. For these reasons, we tend to spec our orders of rear hub motors with 14mm axles, and the front motors with 12mm so that the axle hardware can fit inside the front dropout cavity. The exception is the new Crystalyte H motors, which are 1/2" (~12.5mm) both front and rear.

So, just learning, I take it to mean that 14mm and 12mm here would be the length of the contact area of the part of the axle that is actually shaved-off to slot into the dropouts. I didn't know there was a variance in bikes. But I guess it is possible, because a Dropout provides an open end to slide it in, hence variable axle widths are possible, as long as the 2 flat sides remain equidistant and parallel. Hm. So, I take it, it's not so much the thickness of the cylindrical axle that's important at indicating strength here, but the amount of dropout-axle surface area that will contact the small area of the dropouts, which takes so much force. In other words, I take it the fear isn't so much snapping the center of the axle in two, but the axle twisting out of the dropout.

 

[mchlpeel]

As the 3mm2 phase and freewheel version seems to be the more robust option I have a question

What is the largest freewheel that the leaf can take on standard drop outs?

 

[Kodin]

Look at my previous post; I believe I have a post containing details on dimensions, if you need new measurements taken let me know. Another side note: the new-style cover for the cassette models will not take a brake rotor that has a sleeve that goes over the outer diameter. The one they shipped me for the freewheel version fits just fine. (Good example: I bought an Xon brake rotor that has a super thick aluminum center carrier. So thick in fact, that it extends onto the outside of the hub by a bit and is bored to do so.) If that makes no sense I can always add more pictures.... Just ask.

Also, Nep? They made the shaft as large as they could to still admit the freewheel. Standard bike hardware has set dimensions, and was never designed for this sort of application. Almost tempted to get a megarange freewheel and go to like a 60 tooth front large chainring if I manage to damage the freehub shaft ever. There's about the same amount of meat on both sides of the axle as well considering the through-drilling for the phase wires. It's just on the hub side it's much further toward the end of the shaft. Other option is I cut a new shaft out of CrMo eventually and get it vacuum-hardened.

 

[mchlpeel]

Hey kodin I read your post but you only mentioned that a 9 speed fits on 145-150mm

Standard width is 135mm

I'm thinking about adding one of em3ev's freewheels to my order. What size would fit comfortably.

Specs:

7sp:

11T-32T, 36mm wide, 0.53kg

8sp:

11T-32T, 40mm wide, 0.62kg

9sp:

11T-32T, 40mm wide, 0.58kg

10sp:

11T-32T, 40mm wide, 0.62kg

 

[Kodin]

As pictured above, this brings you to the ~130-135mm mark for this side of the shaft. Freewheels will add a bit more extra; if you're lucky you can stretch the frame a bit to admit a 8/9/10 speed freewheel easily. It's only 5-10mm or so. Lemme know if you need "perfect accuracy" measurements and I'll rig up a jig to do that.

 

[neptronix]

Well, i can tell you that it'll fit a 7 speed cassette if you use about ~2mm of spacers on each side.
Otherwise, 6 speed works as it's shorter.

 

[mchlpeel]

Ideally I need to fit a 9 speed as I have a nice deore xt 9 speed dérailleur and shifters.

The one em3ev supplies is 40mm so I'd need to stretch frame about 7mm or so?

 

[Kodin]

In theory; shaft necks down at ~130 mm, though you will want to add the included washers for better load distribution, so that puts you at 135mm. 7mm extra seems doable, depends on the frame among other things. Keep in mind bending aluminum frames induces stresses, so though I've seen people do it without problems, it's not officially/universally recommended or advised.

 

[SprocketLocket]

Okay dudes. Here is an excerpt of a conversation with LeafBike.
I'm doing a reality check with y'all, to see if what they are claiming is really possible.
They are telling me that they can ship me a motor configured to ANY no-load RPM I want. Is this possible?
Originally, they wanted me to give them any "WithLoad" RPM I requested, which they would then make to order, about which I'm even more incredulous, because there's no way they can know what load I'd put on it. Wanting consistency and verifiability, I clarified that I wanted the No-Load RPM's of each motor in the 200, 300, all the way through 600 RPM (I know the actual ~600's RPM, but wanted to compare their number with Nep's findings). I had assumed there was one in each ~100rpm range. Maybe that's not the case?

Yes,we can do it as Max
speed.(Withload)
ONLY tell us withload,What
max speed would you like?
We can do it.

Then, I ask just for No-Load, specific RPM's.

You just tell us max no-load rpm,we can do it.
Max rpm (No-load) ?
For example: 300 rpm No-load, when you ride on the flat road, speed will be
250rpm.If you climb hill,not steep,when current be 15A, it will be about 180rpm
* We recommend make a little faster speed motor. If it's too fast,our LCD also can set Max any speed.

Earlier in this thread are a list of windings with RPM's. [Edit: post was by Crossbreak, see link below. Did he reverse-engineer LB's windings?] Where did these data come from? Are they conjecture, or verified? I expected to get a repeat of this information from LeafBike, but did not.

from: http://endless-sphere.com/forums/viewtopic.php?f=30&t=66489&start=50#p1013925
361 rpm (9 *7 turn)
421 rpm (10*6 turn)
506 rpm (13*5 turn)
632 rpm (16*4 turn)

 

[SprocketLocket]

They're still telling me this same story. Although the person doing the emails seemed tech-savvy before, I'm now getting the feeling he doesn't know what he's talking about, or is blowing smoke. Or I'm wrong. How should I reply back?

No load RPM is Max RPM,when wheel work in the air. Current is about 1.2A. (I said No load that is No body ride it)
We customize motor as Max RPM No load.

The email continues with yet another explanation of an example of no-load, giving a theoretical RPM number and then explaining it would be slower when someone rode it, and even slower going up a hill (duh).

Further doubt cast upon this situation by crossbreak in the link in my previous post:

netronix ordered the 731 rpm one and got ~632 rpm no load @ 48 V

So my faith that they will send me a no-load spec I order is very low. I don't want to be stuck with a wind I didn't want. Looking for suggestions on what best to say to them or ask them.

 

[flipsx3]

I run this motor with the supplied controller from leafbike at 67v and get 51mph no venting drive it 10miles a day have put 500miles on the motor so far.

 

[neptronix]

Further doubt cast upon this situation by crossbreak in the link in my previous post:

netronix ordered the 731 rpm one and got ~632 rpm no load @ 48 V

So my faith that they will send me a no-load spec I order is very low. I don't want to be stuck with a wind I didn't want. Looking for suggestions on what best to say to them or ask them.

Their no-load figure was probably taken on a fully charged 48v nominal battery ( 54v or higher? ). What i paid attention to is the motor performance graph, which showed what looked like at 48.24v. This showed ~602 rpm under minimal load.

http://www.leafbike.com/products/di...w-rear-hub-motor-bike-conversion-kit-987.html

I'm pretty confident that the 5T model will produce about 500-550rpm under no load at the same voltage ( 48.24v )

 

[John in CR]

5 months of talking and 11 pages of posts with "high efficiency" in the subject of every post, yet no one can bother providing the simple info for Miles' motor spreadsheet. There's a lot more to great efficiency than thin lams. eg A significantly lower pole count can have greater effect, however, the biggie is a design that fits more copper on each tooth since copper losses are dominant, especially with hubmotors.

 

[SprocketLocket]

I re-read this entire thread thru the day, and took notes.
I tracked down the provenance of multiple no-load RPM figures.

731rpm
This one sources from LeafBike's product description on their bare motor page. It is apparently TOTALLY ERRONEOUS. This figure was erroneously taken by LeafBike from their own 1000W motor graph (wrong motor!), which wrong graph is also on that page (no wonder). This 731rpm figure has been the source of a lot of misunderstanding, confusion, and rationalization in this and other threads. Even Neptronix assumed they just used a higher voltage during a test to get this number. But if you look at the 1000W graph (link further down), you will see it.

632rpm
This was an actual measurement taken by Neptronix at 48V with a confirmed 16*4-Turn motor.

"About 630rpm"
This is supposedly what LeafBike claimed was the no-load RPM of the default 4-turn motor in a private email to ebikeaddict in March 2015. It is unclear if they're just repeating Neptronix's measurement someone may have sent them.
Source: post on page 6, 2015 March 21: http://endless-sphere.com/forums/viewtopic.php?f=30&t=66489&start=125#p1046427

602rpm
This is the no-load rpm found in the PROPER, 1500W motor graph, at 48V, assumably of the same 4-turn motor. I cannot explain the disparity, especially in light of LeafBike's own email quote to ebikeaddict above. We'd obviously trust Nep's experience more.
Source: http://www.leafbike.com/u_file/images/15_01_05/0a32577ead.jpg

"About 520rpm"
Estimate of the 13*5-Turn motor, provided by LeafBike to ebikeaddict. Source: same post as above.

Ebikeaddict, could you please post LeafBike's email verbatim?
TeslaNV, could you please do likewise?

---

The bare motor page for 1500W STILL has the incorrect graph and erroneous "731" RPM right in the product details. This is really bad, especially considering Neptronix's nastygram to them, and their commitment to correct it.
http://www.leafbike.com/products/e-bike-hub-motor/gearless-20-24-26-700c-28-inch/newest-48v-1500w-rear-spoke-hub-motor-electric-bike-motor-998.html

Their "kit" page for 1500W has the correct graph. Notice how there is no mention of 731 RPM. In fact, there is no mention of any RPM!
http://www.leafbike.com/products/diy-bike-conversion-kit/26-inch-electric-hub-motor-kit/newest-26-inch-48v-1500w-rear-hub-motor-bike-conversion-kit-987.html

I hope that clears things up, and absolves me of Neptronix's "what's a leafmotor?" dig.

Surprised no one has commented about their theoretically impossible 'made to order' no-load RPM. So 2 people here (ebikeaddict and teslanv) managed to get them to talk about winding numbers. Not sure if it's secrecy or ignorance at leafbike (they claimed secrecy to TeslaNV, but I just do not get that!). I suppose my tactic will be to play "no more mister nice-guy" and get them to fwd the questions to the factory.

---

Extrapolation time.
630 to 520, gaining a turn to 5.
520 is .825 of 630.
520 * .825 = 429 extrapolated RPM for 6-Turn.
429 * .825 = 353 extrapolated RPM for 7-Turn.

So, to put this in a list:
520rpm for 5-Turn (LeafBike estimate)
429rpm for 6-Turn (my extrapolation)
353rpm for 7-Turn (less-reliable extrapolation)

Comparing to CrossBreak's own extrapolations, based on Neptronix's 632 measurement:
506rpm for 5-Turn
421rpm for 6-Turn
361rpm for 7-Turn

You can see, the numbers are surprisingly similar.

CrossBreak-if you read this-I'm not a motor expert, could someone please describe how he was able to figure this back then?

Nep also estimated or guesstimated that the 5-Turn would produce 500-500rpm under no-load.

IMO we shouldn't have to go through these hystrionics, while LeafBike plays coy.


The numbers become even more strikingly similar with speed estimates. I estimated for 29-inch because that's what I have. (To change for yours, change the wheel circumference in the formula below.)

29-inch (91-inch circumference) no-load + estimated loaded speed [from crossbreak's estimations]:
361rpm x 60 = 21660 x 91" = 1971060 / 63,360 = 31.1mph x80% = 24.9mph
421rpm x 60 = 25260 x 91" = 2298660 / 63,360 = 36.3mph x80% = 29.0mph
506rpm x 60 = 30360 x 91" = 2762760 / 63,360 = 43.6mph x80% = 34.8mph
632rpm x 60 = 37920 x 91" = 3450720 / 63,360 = 54.4mph x80% = 43.5mph

New extrapolated no-load estimates for 29-inch wheel [from Leafbike's 630 and 520rpm figures]:
353rpm x 60 = 21180 x 91" = 1922730 / 63,360 = 30.4mph x80% = 24.3mph
429rpm x 60 = 25740 x 91" = 2342340 / 63,360 = 57.0mph x80% = 29.6mph
520rpm x 60 = 31200 x 91" = 2839200 / 63,360 = 44.8mph x80% = 35.8mph
630rpm x 60 = 37800 x 91" = 3439800 / 63,360 = 54.3mph x80% = 43.4mph

Very consistent outcomes.
Am I wrong to think that real-life estimates for low-rpm motors may be a little faster than 80%, due to more torque, and less wind resistance?

Merging & simplifying the 2 outcomes (real-world estimate [80% of no-load] for 29-inch wheel):
7-Turn: 24mph
6-Turn: 29mph
5-Turn: 35mph
4-Turn: 43mph

---

Per Neps' guess that the 1500W 5T would probably be similar to a MAC 8T, and since the MAC is a baseline many of us are familiar with, I couldn't help but wonder about my further speculation:
4-Turn: Similar to MAC 6T?
5-Turn: Similar to MAC 8T?
6-Turn: Similar to MAC 10T?
7-Turn: Similar to MAC 12T?
By "similar", I mean 'roughly analagous'--not in terms of exact performance, but more in terms of "if you're the kind of person who would buy this MAC version, you'd likely be in the market for this version of the 1500W".

---

Members known to have purchased the 1500W:
[Updated: thanks to mchlpeel]
Neptronix (16*4T, 632rpm no-load @48V)
FluxShifter (13*5T, 435rpm no-load @48V)
Dmwahl (13*5T)
Kodin (13*5T)
Burtie
Flipsx3

The latter 3, please post any measurements, or even any anecdotes, whatever you can give us, when you get them! I forgot to write down which version was bought, if it was said (I think Kodin got 5T). Kodin's controller should be arriving in several days.

---

"Digital laser tachometers":
Could someone tell me if this is the kind of thing yous were recommending as available on Ebay "for about 10 dollars". The cheapest of these kinds, by my search, began about $11.41 from overseas.
http://www.ebay.com/itm/USA-Digital-Photo-Laser-Tachometer-Non-Contact-Tach-RPM-Tester/201333601318

 

[mchlpeel]

Thanks for clearing all that up into one post.

After re-reading,

Neptronix has/had the 16x4T

Dmwahl has the 13x5T

Kodin has the 13x5T

unsure about Burtie

 

[spinningmagnets]

Awesome post, SprocketLocket. Thanks for the research and also for posting it here. What I most want now is for someone to find out what Leaf's internal nomenclature is for these four winds. It sounds like they want to sell bulk motors to an E-bike retailer who already knows what they want. They might provide good customer service and attention if you are buying 100 unlaced motors, but for the single-motor guy...good luck.

"Wheel diameter to speed/RPM chart"
http://www.endless-sphere.com/forums/viewtopic.php?f=28&t=16114

 

[FluxShifter]

I run a 1500w 13*5t. No load speed at 48v was 485 rpm. I hit 52 mph yesterday on HOC 18s 120% throttle slight downhill. Fastest I've ever gone on a bike. Twas a blast .

 

[mchlpeel]

What is the correct formula for extrapolating rpm based on voltage

For example

At 48v the 13x5t was producing 485rpm (assuming this is 12s?)

What would the rpm be for a 14s battery? (58.8v HOC)


EDIT

So worked out extrapolation for 14s (50.4v nominal)

485rpm / 48v = 10.1 Rpm per volt

10.1 x 50.4 = 509rpm

In 26" wheel 509rpm = 39.35mph unloaded

Or 31.4mph when taking 20% to get loaded speed

 

[SprocketLocket]

Welcome

FluxShifter, thanks for the info. How did you actually order (and prep the order) for 13*5T specifically? Could you possibly post key excerpts of communications you had with them?
We need to see what wording has "worked" with them, since a number of us have been told "use max rpm only" [not turns] to order.
I sent them a nastygram about this, and other things. We'll see how they respond.

I'm wondering if they actually sent you a 6-turn, or what. 485rpm is just a little above the average between LeafBike's stated no-load of the 5T (520rpm), and my estimate of the 6T (429rpm). <headscratch> Do you remember your exact voltage when you did the test?


Flipsx3, did you order the default motor? Do you know the safe voltage range of their controller? Is 67V the highest you've gone, and is that 18S LiFePO4? They give no voltage range on the page for their 48V 1500W controller.
http://www.leafbike.com/products/brushless-motor-controller/48v-lcd-motor-controller/48v-1500w-bike-motor-controller-700.html

 

[FluxShifter]

When ordering I stated 1500w 13*5t 500rpm wind. Add temp sensor. It's that simple. I had my client order one with a 24mm rim. When I inspected the wheel it had a flat spot and was cracked at the eyelet . Ordered the same part number rim from Jenson USA and received a much better quality rim. I did have to drill out the eyelets for the 12ga nipples PITA. Laced it up. Customer was happy. Still trying to get leafbike.com or DHL to pay for the damage.

My rpm measurment is on the low side due to drained battery so it's probably closer to 500 rpm no load. 13*5t is a good balance between speed and torque. Does get up to 110oC quick when you hit it with 5000w.

The leaf controller has 63v caps. It's good if your building a bike for a easy plug and play operation.

With 13s (54v HOC) I had a no load 526 rpm. I don't have a tach so these rpms are based off mph on my display. My 13*5t has a 10.1Kv averaging the 2 rpm readings I observed.