Leaf / leafmotor / leafbike high efficiency 1500w motor

[MLC]

Hi All,

Cool stuff. I hope you find time to collect data, CD.

I sometimes manage heat in electronic assemblies at work and have some thoughts...

Air flow through (or around) the sinks will play a major part in their efficacy. The pictured implementation (a dense array) might result in smooth air flow over the sinks (rather than through). If so, not ideal.

I'm guessing you're going for max surface area (good theory). But, half the surface area with turbulent flow through the sinks, could be better than 100% coverage with smooth air flow over the sinks. When sinks are effective, you don't have to worry about 100% coverage since heat will "run" to them. (Heat flow proportional to delta-T)

To address air flow and the contact area (sink-to-hub) issue, you might want to try (1) using long-narrow sinks rather than square, (2) leave a space between them and (3) mount them at alternating diagonals (slight "zig-zag").

Lastly, as Neptronix said, the thermal resistance between the sink and hub (how you fix the sinks) is a primary factor. The combination of narrower sinks plus thermal epoxy would make a very big difference (over squares with foam).

Found these after about two minutes of searching. I didn't look closely; there may be better choices out there...
http://www.digikey.com/product-detail/en/assmann-wsw-components/V5619A/AE10812-ND/3511388
http://www.omega.com/pptst/OB-100_OB-200_OT-200.html

Maybe you'll consider these ideas in one of your experiments. Have fun!

Note: I plan to make my first, 16 mile commute via ebike next week. (1500W, 5T LB and 14S on GT Avalanche)

Best,
MLC

 

[Cowardlyduck]

Thanks for the feedback MLC.

There are plenty of way's this can be done better, I agree, however I just threw this together very quickly to just see how it goes initially.

It's funny you should mention having them more spaced out as that's what I originally did with exactly what you say in mind. I was just trying to do it by eye though and I found it too difficult to space them evenly which looked crap, and I was a bit worried about the effect it might have on the wheel's balance.
I might re-do it at some stage and remove every second heat-sink and re-secure it all with thermal epoxy.

One of the reasons I didn't use thermal epoxy to begin with was, I couldn't find any reasonably priced, in large quantities, with high thermal conductivity and no requirement to heat it to set it. The best stuff in the link needs to be heated up to 205C to set....not doing that to my hub! The other stuff (OB-101) that sets at room temp looks good, and I might get some. Thanks!

I actually have some good Arctic Silver thermal epoxy already which I used for this:

But I used nearly all of the 10g tubes to do that job and it's not cheap.


Cheers

 

[Cowardlyduck]

I did some backyard testing of the motor heat-sinks today.

Well, they're not that great...yet!
It seems everyone's hypothesis on the ineffectiveness of the thermal padding was correct, and not much heat is really getting through to the heat-sinks.
When the side covers are very hot (not quite too hot to touch, but close), the heat-sinks are only just warm. This is when the motor is about 90C on the windings. If I let the bike sit for a while, then the heat-sinks got hotter, but still not as much as the side covers.
I'm not going to be ready to remove and do it all again for a while, so more riding will be done with it like it currently is first.

When I do get around to fixing it up/making it better, I think I might use some more Arctic Silver Epoxy. I used it for my controller heat-sink attachment and it's working great.
This is the stuff I bought last time:
http://www.ebay.com.au/itm/Arctic-Silver-Thermal-Adhesive-paste-7g-epoxy-artic-/221289889725?
I've since found it slightly cheaper for a larger quantity:
http://www.ebay.com.au/itm/Arctic-Silver-Alumina-Thermal-Permanent-Adhesive-5g-Kit-/280853528070?

I think, I might be able to make 15g (3 x 5g), plus what I still have left from the controller mod, stretch across 3 motors (this Leaf, my HS4080, plus a geared motor that keeps getting too hot) if I half the number of heat-sinks as that would leave about 12-15, and if the turbulent air theory holds true, be even more effective.

Cheers

 

[Cowardlyduck]

If you have some wire you may be able to strap them down tightly- it pressie makes all the difference with the pads. Thanks for sharing.

The difficulty is in making it tight. I tried this with the string in the above example, and where it gets difficult is when you need to tie it tight in-between the heat-sink 'teeth'.
I did consider using a brake cable inner, however it would be very difficult to tie it into a knot to make it tight IMO.

I've ordered the 3 x Arctic silver now, so I'll try that next I think.

Cheers

 

[spinningmagnets]

I am a fan of the Leaf motor 1500W 205/H35. That being said, I am aware of many new ebikers who want to "jump in" and figure they will do more homework later.

If your hub is getting very hot, there is a chance you are running it frequently in its inefficient range. Improving the heat shedding is (of course) an option. However, doing that ensures you will never get as much range from your battery as you could, compared to fixing the generation of waste-heat in the first place.

The next step up in hub size is the MXUS 205/H45, but it also is wide enough that most owners only use a single-speed (although two sprockets will fit), so...I want to compare a "cooling added" Leaf to the MXUS.

The Leaf is lighter, but only slightly less expensive. However, I am told it easily fits a 6-speed freewheel into common 135mm drop-outs. If you want that, then the choice is clear.

For Leaf systems, the next consideration should be how small of a wheel would you be willing to use? A 5-turn Leaf in a 26-inch wheel might be roughly the same top-speed as a 4-turn Leaf in a 20-inch wheel with a fatter 3.0 inch tire (with both systems at the same voltage). In this situation, the 20-inch wheel would automatically run cooler, more efficiently, and provide longer range from the same battery.

Once the wheel diameter and voltage are settled on by each builder, some may still experience enough heat that they are now considering ventilation or Ferro-Fluid.

I recommend starting with FF. You can always add ventilation later, and if you decide that you want to use only ventilation, the FF can be removed. I doubt moisture incursion will ever be an issue, since hot motors will boil off any condensation from humidity that might occur.

 

[madin88]

I doubt moisture incursion will ever be an issue, since hot motors will boil off any condensation from humidity that might occur.

following situation:

if the motor heats up, air goes out because of the pressure.
now you ride in rain and everything is wet (where the wires going in, on the bearings,..)
if the motor now cools down, it will suck that water in because the pressure inside the motor gets lower as outside
so air will always move in and out and the sealings get stressed.

a pressure valve (like on peli cases for instance) probabaly would make a sealed motor lasting forever. such valves often are installed on water proof gadegets (the cycle satiatior has one as well)
will try it definitely out

 

[Samd]

A Mxus 45H will fit a screw on freewheel, Ron. It's just that over 2kw it seldom makes sense.


Sent from my iPhone using Tapatalk

 

[Cowardlyduck]

Here's the video of the ride myself, Lash, 9lives and his son did on the weekend.
I was riding using my newly heat-sinked Leaf motor:
[youtube]goGQPJwP__g[/youtube]
Still uploading now, but should be done in an hour.

I was pushing 5KW peaks, and only hit my thermal rollback of 135C once, and mostly managed to keep up with the other guy's setups running much more powerful (and heavy) MXUS, HS4065 and QS motors.

I did a hand check of the heat-sinks a few times throughout the ride and they did seem to be getting hot along with the motor. I also did some further testing at home and they do definitely take in some of the heat even with the crappy thermal padding and string.
I also noticed a slightly faster cooling off time when pausing at the top of hills and cruising, so I would say they are definitely helping.

There was also 0 noticeable change in weight or handling of the rear wheel, which is to be expected with less than 100g of heatsinks. This is important to me as all my efforts (venting, Ferro Fluid, Fan Cooling, and Heat-sinks) with the Leaf motor have been about performance without the penalty of added weight and so far it seems to be paying off.

Cheers

 

[CobraJet]

Cool video! What was the smoke all about, Lipo fire or?

BTW, where was this?

 

[Cowardlyduck]

Cool video! What was the smoke all about, Lipo fire or?

Thanks!
It was a BMS/Balance wire fire, not LiPo.
Details and discussion here:
https://endless-sphere.com/forums/viewtopic.php?f=2&t=61183&p=1197111#p1196526

BTW, where was this?

It was in the South of Canberra along several ridge lines here and here.

Cheers

 

[GmagNeato]

Cool video! What was the smoke all about, Lipo fire or?

Thanks!
It was a BMS/Balance wire fire, not LiPo.

Oh I was under the impression that one of the lipo packs blew out when he said the corner of the pack "let go"

If it was as you say a lead that disconnected from the main harness then it should definitely be preventable. I know I'll be checking and reinforcing all my series lead connections tonight for sure! And ordering lots of Co2 cartridges as well! Thanks for sharing this CD!

 

[ShfyTwo]

can someone tell me how to wire the 10K Thermostat? I have the white wire from the hall connector, where do I ground? My 1st time hooking this up.

 

[Jabotical]

So is there any potential practical advantage to the 1000w variant of this motor over the 1500w one? I've read through this whole thread, and it sounds like the consensus is no, but I'm a bit new to many of these concepts.

I recently got sucked into the whole online ebike world, and have been devouring things here and elsewhere for a few days. Perhaps I should have lurked a little longer before pulling the trigger, though: For this first ebike build of mine, I liked the sound of the reportedly higher efficiency and reduced cogging of the leafbike motors, so I set my teeth against the high shipping cost and ordered a 1000w one from them last night. I was aware of the 1500w one, but since it was a tad bigger and a couple pounds heavier, and since I'm not really interested in going above 30mph (for at least a while), I figured I'd be better off starting with the smaller one.

But now that I've read through more (all) of this thread, it seems as though the torque starting off may suffer in comparison to the 1500w model? That's really my main concern (along with maintaining <30 mph top speed for as long as possible, of course, with or without pedaling). I (naturally) want the option of zippy acceleration. Admittedly I'll have little to compare it to, but I did start wondering then if the regen braking would be less effective as well, etc. A couple pounds either way will likely not make much of a difference, and neither does $50 for me.

So should I be frantically trying to get a hold of leafbike to change my order, if that's still possible? Or is there something I can tell myself to feel better about having the 1000w version? I'm a little under 200 lbs and I live in Phoenix where there aren't any hills to speak of. I got a 35A controller to go with it, and one of lunacycle's 52v batteries. But if the 1500w version is more efficient to accelerate/maintain speed relative to its weight, then it seems like it's a no-brainer regardless. If so, why do they even have the other, really? Are there any meaningful tradeoffs in favor of the 1000w variant?

Thanks for any insight/reassurance/condolences!

 

[GmagNeato]

So is there any potential practical advantage to the 1000w variant of this motor over the 1500w one? I've read through this whole thread, and it sounds like the consensus is no, but I'm a bit new to many of these concepts.

I recently got sucked into the whole online ebike world, and have been devouring things here and elsewhere for a few days. Perhaps I should have lurked a little longer before pulling the trigger, though: For this first ebike build of mine, I liked the sound of the reportedly higher efficiency and reduced cogging of the leafbike motors, so I set my teeth against the high shipping cost and ordered a 1000w one from them last night. I was aware of the 1500w one, but since it was a tad bigger and a couple pounds heavier, and since I'm not really interested in going above 30mph (for at least a while), I figured I'd be better off starting with the smaller one.

But now that I've read through more (all) of this thread, it seems as though the torque starting off may suffer in comparison to the 1500w model? That's really my main concern (along with maintaining <30 mph top speed for as long as possible, of course, with or without pedaling). I (naturally) want the option of zippy acceleration. Admittedly I'll have little to compare it to, but I did start wondering then if the regen braking would be less effective as well, etc. A couple pounds either way will likely not make much of a difference, and neither does $50 for me.

So should I be frantically trying to get a hold of leafbike to change my order, if that's still possible? Or is there something I can tell myself to feel better about having the 1000w version? I'm a little under 200 lbs and I live in Phoenix where there aren't any hills to speak of. I got a 35A controller to go with it, and one of lunacycle's 52v batteries. But if the 1500w version is more efficient to accelerate/maintain speed relative to its weight, then it seems like it's a no-brainer regardless. If so, why do they even have the other, really? Are there any meaningful tradeoffs in favor of the 1000w variant?

Thanks for any insight/reassurance/condolences!

Looking at the website the efficiency is 90.5% for both options. The 1000w is lighter (by 2lbs), more narrow (by 5mm), and cheaper (by $25). I think the larger magnets and higher copper fill (presumably) of the 1500w outweigh the reduced weight and cost of the 1000w IMPO.

I doubt it would be a problem for them to change your order if you've only just submitted it last night, but I'd do so quickly. You also need to specify to them which turn count you'd like. I think they ship the 4T as standard. For 30mph at 52v the 5T will be a better option again IMPO. The 5T leaf motor is now on the motor simulator at http://www.ebikes.ca/tools/simulator.html. A great tool to get an idea what to expect at different voltages, amperage, loads, etc. Good luck!

 

[Jabotical]

Looking at the website the efficiency is 90.5% for both options. The 1000w is lighter (by 2lbs), more narrow (by 5mm), and cheaper (by $25). I think the larger magnets and higher copper fill (presumably) of the 1500w outweigh the reduced weight and cost of the 1000w IMPO.

I doubt it would be a problem for them to change your order if you've only just submitted it last night, but I'd do so quickly. You also need to specify to them which turn count you'd like. I think they ship the 4T as standard. For 30mph at 52v the 5T will be a better option again IMPO. The 5T leaf motor is now on the motor simulator at http://www.ebikes.ca/tools/simulator.html. A great tool to get an idea what to expect at different voltages, amperage, loads, etc. Good luck!

Thank you for responding so quickly! I really appreciate your confirmation, and the valuable suggestion about turn counts that I was too intimidated to even ask about.
I started playing with the ebikes.ca simulator, which is indeed a neat tool. I can't really use it to compare a 5T leafbikes 1500w motor vs a 4T one though, because there's no 4T option, right?
If you (or anyone) happens to see this in the near future (I only have a couple hours to respond to leafbikes), what then is the tradeoff between 5T and 4T? Presumably since 5T's top speed is best suited to ~30 mph, I get better low-end torque compared to the 4T winding, whose natural/top speed is higher?

 

[markz]

4T would have a higher RPM per voltage then a 5T. You could compare the MXUS 3000 motors would give you similar differences and compare to the Leaf graph. Also if you put a 4T in a smaller diameter wheel you get a cooler running motor with more torque, but that smaller diameter rim then reduces your speed. Trade offs.

 

[Jabotical]

And the advantage of a lower RPM per volt is presumably higher torque at the same voltage? Or is it more about increased efficiency at those RPMS?

Yes, it's always all about the tradeoffs. Makes sense that speed is reduced and torque increased with smaller diameter rims. I'm beginning to remember some of my relevant coursework from long ago.

Now that I've updated my order to the 5T wound 1500w version (partially on faith), I will fiddle with that simulator more and read about windings. The physics of it interests me, I just haven't been exposed to the details of motor internals much yet in spite of being technically-inclined professionally and personally.

I don't see any MSUS motors on the simulator, btw -- there are several 4xxx MXUS ones, if that's what you meant?

I really like how you can superimpose two scenarios on top of each other in that tool. Comparing against some of the geared variants really cements for me the advantages and disadvantages either way for those types of systems.

 

[markz]

Found the Leaf dimensions here, had to click on the drop out length picture.
Someone should add it to the first page.
http://www.leafbike.com/u_file/images/14_06_01/6cf2988173.jpg

Im gunna splurge on the Leaf 1500W package deal...
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

Is their 12 FET 48V controller any good?
Whats the number on the FETS? 3077? 4110? 4115?
Programmable?
Hows the display?
Rim comes warped/damaged or what?

 

[GmagNeato]

Found the Leaf dimensions here, had to click on the drop out length picture.
Someone should add it to the first page.
http://www.leafbike.com/u_file/images/14_06_01/6cf2988173.jpg

I was able to cram a 7 speed cassette in mine with 135mm dropouts.

Im gunna splurge on the Leaf 1500W package deal...
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

Is their 12 FET 48V controller any good? Hasn't failed on me yet have about 800 miles on it. You will need to change the fuse though as I blew all the ones that came with it in the first few rides. I believe I went with 40a and haven't blown one yet. I've run it at 48v, 52v, and even 60v (all nominal). I've found the max wattage it will ouput is 2000w.

Whats the number on the FETS? 3077? 4110? 4115? Not sure as I haven't opened up the controller yet. Leaf wouldn't tell me either. They just said "high quality" :lol: I think I'll pop mine open to take a peek soon as I'm not currently riding this bike and will post my findings.

Programmable? It is somewhat programmable from the display, but it's a fairly cumbersome process. Leaf only includes a basic manual in chinglish, but there are better ones out there. Here's a good one http://www.trikefactory.de/download/KT-LCD3.pdf

Hows the display? Not bad. Backlight is nice and bright. Gives you all the info you should need including a watt meter. An amp meter would be a nice addition though.

Rim comes warped/damaged or what?

Not really sure as I changed wheels and relaced immediately upon receiving mine (they sent the wrong wheel).

 

[GmagNeato]

And the advantage of a lower RPM per volt is presumably higher torque at the same voltage? Or is it more about increased efficiency at those RPMS?

Yes, it's always all about the tradeoffs. Makes sense that speed is reduced and torque increased with smaller diameter rims. I'm beginning to remember some of my relevant coursework from long ago.

Now that I've updated my order to the 5T wound 1500w version (partially on faith), I will fiddle with that simulator more and read about windings. The physics of it interests me, I just haven't been exposed to the details of motor internals much yet in spite of being technically-inclined professionally and personally.

I don't see any MSUS motors on the simulator, btw -- there are several 4xxx MXUS ones, if that's what you meant?

I really like how you can superimpose two scenarios on top of each other in that tool. Comparing against some of the geared variants really cements for me the advantages and disadvantages either way for those types of systems.

I'm no expert on these concepts, but here is the way I understand it. Ultimately, given the same amount of copper in each motor, the turn count dictates top speed at a given voltage. Nothing more. To increase torque at that voltage, you need more amps (which means beefier wiring and controller) and/or a smaller wheel (which as Markz stated will reduce your top speed accordingly). You can offset the reduced speed by increasing the voltage, but you said you've already went with a 52v battery (good choice) and a 35a controller. I'm assuming you are going with a 26" wheel. Your selection of battery (52v) works well with the 5T in a 26" wheel for speeds at 30-35mph (it's nice to have a little headroom) in my experience. It won't be a wheelie machine, but I think you'll be pleased with the performance.

 

[Jabotical]

Thanks very much Gmag! I don't need a wheelie machine for my first attempt/commuter, but a little headroom is indeed most welcome. Your reply made me feel good about my (already ordered) selections, which is of course what I was hoping for. And cemented several important concepts regarding turn count.

I went on to further research the effects of different windings on the (otherwise) same motor, to follow up your excellent introduction. The Canadian ebikes site unsurprisingly ended up having one of the more clear, accurate, and in-depth treatments of the subject, complete with meaningful equations: http://www.ebikes.ca/learn/hub-motors.html. Elsewhere the internet seems rife with misinformation and slightly skewed understandings. Claims that higher windings result in increased torque (relative to input power) are rampant, as are assertions that the lower current of a higher turn-count motor (lower K) causes less heat and thus better performance.

As the linked page points out, even though for a given resistance less current does mean lower power losses to heat, when you reduce the current necessary to produce a given torque (and speed) by increasing the motor turn count, you do this by increasing the resistance (because you have longer, skinnier wires) by exactly the amount which cancels out the efficiency gains of this reduced current. So rewinding the motor (with the same copper amount) indeed doesn't change the motor's fundamental performance in any way -- the same power is required for the same torque and the same speed regardless of the value of K.

What, then, is the point of different windings? That was the question I was initially struggling with. I.e. why not always use motors with the highest K values possible, to allow for the highest top speed without impacting lower speed performance in any way? First, it does appears that there may sometimes be a little bit smoother performance at lower rpms to be had with higher turn count systems, which is not surprising really given the higher pole densities. Mainly, though, I gather it mostly comes down to a tradeoff of more current vs more voltage for a given performance. Twice the turn count, for instance, means the same torque and speed can be had with half the current and twice the voltage.

Since we tend to have systems that operate at fairly fixed voltages, and batteries that are limited in how much current they can source, it makes sense to me that it would be desirable to arrange for the lowest K (highest turn count) motor variant you can get that will provide sufficient top speed at your voltage / wheel size (which is essentially the advice I was given). Because then your battery doesn't have to provide as much current to give you the torque you want within the speed range that's important to you... and this might in fact in some cases legitimately allow for more torque at those lower ranges -- not more torque per watt, but again since the voltage tends to be constant, more torque per amp is more significant than I had initially considered. (Though as you said, it does also remain an early-design option to increase the voltage as a parameter if the components can all handle it)

I apologize if this is growing off-topic for this thread, but the question and your answers have been very illuminating for me, including in regards to this particular motor and why someone would want the 5T vs the 4T, etc. version.

As a side confirmatory note, leafbikes was immediately obliging about switching my order to a 5T 1500W version, and just charged me the difference, even though they'd already packed up my previous order. I have to endorse them from a customer service perspective because they were so accommodating.

 

[Cowardlyduck]

Jabotical, your right this talk about the best wind motor has all been discussed previously at length, but the discussion on this Leaf Motor thread is nothing compared to the similar discussion which happened on the MXUS thread:
https://endless-sphere.com/forums/viewtopic.php?f=30&t=63142
I'm not sure where exactly they started talking about the different winding options, so you might need to skim through it to find it, but it's well worth it if you really want to understand the differences better.
From what I learned, my understanding is it is almost always better to go with a lower turn winding except when these limitations exist:
-Lower amp capable controller
-Lower copper fill
-Very large wheel (i.e. 29er)
-Every last bit of efficiency matters to you

Apart from those reasons, I would say you are always better off going with a lower turn wind motor. The battery amps doesn't matter as it's the phase amps that come into play when using a lower turn motor, which is why the controller amps matters.

Anyway, on another note, this is what my Leaf motor looks like now:

As suggested, I re-attached the heat-sinks with Thermal epoxy and spaced them out a bit. I spaced them just by sight, so on the other side they get a bit squashed up together as the last few didn't quite fit perfectly. Hopefully it doesn't throw the balance off.

This should really help maintain the light weight, but semi-powerful capabilities of my Fighter and allow 5-6KW bursts.

Cheers

 

[Samd]

This is a very good mod CD - the induction in the magnets causes a lot of heat.
I saw Sketch doing the same on ES facebook also- and the side by side comparison on his qs205 versus an unmodified one was orders of magnitude different.

 

[Cowardlyduck]

Thanks Sam. Yeah I also saw that Facebook comparison although he managed to find some radiator heatsink fins that looks like they might be easier to mount.
Unfortunately I'm not running FF any more, so the heat transfer from the stator windings won't be very high, but I did not realise the induction also generated heat, so it's good to know it will provide some benefit.
In order to do mine, I had to attach them 1-2 at a time over a long period so that they could adequately set before rotating the wheel to do next. I did the same on my HS4080, but with only 8 heatsinks.

Cheers

 

[neptronix]

From what I learned, my understanding is it is almost always better to go with a lower turn winding except when these limitations exist:
-Lower amp capable controller
-Lower copper fill
-Very large wheel (i.e. 29er)
-Every last bit of efficiency matters to you

I totally agree with that. The advantage of having a low cell count means that you can use one battery for a lot of different motors of different power ratings ( as i do! ), and you have more charger options available to you, too.

Also, you can adjust your power level much easier by varying amps instead of volts. I hate reconfiguring RC Lipos just to make one motor happy.

I've maxed the power out on motors with high turn, medium turn, and low turn counts for years. I don't notice the difference.

The key to getting high efficiency out of your high turn count setup is to have huge battery wires and connectors + the lowest voltage, hence lowest resistance FET you can buy ( 3077's ). Low turn count motors suck a lot of amps and the efficiency loss usually just comes from resistance somewhere between the battery and the controller.

Also, you need a stout battery.. battery inefficiency ( sag ) is yet another thing that makes people think that low turn count motors are inefficient.

Spec your system out right and the low turn count motor will be as efficient as any other.