Leaf / leafmotor / leafbike high efficiency 1500w motor

[ABritInNY]

That's motor power, not controller power.

I'm talking about controller power ( which is what you see on a cycle analyst )

I too have the 4T motor that you have, which I finally got working to full power for the first time today, (thanks 4 the suggestions Tek,) I also have a supposedly calibrated CA, but mine is the reverse setup to your graph, with 80v batt * 64a controller setup, but wouldn't the wattage be the same? @around 5kw.
Now I'm trying to make sense of this graph data to best setup this motor. I'm trying to work out it's fastest sustainable constant speed in a 26" rim, and compare it to my tests today. The roads aren't flat around me, so it's different to any simulation, but I think mine pulls like 2400w? at 40mph on flat ground and I'm quite light, only 125lbs, OR perhaps the tears were too heavily flowing lol and I saw incorrectly.
When I disconnected the CA and used Runtastic with GPS it said I went 54mph but took near 6kw of power to do it, and I hit 104C in less than 10 mins riding I like fast, but I NEED to understand this simulator stuff for my motors sake
Wow, seems as soon as you digest one hard topic, the next confusing situation rears it's head!
Do you possibly have a 'safe' top power/speed suggestion for the 4T? One that lines up with both your experience AND what the simulator says (or sumthin in between poss.)

 

[ABritInNY]

Actually, the motor power (or actually load) is sort of the key here.

If you ignore the motor simulation entirely and just look at the black Load Line you see the power necessary to propel the bike at various speeds ('load' on the sim tables). 'Motor Power' equals 'Load' (the power to propel) at the vehicle terminal max speed. In your plot we see that for the configured bike it takes 2277W to push the bike at 40.2 mph -- just to overcome rolling and aero drag.
Even with a 100% efficient motor/drive that bike cannot reach 40mph on 2000W.

The Load Line calculation is basic physics and has nothing to do with the motor simulation. It uses the same equations as every other bike calculator.

So - since the 'bike calculator' Load Line part must be correct, that means GIGO is at work (garbage in, garbage out) and the configured parameters for the vehicle are incorrect.
For instance:

• if you measured your speed in a tuck, then the Power to Propel drops from 2277W to 1610W.
• if your test was also on a barely perceptible 1% downgrade, then the Power to Propel drops further from 2277W to 1439W -- a whopping 838W difference.

Interestingly, the tuck case above requires 1610W/.8 = 2012W if the motor/drive is running at 80% efficiency.

Anyhow - at higher speeds, there are some slippery effects that can creep in considering that major determining parameters for the Load Line (not the motor simulator) are unmeasured guesses.

I'm trying to see if I'm starting to understand this, by my understanding, if we wanted to see the exact real power draw, couldn't we just set the CAv3 into speed throttle mode, set 40.2 mph as the speed limit on a windless day on flat ground, while tucking in as best as possible, and look to see what power draw it settles to continuously after the initial acceleration peak is over? or is that all wrong?

 

[Kodin]

Keep in mind shunts may have up to 5% deviation and therefore what you "measure" with one CA may not be the same as another. Also depends on ambient temps as well. Only way to know for sure is to use a higher end meter, (4-wire measurement preferred) to measure actual resistance values as accurately as possible across the shunt/wires/connectors and program in the shunt value offset into the CA to help compensate. If the shunt heats up in operation, may want to take a second measure to see what the value changes to at operating temps as well.

 

[ABritInNY]

I don't have an external shunt, and use my built in controller shunts, of which there are 3 in series according to the builder,
Lyen 18fet 4110 dropping the 4ohm to a 1.333 intern resistance. But even knowing this abt my controller and my custom batt details, each cell apparently has a ≤1.5mΩ (AC impedance, 1000 hz) and I have 24s1p 30ah lifepo4 pack but IDK how to calculate the impedance's the sim needs? Also I have the 4T in 26" and all I'm looking for is it's most efficient speed band, so as to have an efficient yet powerful setup, that won't overheat the motor so quickly. With my current knowledge level it's painfully obv to me, that all this is @best 25% understood by me any easier help explaining would help me alot.

 

[Kodin]

The speed band is affected by a lot of things, but the main one would be ambient temperatures and heat build up in the motor. If you don't already have venting or ferrofluid, get some ferrofluid and drop in the required amount. (somewhere around 5 ml, add 4.5 and start adding tiny amounts and testing until you see your unloaded max speed wattage start to creep up by 5-10 watts)

 

[ABritInNY]

Yes I saw Justin 5ml test data ferrofluid on another thread, but just like venting it too requires drilling (if only 1 small one, and might be messy,) plus careful silicon sealing of the cover, I think I may just opt for the 4T Mxus v3 with aluminium stator for better cooling but after reading Neptronix's comparison between it and the leaf, I'm not so sure :?

 

[Kodin]

Thermal grease works too. Also you could simply put 5 ml with one cover off and close up the whole thing; wouldn't be exact but it sure would be close...

 

[madin88]

Yes I saw Justin 5ml test data ferrofluid on another thread, but just like venting it too requires drilling (if only 1 small one, and might be messy,) plus careful silicon sealing of the cover, I think I may just opt for the 4T Mxus v3 with aluminium stator for better cooling but after reading Neptronix's comparison between it and the leaf, I'm not so sure :?

with all respect to neptronix, but i would not give that much on his comparsion (the Leaf was vented while the MXUS was sealed iirc)

 

[John in CR]

Yes I saw Justin 5ml test data ferrofluid on another thread, but just like venting it too requires drilling (if only 1 small one, and might be messy,) plus careful silicon sealing of the cover, I think I may just opt for the 4T Mxus v3 with aluminium stator for better cooling but after reading Neptronix's comparison between it and the leaf, I'm not so sure :?

with all respect to neptronix, but i would not give that much on his comparsion (the Leaf was vented while the MXUS was sealed iirc)

Not to mention that he's the fanboy who started this thread with the ridiculous title calling it a high efficiency motor when it clearly isn't.

 

[John in CR]

Thermal grease works too.

What is that supposed to mean?

 

[Allex]

Thermal grease between covers an magnet ring, gives you slightly better heat transfer out to the covers

 

[John in CR]

Thermal grease between covers an magnet ring, gives you slightly better heat transfer out to the covers

I'd argue that the detriment of opening a factory sealed motor to do that would outweigh the tiny gain, and if someone measured an actual gain it was probably on a motor that had been opened many times, a big no-no AFAIC. It's penny wise and pound foolish, because focusing on something like that takes away from efforts where the real gains are. eg A controller setting change where you couldn't perceive a difference in performance could easily make an order of magnitude more difference in terms of heat. Then there's wheel size, where a tiny bit smaller wheel would have a similar effect. Of course, then there's the third elephant in the room, which is use an actual high efficiency motor.

I run more power through hubmotors than anyone, so I have to address heat more than anyone. There's about as much chance of me putting thermal past on a side cover flange as there is of me going to the trouble to polish that interface for an even greater benefit. I prefer to focus on the things that matter, maybe that's why despite running insane power and pushing bigger loads that I probably have a cooler motor than anyone running 3kw or more through a hubbie. :wink:

John

 

[Kodin]

Not exactly what I meant; you seal the side covers with thermal paste rather than silicone sealant; accomplishes the same goal of sealing the covers, and gives you a better interface than (generally thermally insulative) silicone sealer. Also, the ferrofluid tests have been shown to be as effective as venting a motor. Regarding the thermal paste, plain old grease would likely do just as well for actually sealing. It's just if you're going to seal, might as well seal with something conductive. Yes, polishing the covers' interface joints would likely be more beneficial, however not too many people can do that on their own. Further, polishing will likely very quickly increase the press-fit tolerances of the covers to the rotor, possibly causing issues with central alignment of the rotor if one goes too far. Regardless, there's nothing magical about popping open a factory-sealed hubmotor. If done right, there's literally no difference when closed back up. Further, painting the inside of the covers matte black seems to improve cooling slightly, and adding ferrofluid adds a noticeable increase in cooling ability. If you're pushing massive power through a motor and exceeding it's ability to cool, then yes, you'll likely have to switch to fluid circulation rather than passive options like ferrofluid or ATF. At that point however, you're far better off moving to a larger motor rated for the current you are pushing. At the power levels the Leafbike motor can run, it's pretty damn efficient compared to most of what's out there.

 

[GmagNeato]

.. At the power levels the Leafbike motor can run, it's pretty damn efficient compared to most of what's out there.

Proven truth. Not just on paper. I have done none of the cooling mods mentioned. I typically run at an average of 1500w with frequent peaks exceeding 2000w, in a 26" wheel with the 5T, and I have NEVER seen temps exceeding ambient, even on 7-10 mile rides on 70 degree days

I will be pushing more power through it soon, and have decided that if I do any cooling mods it will be the ferrofluid. Imo, it seems like the biggest return for the smallest investment of all the mods I've seen.

Kodin, if you don't mind, what are your power levels/wheel size and what speeds are you seeing with your 5T? Have you modded it at all?

 

[Allex]

What tempsensor is inside leaf motors?

 

[GmagNeato]

What tempsensor is inside leaf motors?

I'm not sure what model it is. I haven't opened mine up yet. I've been wondering this as well for Adaptto compatibility reasons, so I will need to find out. The temp does fluctuate with ambient though and whenever I've checked for calibration it's always been within a degree or so.

 

[madin88]

I typically run at an average of 1500w with frequent peaks exceeding 2000w, in a 26" wheel with the 5T, and I have NEVER seen temps exceeding ambient, even on 7-10 mile rides on 70 degree days

There must be a problem with the temp sensor because you should see at least a few degree more as ambient.

 

[Allex]

He probably means side cover temps which is totally different from the insides

 

[GmagNeato]

I typically run at an average of 1500w with frequent peaks exceeding 2000w, in a 26" wheel with the 5T, and I have NEVER seen temps exceeding ambient, even on 7-10 mile rides on 70 degree days

There must be a problem with the temp sensor because you should see at least a few degree more as ambient.

My mistake. I worded that all wrong. I should have been more precise. The temp reading fluctuates consistently with ambient temps. It typically does exceed it while in operation (the motor is warmer than the air and internals warmer still), but this variance is never very dynamic. But again, these are pretty modest power levels.

I can't really say how accurate the sensor is, but I believe it's working well enough to give a relative idea of performance.

 

[John in CR]

Whatever model it is, it's the magic kind that helps the motor defy the laws of physics to remain at or below ambient temperature.

 

[Allex]

Got my Magic leaf.
Not sure if it is 30 or 35mm it looks super thin, I am used with motors like mxus3k
Ordered 1500Watt 35mm stator and asked them to lace it in extra strong wheel and throw in extra rim and spokes, this is what I've got. Is it same rim as you guys get from them or is this one different?
Dont have the calipers to measure but I asked for phase wire upgrade and they look thicker than Mxus.

Thats a 12G spoke

 

[ABritInNY]

I typically run at an average of 1500w with frequent peaks exceeding 2000w, in a 26" wheel with the 5T, and I have NEVER seen temps exceeding ambient, even on 7-10 mile rides on 70 degree days

There must be a problem with the temp sensor because you should see at least a few degree more as ambient.

I think he may be possibly be confusing his C and F degree scale, if the temp says 75 on his CA from the sensor, its C, and that would be approx 180f, of the top of my head... just a guess tho...
Ok read a lil further, seems I agree with the sentiment that you must have a magic motor.
Don't get me wrong I LOVE the leaf, the power is amazing and if you keep to it's spec sheet it IS really efficient and can run over 2kw constant @over 80% efficient, so if that's all you want power wise you won't need ferrrofluid or any other cooling mods. It's the higher power draws, associated heat probs, that we're discussing how to better manage here.

 

[ABritInNY]

What tempsensor is inside leaf motors?

Mine came wired with a 10k therm in the windings, not linear so no prob math

 

[Allex]

Haha,
The PVC coating of the wires is good up to 105 °C so this isn't good for very hot stuff.
Thermistor temperature range -55 °C to 125 °C
Guess I will have to swap it for adaptto thermistor kty84

They just tucked it in, no thermal epoxy, nothing, so the measurements will def. not be accurate.
Of course the halls are frocked up, the middle one is deep below

 

[Cowardlyduck]

Got my Magic leaf.
Not sure if it is 30 or 35mm it looks super thin, I am used with motors like mxus3k
Ordered 1500Watt 35mm stator and asked them to lace it in extra strong wheel and throw in extra rim and spokes, this is what I've got. Is it same rim as you guys get from them or is this one different?
Dont have the calipers to measure but I asked for phase wire upgrade and they look thicker than Mxus.

Yeah, looks like you have the 35mm as the wire doesn't come out the end of the axle like the 30mm does...unless they changed that now.

The thicker wires are a good start, but I still replaced them with 12awg silicone wire. I reused the Leaf phase wires in my 28mm '1000W' Golden motor and they are working great for it's power levels. 8)

Cheers