New Crystalyte Motor series (HT35 / HS35 and HT24 / HS24)

[Spacey]

I bought a sensorless version But do I benefit any extra power speed from that extra 4mm?

And here is the $1,000 question.....will I be able to add halls to the wider sensorless version?

 

[Andje]

if anything that is what i would be upset about; in justin's pictures of his sensorless version it "looks" to me like i could fit halls on there, under the laminations. It wouldn't be easy and I would bow to his judgement, but I just replaced mine on my 9c and had no problem tucking them in flush; the lamination diameter fluctuation was more then the halls added certainly.
And we will definitely loose some power. Actually I have no idea if it's a linear relationship, so I again will bow to someone who can do the math (edit-look down, i trust lfp). Prbly have to wait for a dyno test for the final verdict though. And then a side by side race of a sensored and sensorless version at same V/A.

 

[methods]

And here is the $1,000 question.....will I be able to add halls to the wider sensorless version?

Yes - where there is a will there is a way.
Even if you have to machine the cover down another few thousandths it can be done.

-methods

 

[methods]

... in justin's pictures of his sensorless version it "looks" to me like i could fit halls on there, under the laminations.

Right - but what you or I could conceivably do once does not necessarily equate to a profitable manufacturing process for Kenny. Having "skilled workers" perform a "tight operation" thousands of times can result in a reliability nightmare.

It may just be that Kenny wants to make the sensorless more attractive.... who knows. If he gets enough pressure from enough directions I am sure we will eventually see senors on the 35mm version. I told him that I would forward pictures when we figured out a good way to do it.

I am holding out for the long shot: He used magic metal (which can only be extracted from a pandas ass on leap year) for the 31mm stators In my fantasy world these would perform about 3X better than the 35mm

-methods

 

[Andje]

haha, yeah i totally agree on the manufacturing point. I've only opened a couple hubs and the halls on both were terribly installed.
I just meant it looked possible, so those of us on here in the know could do it.
Perhaps it is, as you, impossibletoobtainium. If so i will sell it in grams and live off the profits for years!

 

[chroot]

ugh 32mm? Is that mean I ordered sensored HS3540 isn't exact 35mm. Look like I wasted money on this HS3540 for nothing. Am I misinformed? Not nice

Need test dyno compare 9C 9x7 vs HS3540 (32mm with sensored)

 

[methods]

Chroot always has a way with words.....

-methods

 

[liveforphysics]

ugh 32mm? Is that mean I ordered sensored HS3540 isn't exact 35mm. Look like I wasted money on this HS3540 for nothing. Am I misinformed? Not nice

Need test dyno compare 9C 9x7 vs HS3540 (32mm with sensored)

This motor, 32mm or 35mm is still going to be an animal.

Justin's numbers weren't finding saturation (no knee), so those torque values weren't the torque limit of the 35mm stator.

It's entirely possible the 32mm is capable of an identical dyno curve (depending on how it's wound and powered etc). And the 32mm should have slightly reduced parasitic freewheeling torque than the 35mm.

The outside of the case is also physically identical, meaning for a given temperature of the case, it's going to cool identically.


The 35mm stator should have roughly ~12% higher torque saturation point, but if you weren't planning on running at it's saturation point (which nobody but the most extreme LiPo + super hacked controller guys can do), then it's really not going to make a lick of difference in practice.

 

[neptronix]

LFP; theoretically, how would you compare this motor to a 5xxx series crystalyte?

I have been going back and forth for weeks now wondering if i should buy one of these. I figured drastically lighter = less capable than the 5xxx series..

 

[Andje]

thanks lfp, i will edit my above comment, don't want to spook anyone with my ignorance. It is good to hear that it won't make as much difference as I assumed.

 

[liveforphysics]

LFP; theoretically, how would you compare this motor to a 5xxx series crystalyte?

I have been going back and forth for weeks now wondering if i should buy one of these. I figured drastically lighter = less capable than the 5xxx series..

Mass has no bearing on the rate you can dissipate thermal energy. Less weight just effects the rate you reach thermal equilibrium, but zero effect on continuous energy dissipation (or continuous power in this case).

The 53xx series has a boat load of weight in it that is not material actively contributing towards the motor function. This motor has a design with less dead-weight in it.

As far as an ultimate comparison, it will require somebody with more advanced skills. Member Biff could do it (but he is a very busy guy right now).

 

[chroot]

Maybe you may not understand the deaf culture. Like I am not great in english as primary in america sign language.

Chroot always has a way with words.....

-methods

 

[chroot]

@liveforphysics - thanks for explain and I understand little difference.

 

[justin_le]

And we will definitely loose some power. Actually I have no idea if it's a linear relationship, so I again will bow to someone who can do the math (edit-look down, i trust lfp). Prbly have to wait for a dyno test for the final verdict though. And then a side by side race of a sensored and sensorless version at same V/A.

There's no need to wait for a dyno test to have a really good idea of what's going on. The motor KV/winding is directly proportional to the flux linkage. To a first order, we would indeed assume that the 30mm stator would have 30/35 = 86% of the flux linkage of a 35mm stack, and hence 86% of the power and torque capability.

Under that assumption, for the same winding, the 30mm should spin 16% FASTER than the 35mm stator. Now, the fact that Crystalyte supplied the HS sensored motors with a 6-turn rather than a 5turn wind implies that this was probably the case:
30/35 * 6/5 = 1.03, so basically by using a 6 rather than 5 turn stator Crystalyte can get within 3% of the V/rpm of the 35mm stack and pass it off as the same 40kph motor spec.

If the 30mm lamination is really super high-saturation-high-permeability-magic-panda-metal-alloy and behaved the same or better as the 35mm pot metal sensorless crap stator, then your 10x6 motors should be spinning at least 17% slower than the 12x5's in the sensorless wind.

So Luke, or Methods, or somebody, give us as exact as you can a spec for your motor KV. Take the unloaded RPM at a known volts, subtract off the I*R loss from the windings knowing the winding resistance and no-load current, then divide.

The 35mm 12x5 has a KV of 0.92 Rad/sec/volt

-Justin

 

[justin_le]

LFP; theoretically, how would you compare this motor to a 5xxx series crystalyte?

I have been going back and forth for weeks now wondering if i should buy one of these. I figured drastically lighter = less capable than the 5xxx series..

The 53xx series has a boat load of weight in it that is not material actively contributing towards the motor function. This motor has a design with less dead-weight in it.

As far as an ultimate comparison, it will require somebody with more advanced skills. Member Biff could do it (but he is a very busy guy right now).

Still way early for even beta release but we've been working on doing a thermal model of the hubs to have a heating characteristic time show up in the simulator. So if you go to the current development site (please don't comment on layout, obviously a work in progress right now and may be broken at random times):
http://www.ebikes.ca/simulator3

There is a text field showing "overheat in" which is an estimate on how long it will take for the motor stator to get to 150 degrees C at the particular cursor location. Right now it's based on heating tests that were done with the motor static and not spinning, so it's pretty conservative, but does give a good comparative sense of how the different motors stack up in their ability to 'take it'.

Luke is correct that the X5 carries a lot of extra metal weight that isn't core to the function as a motor. But this extra mass DOES mean that it can handle an overvolting dump of energy for a longer period before overheating. In principle you could get exactly the same effect with one of the lighter 9C stlye motors by bolting some huge chunks of aluminum heat sink to the stator to increase the motor's internal heat capacity. It would also make it heavier again, like the X5....

-Justin

 

[chroot]

@Justin - Really appreciated your explanatory really very crystal clear and simulator3 is very nice!

 

[Andje]

thanks justin, that is the math understanding i lacked. very much appreciated.

 

[liveforphysics]

4 pieces of 14awg fit in the wound slot. This was inspiration to do a re-wind.

Grinder makes fast work of end-turns.

8 pieces of 14awg technically does fit... however, winding 7 in there was massively difficult on it's own, so I'm going for 7 turns.

Slot dimensions for the motor tweakers out there.

Magnets are 35mm long, and this wide:
http://img703.imageshack.us/img703/2976/imag0990.jpg

 

[liveforphysics]

Factory my motor had 19.6mm^2 of copper fill cross-section in each slot. (10 pieces of 25awg wire each wrapped 6 times around each tooth, 12 sections per slot.)

When I finish with the 7-turns of 14awg, it will have 29.1mm^2 of copper fill cross section in each slot.


Roughly a 50% improvement in copper fill, and much less end-turn loss. Should do very good things.

 

[Hyena]

4 pieces of 14awg fit in the wound slot. This was inspiration to do a re-wind.

That's badass man 8)
Did you even test the stock performance ? :lol:

How are you planning to run it ? 100v 100a + I imagine ?
I look forward to seeing the videos of it in action.
Actually, if you've got time to throw the camera on a tripod or something I'd be interested to see a clip of just how long it takes to do a small section of that rewiring. I'm guessing that wire is pretty stiff and rewinding will be a bitch of a job but it should be a destroyer of worlds when finished

 

[dbaker]

Amazing, Luke Hub motors taken to another level Is this going to run in the Death Race What are your expectations for change in motor impedance that the controller will see

 

[keyne]

Luke, could you draw a quick sketch or post a link to a good web page on how to rewire one of these motors?

 

[esoria]

Very nice I'd love to rewind a motor, but
I do not know what to do

 

[Andje]

If it only took say 20-40 hours for someone who has never done it before, and the performance was substantially improved, i'd prbly attempt it at some point. It's a scary task mentally, but i bet it's more just finicky and hard in practice with good understanding and directions. Then again, might be a waste of time for me; bet i will be happy with the performance of this hub as i already am with my 9c.

 

[methods]

Very nice I'd love to rewind a motor, but
I do not know what to do

Neither does Luke

But... that is what makes him a bad-ass. The willingness to jump in with both feet, every time, no matter how scary or expensive. The resources are here - if you want to do it you can. :wink:

-methods