Brushless Efficiency-Do TF Hubs Really Work Differently?

[xyster]

Contributors to this thread from our friends at the TidalForce forum seem to claim that TF hubmotors operate differently and more efficiently than other brushless motors like the crystalyte.

Comments?

I'm not seeing how it works any differently, besides that TF's may use more poles, and so be a little smoother.

But if true, a used 750w-1000w TF motor might make a great alternative (I don't know how much they go for though...)

http://groups.google.com/group/Tidalforce/browse_thread/thread/938863b34d1e7078

I always wanted to know if the WC system uses all the coils at the same time
and if so, how that would affect efficiency.  I took a quick look just now
at the WC site and noticed this
http://www.wavecrestlabs.com/technology/controls.html  (not seen it before?)
and it clearly states each coil can fire when needed.  This is SOOOOO cool
and I guessed this was the case.
The point is that most dual type motors are simply tapped.  This is not the
most efficient way and the fact the WC system is NOT tapped and uses the
coils in various configurations to get you that efficiency is very exciting
(to me anyways).

R, I only notice perfectly smooth power from my TF whereas my
crystalight has noticable pulse/vibration called "coging effect" by a
C-lite dealer/engineer. I'm thinking your first post explains this.
The next test will be to test each motor on the same battery and
course to cpmpare effeciency.
I'm finally getting a replacement 504 hub from Justin which is in the
mail and will go on a new M750 frame and will be used for such
comparisons.
Al
M750X
Clite to Be

Interesting Al,
I'm assuming the 504 is a single speed? If so, then you would find the
results to be slightly misleading due to the power band.  
The thing is that you would need to test 36v and not just in a straight
(round course).  It would be interesting to hear a non stop speed comparison
but in traffic or any kind of stop start situation, would,  I assume give a
truer comparison.  But even then it gets confusing because the TF has a much
broader power band (Multi Speed) and I am not sure about the 504? (Or any
other kits to be honest as they are either single or dual tapped. TF can
switch on the fly and in milliseconds using SEVEN POLES... OMG!!!) so it
does get kind of confusing if you try and compare true efficiency to two
very different motors. And we have not even gone into available torque!
64lbs of Newton Meters @ 36v is something to be hold.  
The only real tests would be dynamo bench tests.  You could send variable
speed data to the dynamo and make out what is happening and also load up the
ride to give some realistic load tests to boot.  But your data will tell a
story none the less.
Cheers Al & look forward to your findings, I LOVE this stuff!

 

[lemmiwinks]

I don't know, but is the TF motor a variable reluctance motor?
http://www.freescale.com/webapp/sps/site/overview.jsp?nodeId=02nQXGrrlPb02R

I recall bocabikeguy saying something about it once, but I usually tuned out of the TF vs Crystalyte arguments on Power-Assist/E-motor-assist pretty quickly. Plus I take everything bocabikeguy says with a sizeable grain of salt. The now defunct lectra EMB http://www.electricmotorbike.org/index.php?page= used such a motor. It might have been a pro TF person claiming it was a VR motor and bocabike refuted the claim, I really don't remember.

 

[xyster]

How many poles do crystalyte hubmotors use?

I looked at this great animation linked from another thread:
http://www.servomag.com/flash/4-pole/smi-motor007.htm

How can you have 7 poles like some of the TF folks claim?
Is there another meaning to the term I'm missing?

Doesn't every south have to have a north, meaning only even numbers of poles? (excepting of course monopoles thought possibly to exist in the beginning of the universe )
http://en.wikipedia.org/wiki/Magnetic_monopole

 

[lemmiwinks]

How many poles do crystalyte hubmotors use?

Stuffed if I know, sorry. What I know about brushless motors you could chisel on the back of an aspro with a crowbar (old Aussie colloquialism meaning I know diddly). That goes doubly for Crystalyte motors.

 

[fechter]

The TF motor has magnets, so it is not a VR motor.

Wavecrest always hyped their "adaptive" something or other (I forget what they called it). My take on it was it was a form of dynamic timing advance.

Somewhere I remember somebody did a test of a TF vs. Crystalyte motor on an oval track and measured the amp-hours vs. distance etc.
From their data, it looked like a Xlyte 4 series was more efficient.

 

[xyster]

What I know about brushless motors you could chisel on the back of an aspro with a crowbar

That's great
What's an "aspro"?

 

[lemmiwinks]

That's great
What's an "aspro"?

Aussie slang for an Asprin. We like to abbreviate stuff 8)

 

[Lessss]

What didn't you go to Uni?

 

[xyster]

What didn't you go to Uni?

I prefer two wheels

Yes, I went to uni and grad'd from Colli. There I learned about "syllables". "Asprin" has two syllables. "Aspro" has two syllables. "Aspro" is not an abbreviation! Though it does save writing an extra letter...
silly, lazy Aussies!

(I'm still bitter from the abuse I suffered at the hands of an emu while visiting Australia when I was little And I'm also still disappointed I never saw a wild Koala, though I looked and looked for hours where wild Koalas were said to live! You guys owe me! One day I'll be back!)

 

[lemmiwinks]

:lol: I've seen a Koala while walking the dogs, they're not that great. :wink:

 

[fechter]

There's one....

 

[lemmiwinks]

Here's a lame joke for ya'll:

Q: Why did the Koala sit in the fork of the tree?



















A: Because the knife was too sharp!

 

[patrick_mahoney]

I'd seen the same data that fetcher mentioned - in his post way up above before we started talking about koalas.

It took me a bit on Google but I finally found it:
http://groups.msn.com/TidalForce/featurerequests.msnw?action=get_message&mview=0&ID_Message=5196&LastModified=4675568376462446442&all_topics=1

And then reprinting Devin's picture/graph down below. The data looks like Fetcher mentioned... TF motors are not inherently more efficient - although the whole regen thing probably makes them move efficient over the long haul. But the TF motor is markedly less efficient. It takes ~15A to get 20mph at 36V on the 5304, it takes ~18A to do the same thing on the 750X at 36V. It's not entirely apples to apples - but it's pretty close.

 

[Toorbough ULL-Zeveigh]

Isn't it kinda pointless to have elaborate bar graphs comparing amps & amp-hours of systems running different voltages?

What insight does that give? More than anything it shows off Devins ingnorance of basic electricity.

 

[fechter]

That's the one I was talking about Patrick. Good work.

Look carefully at the data.
Just ignore the 48v numbers and compare the 36v numbers. It definitely looks like the Xlyte motor does a bit better.

Now, if I could modify the Xlyte controller for regen, it would really kick butt.

 

[Toorbough ULL-Zeveigh]

It's kinda like comparing a regulated power supply with an unregulated pulsing DC supply. An unregulated supply is more efficient because no power is lost to the regulator, for a narrow range of operating conditions anyway.

You have to give up something to get something. It costs some power to get a broader band of smooth power delivery from zero rpm.

 

[xyster]

Any idea what Al at the TF forum is talking about here?
Unless I'm missing something, which isn't exactly unlikely, seven poles shouldn't even be possible!

But even then it gets confusing because the TF has a much
broader power band (Multi Speed) and I am not sure about the 504? (Or any
other kits to be honest as they are either single or dual tapped. TF can
switch on the fly and in milliseconds using SEVEN POLES... OMG!!!)

 

[fechter]

I can't believe the TF motor actually uses seven phases.
The number of stator poles has to be a multiple of 3 if it's a 3 phase motor. 3 x 7 = 21. I could easily see 7 windings per phase.

You could make a BLDC motor with more than 3 phases, but I haven't seen too many.

Someplace there was a picture of the insides of a TF motor.

 

[patrick_mahoney]

A little more Googling and I found this (search terms "tidalforce stator").

http://www.greenspeed.us/wavecrest_electric_motor.htm

From the page:

Inside of the hub is a multi-phase, DC brushless motor, arranged so that the rotor surrounds and rotates around the center-mounted stator. The rotor is made up of a series of independently-controlled electromagnetic coils driven by a proprietary power electronics module. The housing is backed by an iron plate, a necessary part of any electric motor. The rotor turns on Timken bearings.

The magic is in the electronics and algorithms that control the coils. The system can be updated every 100 milliseconds, keeping track of the position of the rotor and optimizing when the coils need to be turned on or off. The algorithms used to control the coils are the reason that WaveCrest has been able to demonstrate unheard of torque densities. "

The first paragraph is a bit vague - it described any brushless motor that I can think of. But the second one which describes the "magic" seems to match Fetcher's analysis that it's got a variable timing advance system.

Still, if Devin's data is right, I just don't see the advantage to the TF motor. The whole system, I'll admit is cool - but the motor seems to have less torque and seems less efficient at 36V than the Crystalyte motor. The torque seems significantly lower (16 seconds to accelerate to 20mph vs. 12 seconds on the Crystalyte). The top speed is a bit higher than the Crystalyte at the same voltage, so one would expect a bit of a torque penalty but the two don't seem to match up.

It's hard to do a fair comparison of different motors directly unless you can gear them, or you you can rewind them to have matching Kv's but the motors are so close that you can extrapolate a bit from the graphs to try to even things out and even when you do that, the motor doesn't seem to come out ahead. If I had to choose between the two motors, I'd choose the Crystalyte as the better of the two.

 

[Lowell]

Isn't it kinda pointless to have elaborate bar graphs comparing amps & amp-hours of systems running different voltages?

What insight does that give? More than anything it shows off Devins ingnorance of basic electricity.

Some people only have 20 amp controllers, so it's not entirely pointless...