GL-2 Hub Motor Kit
Has Higher Torque Characteristics than GL-1
This is a gearless brushless motor
The GL-2 I was fortunate to have the opportunity to test on a whizz bang
very expensive testing machine at an ebike factory in China in late March '06.
The following results were obtained. (a little bit of info at top of this page
about testing machines http://www.users.bigpond.com/solarbbq/torquegraphs.htm),
so I can vouch for these results as I got them myself (with a bit of help from
the technician I must admit, well lets be honest, I helped him!)
Unloaded:
0.912amp
(@321rpm)
Maximum Torque: 48.42Nm
(@ 18.87amp)
Maximum Efficiency: 84.89%
(@281rpm@12.38Nm@8.942amp@364.3watt)
Maximum Power Out: 641.8watts
(@74.86%eff@214rpm@28.64Nm@17.8amp)
Rated
Power:
350watts
(I have no idea how rated power is calculated by the machine)
Pictured below is the motor I tested, diameter 26.33cm weight = 6.5kg:
The motor as it comes is designed to run on 48volts and the controller can
draw up to about 18amps.
I first received some of these motors (12th of June 06) but I am yet to test
the motors. So I wont be able to sell these kits until about November 06 after I test in a bike.
I'm also hoping to test this motor with the
crystalyte 72volt controller and see if it can be overvolted at all ( i.e. will
there be excessive heat produced when the motor is under load).
Below is the torque graph that I helped make in March 06 on a whiz bang test
machine in an ebike factory..
Rated Speed is 320rpm.
From the graph it can be seen:
Maximum Efficiency: 85% (very high indeed) (makes this the highest
efficiency motor I have on the website)
Maxium Torque : 48.42 Nm at 19amps (very high is making this motor almost
comparable with
the x5 crystalyte motor)
Maximum Output in Watts: 650watts (quite a high value would really like
to try this motor on 72volts as time permits and see what it can do!!)
{postscript: I now realise I wasn't told which rim size the graph above
torque figures are for, I would have to guess 16" or 18" as its a
common size used by the company who owns the testing machine, graph below may
help though)
An alternate Torque Graph for the GL-2 motor, torque graph supplied by the
company who makes the motor.
There are some differences between the two graphs noteably the maximum output
power is lower in the graph supplied by the company who makes the motor
(525watts compared with 650watts). Also the stall torque is lower on the second
graph (the one produced by the company who makes the motor) . From the top graph
the stall torque would be around 60Nm.
The torque figures in the graph above are for a 24" rim size ( 48volts was used in the test
above)(I will have to check what rim size the first torque graph was calibrated
for to get the torque values, will dig out the graph soon)
So for a 20" rim: 24/20* 44 = 53Nm max. torque
for a 26" rim size: 24/26*44 = 41Nm max. torque
This puts this motor in terms of torque above the 408 cystalyte but below the
X5304 crystalyte motor (even when the X5 is run on 36volts!!).
Expecting to have time to test this motor by the end of November.
Example of the GL-2 spoked into a 24" rim.
Kit will come with regenerative braking type controller, throttle, brake
handles and spokes to suit your rim. Will also include a 3 speed cluster on these motors you will need to spoke the motor into
your own rim. Also a key switch and fuse (20amp)
Inside the motor:
There are 4 parts to the motor and no moving parts excepting the bearings.
Benefits being no maintanence (unless a hall sensor fails)
Face plates are made from an alloy to keep the weight down. Bearings would be
easily replaceable if ever needed.
Both face plates are identical with identical threads. One thread is used for
the gear cluster (a 3 speed cluster is supplied on the motor)
the other side is for a drum or disc type brake. A rubber seal keep stops water
getting into the motor via the axle.
Bearings seem to be good quality. Can be removed by hitting the face plate very
firmly on the ground a few times (the chinese method!( no puller required).
Since the bearings are placed well inside the face plates it means its possible
to remove the threads and get a very narrow motor which can be used as a front
motor application. Minimum width motor will fit is 8cm if threads are removed.
But very strong front forks would be required to avoid any fatigue of front
forks from the torque of the motor.
The axle flat stops just before the motor thread (on face plates), if the motor
was to be used as a front motor the flat on the axle would have to be extended
(can be done with a hacksaw and a grinder and a bit of skill.....see gl-1 page
for instructions)
The windings of the motor are somewhat different to most non-geared
brushless motors. The are very similar to the fisher-paykel washing machine
motor structure ( which I did quite alot of testing on some years ago to use on
a trike, with some success). It means the motor has alot of poles and quite a
large diameter but is quite narrow. This motor is the latest design motor
beginning to be used in ebikes (moped style) in china as they give higher torque
and slightly higher efficiencies than standard motor structures.
What would a chinese motor be without some bamboo!!! Inbetween the iron fingers
(not sure of their proper name) they put strips of bamboo which wedge the
windings in place so they cant move.
The iron core focuses the magnetic field as is the case in most electric motors.
Hall sensors are located as shown.
Hall sensors shown below. They are glued in position as with all the brushless
hub motors I've come across. Picture below right shows the 3 main power wires
(its a 3 phase motor thanks again mr tesla), the three main wires which come out
of the motor divide into a number of copper wires (5 or 6 sorry cant remember
exactly), which then are wound around the iron core. All motor winding is done
by hand as is very common in china.
There are 46 neodymium magnets glued to the iron rotor (the iron helps focus the
magnetic field).
Installation Instructions coming soon.
{Off-topic: way off topic! So this motor is really a step up in
design from the commonly available brushless gearless hub motors on the market.
It gives a small increase in efficiency and a small increase in torque over the
motors that are presently being used. This motor is now being put in the latest
moped style ebikes being produced in china. So its really a state of the
art motor as far as electric bikes are concerned.
The design though has been around for some time a good example being the fisher-paykel
washing machines which have been using this style of motor for maybe 10 years or
so (which they patented). They needed a high torque low rpm motor in which speed
and direction could be controlled electronically.
Just a bit of a guess at what new motor designs will emerge, I think its
very likely that the axial flux motors being used in solar race cars ( as
designed by the CSIRO in Australia and the Northern Territory University) may be
the next design to emerge for use in ebikes due to their high efficiencies (98%)
and high torque to weight ratio. That will be quite a big step up if it occurs.
The Solar race car motors are about 32cm in diameter, the GL-2 is
23cm diameter (not a huge difference). The solar race car motors running on a
bit over 100volts can propel the solar cars at over 100km/hr ( I think up to
about 130km/hr or so). I don't think there is a huge difference in amount of
work involved to produce a GL-2 and a CSIRO style motor (I haven't of course
built one so I could be very wrong!! see this page for design structure of the
CSIRO motor http://www.users.bigpond.com/solarbbq/bikeexp/bexp.htm).
The windings are quite different in the CSIRO motor and possibly would be
simpler to produce than the windings in the GL-2. The magnet array of the
CSIRO motor though would involve quite a bit more work as there are more poles
and they use a halbach array ( the north south poles are aligned so that an iron
backing plate is not required and reduces the weight of the motor
substantially). But using the halbach array means more work
to produce the motor, but I would guess its quite achievable considering the
effort that goes into producing a GL-2 motor.
There is one brand of electric bike using a motor similar (I dont think
it uses the halbach array though) to the CSIRO motor but strangely enough it is
at present probably the worst electric bike on the market ( I wont say the name
in case they read this!! ) The reason I think it is not a good electric bike is
that the motor was a University based project, alot of successful electric bike
and hub motor companies (in china) seem to do alot better when the company is
founded by a motivated individual where commercial success is more imperative.
I'm really quite amazed at what some people have achieved over there in China
off their own backs. A University based project will operate very differently to
a company set up by an individual. The other factor that has contributed to the
poor performance of the bike I'm speaking of is that they have limited the motor
output to around 200watts to comply with australian road rules. This amount of
power is rediculously low for any practical electric bike.
Another example of a University based project that had(has) huge potential but
went (and continues ) to go horribly wrong is the Tongxin company in china, an
interesting an example of the difference between uni based projects and more
commercially based manufacturers.
And while I'm off topic!! There are plans afoot in which ebike companies want to
enter into the electric car market in the near future, a CSIRO style motor at
low cost would certainly make that very viable, I don't have any doubt they will
achieve it!}
beware of spray ebike company: more info
back
to home page
Copyright 1998-2006 by Brett White on all images/text/information and
graphical materials on this website.