Cute 100 + KU90 controller problems

Piksliviksi

100 µW
Joined
Sep 30, 2012
Messages
7
Hi,

This is my first post here, so be gentle, please :D

Just installed Cute Q100 36V 250-350W rear driving motor
together with KU90 controller (from bmsbattery) and I would like to hear your thoughts
on couple of things I noticed:

1. Temperature of the motor: after riding 18km or so, the temperature of the motor was ~66 degrees Celsius (measured from the bottom
of the screw-hole that holds the disk brake). Outside temp was 35 degrees Celsius.
Is that normal? Seems bit high to me. (Outside temp was +10 today, by the way).
Endomondo shows: Distance 17.96 km, Duration 41m:10s, Avg Speed 26.2 km/h. I was going full speed half of the total time,
36-37 kmh.

2. I get occasionally weird behavior from either KU90 or the motor, sometimes when I press the thumb lever, I get
loud noise and no traction, I press again - then the motor starts to work normally.. Sometimes I have to press the lever
2-3 times, before it starts to work. The noise I hear resembles the noise that appears, when your bicycle tyre is
rubbing against the mudguard (fenders in us speak?). Also, I lose traction occasionally, while holding the thumb-lever down.
Then I have to release the lever and press again and it starts to work. My thumb-lever has battery voltage
LED indicators as well, and these lights stay on all the time, while I loose traction.

Also, the on-off switch - when I press the red button on the throttle, nothing seems to happen, except
for that the battery voltage indicator lights switch off. I can still operate the thumb lever and the motor starts to pull.. I find it weird. Shouldn't the
switch turn the controller off? I didn't get any instructions with the controller, and everything that I could do
I have already tried, re-connected the cables, etc.

3. I would like to get a better controller - KBS48051E,30A,24-48V from Kelly maybe? What else would I need to get the regen braking to work?
Would the Controller work with my battery/motor (I have to limit the amps I think to 20-21)?
The battery I have is 36V 12Ah LiFePO4, Headway cells, also from bmsbattery.
 
Somehow you have gotten way off track.
I don't see a listing for a KU90 controller currently at the BMS Battery site, so I don't know what it is, but I doubt that it's the right controller for the Q100
The best controller for the Cute Q100 is the one that comes in the kit, the KU63.
Spending more money on a Kelly controller than the motor costs would be a double waste of money, it would kill the motor ASAP.
Those of us here using the Cute, like the KU63 with an easy shunt mod to around 17A @ 48V. At 36V, the motor could be happy with a couple more Amps, hard to say because we never run mini's on 36V.

The Q100 kit from BMS Battery is very well engineered for such an inexpensive kit, the connectors make it impossible to connect the axillary components incorrectly.
You need to get back to the stock kit and run 48V.
There is no point in trying to make the Q100 something that it's not.

P.S. the Q100 motor is a "free wheeling" motor, it has no regen capacity.
 
If you actually are using a KU90 "universal" controller that looks like this;

Universal_350W.jpg

then you are indeed using the wrong controller for a Cute 100.

The Cute 100 spins faster (internally) than most geared motors and needs a "high speed" controller like the KU63 or KU93 models from BMSBattery.

-R
 
I'm running it with hall sensors, I could try sensorless
as well just to see, how it behaves..

Actually the KU90 I have came originally with the 36V 350W 201rpm Bafang front driving kit
and I just replaced the motor. Front driving Bafang was too heavy for my old soviet
-era road-bike steel fork. The Q100 I got now, is also a 36V and 250-350W motor,
that "250-350W" on Bmsbattery's site made me think, that I could use the same controller
for both, 350W Bafang and Q100.. my mistake

I'd like to get a programmable controller actually,
maybe the 6x4110 Sensorless Mini Monster Controller LYEN Edition?
Or the 20 amp Kelly controller: http://kellycontroller.com/kbs48051l20a24-48v-mini-brushless-dc-controller-p-1169.html.. ?

What's the difference between KU63 and KU65?

The battery I have currently is a 36V 12Ah battery pack,
I was actually thinking about modifying it to 48V, by adding
cells and replacing the BMS..

PS. Here's the bike, Soviet Start-Shossee after the first
trial run (wires just tucked away to an old gas-mask bag on the side):
https://dl.dropbox.com/u/19941815/ss_elekter.jpg
 
Do not continue with that controller. The motor cannot handle that amount of current (22A). The last guy that used it only got about 5 miles before something burned. The correct controller is the KU63 that is set to 15 amps max. The motor can handle about 17/18amps max. To get that, you can either get a 6 FET programmable controller, which is relatively expensive, or solder half the shunt in the KU63. If you up to that level, don't run at full throttle everywhere without pedalling.

KU65 controller has additional panel with switch and LEDs. The LEDs are bright so not very stealthy, although the switch is useful.

Edit: I just looked at your photo. That's not a Q100 motor - much too big. Has it got a label on it? If it's a Q128, hard luck! If it's a BPM you have something to work with, Now everything that's been said in previous posts is irrelavent. Your juddering was probably due to those bullet/barrel phase connections, which break down under load if they're too loose. Most serious builders change them, but they'll work up to a point if you squeeze the barrels before connecting so that they're really tight.
 
How the switch on the throttle works depends how you wire it. There's a thin red wire from the controller next to the thick red battery onethat you must've joined to the thick one somewhere so that it gets 36v and enables the controller otherwise the controller wouldn't work. Instead, you need to run it through the switch and then to the 36v. Connect the LEDs wire to the controller side of the switch as well, and then the LEDs and controller will be activated by the switch so that the LEDs come on when it's active.

Forgot to say: You made a nice job of your project. Should be good when you get it sorted. A second look says it's probably a BPM motor, which is really nice.
 
d8veh said:
.......... To get that, you can either get a 6 FET programmable controller, which is relatively expensive, ....................

??? Not sure about this, as the fully programmable 6 FET controllers are dirt cheap ($22 plus shipping) and are easy to programme, or mod to get pretty much whatever you want, up to around 80 plus V and maybe 30 to 40 A.
 
d8veh said:
That's not a Q100 motor - much too big.

Thats why I have the KU90 - its the original motor that I used,
now that 201rpm Bafang is gone and replaced with the Q100.. the picture is
made when I installed the first motor.

Heres how it looks today: https://www.dropbox.com/s/e71s86w5tod13qk/2012-10-01%2013.44.07.jpg

If I get this controller-confusion sorted out, I will hide the wires,
install some-sort of basket to my rear rack and cover the top and the sides
of the battery with some black thin plastic, so, that when I have something in the
basket, the battery wouldnt show.

Controller will be installed on the fender-side of the rack, below the battery.
Im looking for a stealthy black box, where to put the controller in (should make it more
resistant to humidity though) and I want to install Speedict also to this box along
with the controller.

The problem with the KU63 is that allthough its dirt-cheap @bmsbattery, shipping to
Estonia is about 40USD, which makes it a bit problematic. A bit more money and I can get the
6 FET 3077 Mark II Mini Monster LYEN Edition Controller ($89)..
 
Just be sure any controller you buy is compatible with a high-speed motor like the Q100. The Q100 with a 9:1 reduction spins faster than most other geared motors with 5:1 reduction therefore you need a controller that can handle high speed commutation.

-R
 
Russell said:
Just be sure any controller you buy is compatible with a high-speed motor like the Q100. The Q100 with a 9:1 reduction spins faster than most other geared motors with 5:1 reduction therefore you need a controller that can handle high speed commutation.

-R

Good point, except that pretty much every controller currently available is now capable of commutating at rates far greater than this motor needs. It's true that years ago one or two of the older, pre-MCU, controllers (the analogue Crystalyte springs to mind) weren't capable of driving one or two high speed motors, but pretty much all the controllers I've seen in the last couple of years will handle at least 60 K erpm. IIRC this is a 10 pole pair motor (20 magnets), so even in a small (20") wheel pretty much any controller available will drive it to around 650 rpm or more, given a high enough supply voltage. 650 rpm in a 20" wheel is roughly 38 mph, faster than this motor could ever drive an ordinary bike in terms of power needed.
 
I guess it could be the 700cc wheels, but that looks more like a Cute 85 than a Q100.
I have both a Lyen mini-Monster and a KU63 and there is no advantage to running the Lyen on a Q100!
IT WILL NOT MAKE ANYMORE TORQUE!
HEED THE 17 AMP LIMIT OR YOU WILL DESTROY THE MOTOR!
Buy other things that you will use in the future from BMS Battery to make you feel better about the shipping or buy a controller from Ecrazyman or any of the many vendors on Alibaba!
 
The advantage of using a programmable controller is that you can set the various limits to suit your needs. The Wuxi controllers (a.k.a. Greentime, KU, Hautong etc) are programmable, but unfortunately we don't have the means to use this function, as it's not been released openly. The Xiechang controllers (a.k.a. e-crazyman, Kewin, Lyen etc) have the advantage that the means to programme them is no open and easily accessible.

Amongst the other useful advantages, like being able to set the speed ranges for switched speed, the cruise control settings and the low voltage cut off setting, programmability gives is the ability to vary the phase current to battery current ratio, which is surprisingly useful in getting good throttle response, preventing the motor from drawing too much current at low speed, yet allowing good high speed performance. The crude alternative of soldering the shunt is both risky, in terms of the accuracy being very poor, and also because it desensitises the peak over-current fault protection system in the controller, making it more vulnerable to failure under fault conditions.
 
If you read back though these posts you will see that the OP is fixated on pumping 20 to 30 Amps though his little Cute and a bunch of techincal dilly-dallying isn't going to help him a wit when he burns his motor out.
The Cute 85/Q100 has well defined limits and the parameters of the KU63 suits it very well. The cruise functions beautifuly, there is no need for "speed ranges" and presumably, since his pack had a BMS, it would also have a LVC. At any rate, the Ecrazyman basic controllers are available with various LVC's.
As for the shunt mod, a light coating of solder will put the controller @ 17A plus or minus .5A, everytime. couldn't be easier.
Trying to coax more performance out of the Q100 basic kit is a Fool's Errand.
His money would be, far and away, better spent on another Q100 kit, rear mounted for 2WD.
Aside from D8veh's nicely done conversion, there are some very nice dual Q100 conversions on the German Pedalec forum.
Sometimes, for some folks, KISS is best.
 
motomech said:
I guess it could be the 700cc wheels, but that looks more like a Cute 85 than a Q100.
I have both a Lyen mini-Monster and a KU63 and there is no advantage to running the Lyen on a Q100!
IT WILL NOT MAKE ANYMORE TORQUE!
HEED THE 17 AMP LIMIT OR YOU WILL DESTROY THE MOTOR!
Buy other things that you will use in the future from BMS Battery to make you feel better about the shipping or buy a controller from Ecrazyman or any of the many vendors on Alibaba!

Yes, 700cc wheels, and this is what Bmsbattery sent me, should be the Q100. Had only two stickers on it - 36V and the QC-sign.
But I'm not looking for more torque, I just would program the controller to get the limits I need
for this motor..
 
I think rather than a programable(expensive) controller. you might be better served by the standard controller and a Cycle Analyst.

Here is an impecible Dutch conversion using a Q100 on a road bike. He switches from the KU63 to a programable controller, but in the end, He finds the features of the Cycle Analyst more useable;
http://www.pedelecforum.de/forum/showthread.php?13010-Umbau-Hollandrad-mit-Q100-EB206-36V10Ah-CA

I myself, use a "fast wind" Q100 on the rear with a 24" wheel as a assist motor and it works wonderfully. Just engage the cruise on the 16 -17A shunt modified KU63 and pedal. smooth as silk with no noise what so ever. A "slow wind" Q100 in a 700cc wheel would be about the same thing.
If you are not looking for more power out of the Q100, what are you looking for?
You are over-complicating this.
 
motomech said:
If you read back though these posts you will see that the OP is fixated on pumping 20 to 30 Amps though his little Cute and a bunch of techincal dilly-dallying isn't going to help him a wit when he burns his motor out.
The Cute 85/Q100 has well defined limits and the parameters of the KU63 suits it very well. The cruise functions beautifuly, there is no need for "speed ranges" and presumably, since his pack had a BMS, it would also have a LVC. At any rate, the Ecrazyman basic controllers are available with various LVC's.
As for the shunt mod, a light coating of solder will put the controller @ 17A plus or minus .5A, everytime. couldn't be easier.
Trying to coax more performance out of the Q100 basic kit is a Fool's Errand.
His money would be, far and away, better spent on another Q100 kit, rear mounted for 2WD.
Aside from D8veh's nicely done conversion, there are some very nice dual Q100 conversions on the German Pedalec forum.
Sometimes, for some folks, KISS is best.
Here is an impecible Dutch conversion using a Q100 on a road bike;
http://www.pedelecforum.de/forum/showthread.php?13010-Umbau-Hollandrad-mit-Q100-EB206-36V10Ah-CA

If that's directed at me then you've got your wires well and truly crossed.

I've not once hinted, suggested, or otherwise implied anything about increasing the current, all I've said is that programmability allows a lot of fine tuning to be done to get the best from a particular motor/bike/battery/usage pattern.

Your comments about the shunt mod are misleading, and almost certainly based on a small sample. The stock shunt starts off with a 5% tolerance, which is made worse by the way it's not always seated properly on the board/has different amounts of solder flowing up the pins. There will be differences in the way people add solder, adding further variability, and worst of all, increasing the current limit by reducing the value of the shunt also reduces effectiveness of the peak current overload circuit, making the controller more vulnerable to failure under a fault condition. If you don't believe me, then just take a look at the circuit diagram for these controllers - the problem becomes clear when you see how they operate internally.

It's also a fact, beyond dispute, that adjusting things like the phase current to battery current limit ratio via programming can give beneficial results, particularly when the motor may well have a fairly modest peak current handling capability, as has been reported for this particular motor. This would indeed allow better performance to be obtained overall (defining performance as the mix of endurance, reliability, speed, hill climbing ability, throttle responsiveness etc).
 
motomech said:
I think rather than a programable[expensive) controller. you might be better served by the standard controller and a Cycle Analyst.

Since when has $22 been expensive? That's the cost for a programmable controller that's equivalent, in performance terms, to the small fixed output Wuxi controller sold as the KU series by some vendors.
 
Jeremy Harris said:
motomech said:
I think rather than a programable[expensive) controller. you might be better served by the standard controller and a Cycle Analyst.

Since when has $22 been expensive? That's the cost for a programmable controller that's equivalent, in performance terms, to the small fixed output Wuxi controller sold as the KU series by some vendors.

You are the one talking about a $22 programable controller.
The OP has refered only to the Lyen(actually w/ the cable and custom mods, well over $100 plus shipping) and a Kelly.
Yeah, I would say that's expensive.
 
motomech said:
Jeremy Harris said:
motomech said:
I think rather than a programable[expensive) controller. you might be better served by the standard controller and a Cycle Analyst.

Since when has $22 been expensive? That's the cost for a programmable controller that's equivalent, in performance terms, to the small fixed output Wuxi controller sold as the KU series by some vendors.

You are the one talking about a $22 programable controller.
The OP has refered only to the Lyen(actually w/ the cable and custom mods, well over $100 plus shipping) and a Kelly.
Yeah, I would say that's expensive.

I agree, either the Lyen or the Kelly would be massive overkill for this application, but that's not to say that a bog standard, off-the-shelf, programmable 6 FET controller at $22 wouldn't be, which is all I've suggested. When you can get a fully programmable controller that's potentially more capable than the Wuxi fixed controllers (like the ones labelled KU65), for exactly the same price, why wouldn't you opt for the extra flexibility that programming gives?
 
I agree, either the Lyen or the Kelly would be massive overkill for this application, but that's not to say that a bog standard, off-the-shelf, programmable 6 FET controller at $22 wouldn't be, which is all I've suggested. When you can get a fully programmable controller that's potentially more capable than the Wuxi fixed controllers (like the ones labelled KU65), for exactly the same price, why wouldn't you opt for the extra flexibility that programming gives?
When you can get a fully programmable controller that's potentially more capable than the Wuxi fixed controllers (like the ones labelled KU65), for exactly the same price, why wouldn't you opt for the extra flexibility that programming gives?
Well, my concern would be the same if I bought a programable controller, That I would make things worse by messing with it, and I already know that one can't get regen from a geared motor. I guess both the OP and myself could learn to use it and not shoot ourselves in the foot, but I am still sceptical of any real benefits. By virtue of it's small size and low oputput, the Q100 already is dead smooth at takeoff and as quite as any geared motor can be(under 17.5 A).
Of the many observations I have read of the Q100[and all mini's, to a certain extent], the only prevailing theme is a wish that it had a little more power. Will any aspect of programing, aside from Max. Amps, do this? Especiallin light of the Q100's exteemly sm. gauge wires?
I don't claim to have a vast reservoir of experience with hub motors, but I have spent some time with the Q100 and I think I have a pretty good handle on how far it can be pushed.
Up to constant 750W, it is quite and happy.
Above, it starts to protest in the form of a whinny growl, but it will live if handled carefully.
Above a contant 850W, all bets are off.
 
What kind of cheaper programmable options there are besides Kelly and Lyen?
And from where to get it? I just don't want to shoot myself into foot with this,
buying from X chinese vendor and ending up with a rather cryptic UI
in chinese.. :cry: I'd rather spend some more money and get the controller
from known place.

The reason behind the idea to get the programmable controller is that
I'd like to get some experience on how the different parameters
affect the performance of the motor, using Speedict as a monitor.
I could limit current with the Speedict, it'll show exactly what you've got in real-time,
but still, I would use Speedict as a "fuse", if I mess up with something, I have a back-up
current limiter between the motor and the controller.. 8)
I'm doing research right now and I want to build an electric moped during the
winter months. So, I thought about getting a 20-30 amp 36-48V Kelly, that I could use with
any other 36/48V 250~500W motor in the future as well when testing.. even the
36V 350W 201rpm Bafang I have. Bad idea?

I will get the stock 6-mosfet KU63 (or similar, whatever brand) as well, for everyday use I think
for Q100. I found few options from ebay already.
 
You can buy exactly the same base controller as the Lyen 6 FET from Keywin Ge (e-crazyman on ebay). He's a very reliable and honest seller; I've been buying from him for several years now and currently have around 7 or 8 of his controllers fitted to various bikes, boats etc, with not a single failure in around 3 or more years of use. Lyen gets his controllers from him, then modifies them with better FETs and customised wiring, hence the higher price.

You can get a 6 FET controller from e-crazyman here: http://www.ebay.com/itm/36V-350W-brushless-controller-for-E-bike-scooter-/260947486400?pt=LH_DefaultDomain_0&hash=item3cc1aea6c0

This controller is, in reality, a Xiechang controller (just as the controllers labelled as KU65 by BMS Battery are in reality Wuxi controllers). There is a great mass of information on this forum on programming these controllers and fine tuning their settings. Programming involves connecting the controller to a PC via a serial cable (or USB/serial adapter) and using one of the freely available software packages on this forum to change controller settings.
 
Jeremy Harris said:
Russell said:
Just be sure any controller you buy is compatible with a high-speed motor like the Q100. The Q100 with a 9:1 reduction spins faster than most other geared motors with 5:1 reduction therefore you need a controller that can handle high speed commutation.

-R

Good point, except that pretty much every controller currently available is now capable of commutating at rates far greater than this motor needs. It's true that years ago one or two of the older, pre-MCU, controllers (the analogue Crystalyte springs to mind) weren't capable of driving one or two high speed motors, but pretty much all the controllers I've seen in the last couple of years will handle at least 60 K erpm. IIRC this is a 10 pole pair motor (20 magnets), so even in a small (20") wheel pretty much any controller available will drive it to around 650 rpm or more, given a high enough supply voltage. 650 rpm in a 20" wheel is roughly 38 mph, faster than this motor could ever drive an ordinary bike in terms of power needed.

Well maybe...

I tried 3 controllers that worked perfectly on my Bafang geared motor (an Ecrazyman 6-FET, A KU90, and another one I got with my first Bafang) on a MAC motor and none of them worked properly due to the high commutation frequency of the MAC. I bought the "high-speed" KU93 and it worked great on the MAC.

Now if the OP has a 201 RPM Q100 and runs it on 36V I would agree that most any controller should work, and for that matter the KU90 he is presently using should work . I just bought a Q100 from an E-S member but haven't done anything with it yet or I would know for sure.

which reminds me...note to Piksliviksi

The KU90 has a "learning" function where it will go through the possible phase/hall combinations. First put the bike on a stand since the motor will run on its own during the procedure. Connect the two white wires together then power up the controller. The unit will sequence through the possible combinations which will cause the motor to jerk or even run backwards inside the case. When you see the motor running smoothly forward quickly disconnect the white whites. Cycle power on the controller and it will remember the last state and should run well.

If the programming works the motor will run smoothly and will have a no-load current around 1A on the stand (depends on speed). If the motor runs well at partial throttle but then starts drawing excessive current as it reaches higher speeds then it likely means the commutation frequency is too high for the controller and you'll need a new one.

Link to KU90 connection diagram:
http://www.google.com/url?sa=t&rct=...64CgCg&usg=AFQjCNF1u6UyDG0DrTIH8SFIzX9PhDojhA

-R

PS be sure to check all of your phase and hall connections for pushed out pins, bad crimps, etc
 
Back
Top