KU123 questions

[schwibsi]

Hello,

I'm about to mod a KU123 with different FETs (for 13s) and I'd like to do get rid of the wires I don't need.
Some wires were ripped out by accident so I don't know, what connector they go into.
For example the blue wire in "Q"

Also, the colors of the hall sensors and the color of the corresponding phase wires seems to jumbled.
The green phase wire is U, the U hall sensor is blue
The blue pahse wire is V, the V hall sensor is yellow
The yellow phase wire is W, the W hall sensor is green.

I'm going to use a Mac motor and would like to connect them properly.

Is there anything else I should be changing?

 

[schwibsi]

KU123
U (large) green phase wire
V (large) blue phase wire
W (large) yellow phase wire

K1 blue 3-speed switch
K2 black 3-speed switch
ZF unten green 3-speed switch
U yellow hall sensor
V green hall sensor
W blue hall sensor
S0 green throttle
+4.3V red throttle
XS white Speedlimiter
Q blue cruise control switching minus
ZL green PAS switching minus
SL yellow ???
SW blue ???
XZ red ??? switching +5V

A3 ??? ???
A4 ??? ???
E4 ??? ???
K3 ??? ???
K0 ??? ???
X ??? ???
W1 ??? ???
DS ??? ???
A2 ??? ???
YBS ??? ???
E5 ??? ???
E2 ??? ???
E3 ??? ???
B- ??? ???
B+ ??? ???
S+ ??? ???
S- ??? ???
A2 ??? ???
C1 ??? ???
C2 ??? ???
C3 ??? ???

 

[Marc S.]

Have a look here: http://www.avdweb.nl/solar-bike/electronics/ku63-motor-controller.html
There you will find scematics of the KU63 controller that is allmost identical to the (bigger) KU123. It helped me a lot to set up my system.

I run a Bafang BPM II (sometimes with and sometimes without connected hall sensors) and a small (oil cooled) Bafang SWXH without connected hall sensors with KU123 controllers. From my 4000km experiance with the BPM II the KU123 runs fine without hall sensors most times even from a dead stop without stuttering and to my surprise has a little bit more torque than with connected hall sensors.
The small sensorless SWXH will stutter from a dead stop but runs fine if the wheel starts turning. I haven't connected its hall sensors yet.

Maybe you don't have to bother too much to sort out the hall sensor wires...

 

[-dg]

I run a Bafang BPM II (sometimes with and sometimes without connected hall sensors) and a small (oil cooled) Bafang SWXH without connected hall sensors with KU123 controllers. From my 4000km experiance with the BPM II the KU123 runs fine without hall sensors most times even from a dead stop without stuttering and to my surprise has a little bit more torque than with connected hall sensors.
The small sensorless SWXH will stutter from a dead stop but runs fine if the wheel starts turning. I haven't connected its hall sensors yet.

This is my experience with a front sensorless SWXK with a KU63, it judders a little from a dead stop, but just push off before adding power and it is quite smooth. I don't miss sensors at all.

Not to threadjack, but I'm trying to choose motors for two new builds and your experience is relevant .

First, how did the oil cooling in the SWXH work out? Does it leak? What sort of oil? Is it effective? Is it quieter? How much power are you running with this? Also, how do you like the SWXH, it is one of the motors I'm considering for a 20" wheel bike build.

Second, I'm building a rear drive old steel mountain bike. The motors in contention are the BPM, the BPM II, and the CST. Any comments on these three, particularly, what is the difference between the BPM and the BPM II other than the cover style? Is there any reason to select one over the other? Also, the CST looks (in photos) like a very nicely made motor, but it's not clear that it has any functional advantage over the BPM/BPM II, and it may have less power handling. Any comments on this? Finally, how do you like the BPM II?

 

[Marc S.]

Not to threadjack, but I'm trying to choose motors for two new builds and your experience is relevant .

First, how did the oil cooling in the SWXH work out? Does it leak? What sort of oil? Is it effective? Is it quieter? How much power are you running with this? Also, how do you like the SWXH, it is one of the motors I'm considering for a 20" wheel bike build.

So far the oil cooling works out great. It does leak a little if you fill in too much oil. 50-60ml would be the right amount for the SWXH. 80ml for the BPM II. Don't use too much oil! The first oil filling of my SWXH worked out relativly clean but I used about 80ml for the second filling and that produced quite a mess in the first week till the exessive oil was out...
I use standard automotive transmission oil (ATF). Have a look here: http://endless-sphere.com/forums/viewtopic.php?f=2&t=37972

The ATF filling is effective and the motor runs quiter. More important the internal freewheel clutch now works flawless with oil lubrication at elevated power levels. The beauty of the SWXH is that you can apply the oil by filling it through the holes of the disk brake flange with a big siringe.

Currently I'm 'torturing' my little SWXH to produce between 1200-1500W (12s-14s LiPo at 35A, 1600-2000W off the battery) on my daily 45km commute. On top of that it's a 20" motor in a 26" wheel. So far it's runnung fine but will get quite warm to the touch after longer climbs at 40hm/h. I guess the cover reaches 60°C.

Second, I'm building a rear drive old steel mountain bike. The motors in contention are the BPM, the BPM II, and the CST. Any comments on these three, particularly, what is the difference between the BPM and the BPM II other than the cover style? Is there any reason to select one over the other? Also, the CST looks (in photos) like a very nicely made motor, but it's not clear that it has any functional advantage over the BPM/BPM II, and it may have less power handling. Any comments on this? Finally, how do you like the BPM II?

I can't help with the differention of the BPM, BPM II and CST. It looks like the CST is a BPM internally but without the fast windings available you need for 20" wheels. While I like my BPM II I might choose a MAC now if only for better availability of spare parts, steel gears and to be easier taken apart.
So far my BPM II developed a broken clutch and corroded ball bearings after 4000km before I tried to 'repair' it with an oil filling. That worked as a temporary solution but the clutch started slipping sometimes when accelerating hard.
I ran it with 15s LiPo and 35A (2000W off the battery) for the better part of it's life. Used to commute through the whole winter with lots of snow, parked outside for two month (that probably killed the bearings) and to my surprise the nylon gears are looking like new.

I'd like to take apart my SWXH after I get the BPM II repaired, implant a new clutch and steel gears, fill it with ATF and test how much power the little Bafang SWXH can produce on a daily basis. While it doesn't have the torque of the BPM II it is way more stealth and lighter. Since here in Germany motor power is legally restricted to 250W stealth is a good thing and can arguably be more importand than raw power!

An other idea is to get a new SWXH with the fastest winding for 16" wheels, add three steel gears, maybee alter the freewheel clutch for better lubrication, fill it with ATF, lace it in a 26" wheel and check out how that thing goes with 20s LiPo and 20-25A controller. Since I ride a ICE Sprint tadpole trike with good brakes I'm not afraid of speed...

An other thread you should check out and probably follow its developement is Justins test of hub motor cooling. http://endless-sphere.com/forums/viewtopic.php?f=2&t=48753
So far Justin is testing air cooled direct drive hubs but will get to test oil cooled geared hubs eventually. For me oil filling geared hub motors have proven its usefullness already despite the occational oil dropping. Apart from better heat transfer from the windings the better lubricated internal freewheel clutch makes all the difference.

 

[d8veh]

I have both the BPM and CST, which I run at 36v and 30amps. I'm not sure if the BPM is code 11 or 12 because I stuck a 250w label over it.

Both motors have identical performance. They have the same top speed, and on steep hill-climbs they stay side by side. On the down-hills, they also roll side by side.

Apart from being built better, the CST's main advantage is the cassette spline that makes it a much better choice if you have a bike with decent gears and changers that you don't want to mess up.

 

[-dg]

50-60ml would be the right amount for the SWXH. 80ml for the BPM II.
...
I use standard automotive transmission oil (ATF). Have a look here: http://endless-sphere.com/forums/viewtopic.php?f=2&t=37972

The ATF filling is effective and the motor runs quiter. More important the internal freewheel clutch now works flawless with oil lubrication at elevated power levels. The beauty of the SWXH is that you can apply the oil by filling it through the holes of the disk brake flange with a big siringe.

Currently I'm 'torturing' my little SWXH to produce between 1200-1500W (12s-14s LiPo at 35A, 1600-2000W off the battery) on my daily 45km commute. On top of that it's a 20" motor in a 26" wheel. So far it's runnung fine but will get quite warm to the touch after longer climbs at 40hm/h. I guess the cover reaches 60°C.
...
I can't help with the differention of the BPM, BPM II and CST. It looks like the CST is a BPM internally. While I like my BPM II I might choose a MAC now if only for better availability of spare parts, steel gears and to be easier taken apart.
So far my BPM II developed a broken clutch and corroded ball bearings after 4000km before I tried to 'repair' it with an oil filling. That worked as a temporary solution but the clutch started slipping sometimes when accelerating hard.
I ran it with 15s LiPo and 35A (2000W off the battery) for the better part of it's life. Used to commute through the whole winter with lots of snow, parked outside for two month (that probably killed the bearings) and to my surprise the nylon gears are looking like new.

I'd like to take apart my SWXH after I get the BPM II repaired, implant a new clutch and steel gears, fill it with ATF and test how much power the little Bafang SWXH can produce on a daily basis. While it doesn't have the torque of the BPM II it is way more stealth. Since here in Germany motor power is legally restricted to 250W stealth is a good thing and can arguably be more importand than raw power!

This was very helpful. I read the oil cooling is not snake oil thread some time ago, but this being the internet, opinions varied. Did you do anything to the SWXH besides add oil? Some on the oil thread recommend breather holes, but again the discussion was inconclusive. I await Justins results eagerly. One other concern was that the ATF might degrade the insulation on the phase wires, or the varnish, or the string or goop holding the wire bundles. How long has your SWXH been on oil?

How long have you been hammering that poor SWXH? I put an SWXK on my wifes bike and it seems happy so far at 21A on 12s, but I would never have expected to put 35A into it. Impressive. How long and how steep is that climb?

The CST internal construction is different than the BPM, it looks much cleaner with purpose built bits holding the wiring the sensors etc. There is a teardown photo thread on Pedelecs. The clutches look different, but I don't know how significant these changes are to the user, they may be more design for manufacturing. The other thing is that greenbikekit are offering parts for the BPM.

I agree stealth is good, although not so critical here in California. But I want my bike to be a bike, not an electric scooter. I tried another members bike with a MAC and cell_man 9 FET 30A controller and it was impressively strong. However it was much noisier than the SWXK. Pedestrians definitely noticed it and I found it a little annoying. So I'm hoping the BPM or CST will be strong but not as noisy.

 

[schwibsi]

The most important pin for me would be the on switch, that need's to be connected to the plus battery line.
It's torn off and I don't know, where it goes.

 

[Marc S.]

The most important pin for me would be the on switch, that need's to be connected to the plus battery line.
It's torn off and I don't know, where it goes.

That should be the thin red wire in the front on the left side in this picture:

 

[Marc S.]

This was very helpful. I read the oil cooling is not snake oil thread some time ago, but this being the internet, opinions varied. Did you do anything to the SWXH besides add oil? Some on the oil thread recommend breather holes, but again the discussion was inconclusive. I await Justins results eagerly. One other concern was that the ATF might degrade the insulation on the phase wires, or the varnish, or the string or goop holding the wire bundles. How long has your SWXH been on oil?

I tried a breather tube in the beginning but the SWXH and the BPM II seem happy enough breathing through the motor cable. But that will get quite messy if you fill in too much oil.

How long have you been hammering that poor SWXH? I put an SWXK on my wifes bike and it seems happy so far at 21A on 12s, but I would never have expected to put 35A into it. Impressive. How long and how steep is that climb?

The SWXH had been running about 100km with the first oil filling and than got stored with oil for about 6-7 month. When I reactivated it a month ago it got relaced to a 26" rim, a fresh oil filling and did an other 820km so far.
Before that it went about 5000km in a 20" rim and stock grease lubrication. At least 1000-1500km of that it had to suffer a 15s LiPo battery and 17A controller.

My daily 'hill' is not that steep (2-3°?) but a 15min long climb at 40km/h. Before reaching that 'hill' I hammer it about 30min through the city traffic with max acceleration at the traffic lights with the 14s battery. So far despite my best efforts neither the clutch or the Nylon gears gave up. It got a bit noisier in the last month, though...

The CST internal construction is different than the BPM, it looks much cleaner with purpose built bits holding the wiring the sensors etc. There is a teardown photo thread on Pedelecs. The clutches look different, but I don't know how significant these changes are to the user, they may be more design for manufacturing. The other thing is that greenbikekit are offering parts for the BPM.
I agree stealth is good, although not so critical here in California. But I want my bike to be a bike, not an electric scooter. I tried another members bike with a MAC and cell_man 9 FET 30A controller and it was impressively strong. However it was much noisier than the SWXK. Pedestrians definitely noticed it and I found it a little annoying. So I'm hoping the BPM or CST will be strong but not as noisy.

On my daily commute to work I'd like the electric scooter.
If I ride along for fun or doing a tour on the weekend a 10s battery will smooth out any hill while I'm cycling happily at 20-24km/h.
The other nice part about the ATF filling: The hub motor is not only quiter it got less resistance when pedaling. While the SWXH had barly noticable resistance to begin with the BPM sports a more pronunced drag.

Right now both motors complement each other quite nicely and have a place in my trike. The little SWXH is better as light to medium assist while cycling but can be pressed into more serious work if needed while the BPM is way better at hills and a really good solution pushing my trike quite easy through 20cm of snow in winter.
Snow, a trike and the BPM (or a Mac) is quite a winning combination! I never had that much fun commuting to work in winter!

 

[Samd]

Hello,
I'm about to mod a KU123 with different FETs (for 13s) ..

I run any KU series unmodified as 15S LiPO hot-off-charge as stock - which is 63.0 volts.
All of mine have had 63 volts components or better.

Keep main power wires, and the three red/black/green throttle wires.
Keep the five halls if you like the lack of stutter on takeoff.

Also, before you cut them off, locate the 3speed switch wires. Short whichever two give you the fastest wheel speed. They seem to be shipping the KU controllers of late with middle speed rating if no speed switch fitted. So don't short change yourself by upping your voltage and missing out on more top speed.

Cheers.

 

[-dg]

Before that it went about 5000km in a 20" rim and stock grease lubrication. At least 1000-1500km of that it had to suffer a 15s LiPo battery and 17A controller.

Wow. You are the person who has tried all the things I'm thinking about. I'm building two bikes, one 26" for general use that will get the BPM or CST, and a 20" folder (Xootr Swift) for taking on transit. The folder is to be kept as light as possible since it has to be carried on stairs at the stations. I'm trying to decide between two configurations for it:

A) Small front motor, Bafang SWXU or Cute 100, in a slow wind with 14s lipo to hopefully get the speed back up to around 30 kph and still have some climbing ability. Advantages: keeps the stock cassette drivetrain, stealthy, weighs only 4 kg more than the bare bike with 5 ah batteries. Limited speed helps range with small battery. Disadvantages: limited power, might turn out too slow and need more voltage.

B) SWXH in the rear with 12s lipo. Advantages: plenty of power for this application. Can use 12s batteries maybe (common with other packs). Disadvantages: Needs 11 tooth freewheel, may have to change shifter, heavier, may need more battery due to lack of self discipline, biases bike weight to the rear (harder to carry).

Questions are:

- how fast was the SWXH in the 20" wheel with 15s?

- any experience with SWXU or the Q-100?

- any thoughts on configurations A and B?

 

[-dg]

I have both the BPM and CST, which I run at 36v and 30amps. I'm not sure if the BPM is code 11 or 12 because I stuck a 250w label over it.

Both motors have identical performance. They have the same top speed, and on steep hill-climbs they stay side by side. On the down-hills, they also roll side by side.

Apart from being built better, the CST's main advantage is the cassette spline that makes it a much better choice if you have a bike with decent gears and changers that you don't want to mess up.

Thanks for this. The bike for this build, an older Trek 950, has 7 speed shifters and cassette. I'm fine with 7 speed, have a pile of freewheels, and will replace the shifters anyway since they won't fit on the bars I want to use. So the cassette feature while nice does not really matter for this build. If the BPM and CST perform the same, and the freewheel is acceptable over the cassette, is there any reason to choose the CST over the cheaper BPM? Is it smoother or quieter, or more waterproof or more reliable or better looking or anything detectable without opening it?

 

[Marc S.]

Questions are:

- how fast was the SWXH in the 20" wheel with 15s?

- any experience with SWXU or the Q-100?

- any thoughts on configurations A and B?

1. 40km/h in a recumbent trike.

2. Nope.

3. From the little experiance I had with front wheel motors (sat on two bikes) I'd go with rear wheel drive any day.
11 tooth freewheels are not that hard to come by. Since the SWXH is relativly light I wouldn'd mind the 3kg in the back wheel.

I'm thinking of a light 20" wheel folder as well. Thats actually the only non recumbent bike that I would consider to own after two years with my recumbent trike.
My configuration for the folder would be a faster 16" Bafang SWXH in 20" rear wheel (with ATF lubrication to avoid any free wheel clutch problems from the start), KU123 controller , 12s-14s LiPo/10-15Ah battery, Cycle Analyst V3.
I'd restrict the Motor to 20-25A for every day use but like to keep a 35A setting to embarass sports cars at the traffic lights.

 

[d8veh]

is there any reason to choose the CST over the cheaper BPM? Is it smoother or quieter, or more waterproof or more reliable or better looking or anything detectable without opening it?

No difference that I've noticed except that the waterproofing looks a bit better on the CST. They sealed the wires through the axle. Whichever one you get, it's still better to make a drip-loop on the motor wires and put some gasket sealant or loctite on the disk brake screws.

 

[d8veh]

The most important pin for me would be the on switch, that need's to be connected to the plus battery line.
It's torn off and I don't know, where it goes.

That should be the thin red wire in the front on the left side in this picture:

That must be an old version. Here's the latest.
Those two pads marked ZX next to the end of the thin red wire are for selecting 36v or 48v operation. Connected is 36v, open is 48v.

 

[schwibsi]

I have both the BPM and CST, which I run at 36v and 30amps. I'm not sure if the BPM is code 11 or 12 because I stuck a 250w label over it.

Both motors have identical performance. They have the same top speed, and on steep hill-climbs they stay side by side. On the down-hills, they also roll side by side.

Apart from being built better, the CST's main advantage is the cassette spline that makes it a much better choice if you have a bike with decent gears and changers that you don't want to mess up.

Thanks for this. The bike for this build, an older Trek 950, has 7 speed shifters and cassette. I'm fine with 7 speed, have a pile of freewheels, and will replace the shifters anyway since they won't fit on the bars I want to use. So the cassette feature while nice does not really matter for this build. If the BPM and CST perform the same, and the freewheel is acceptable over the cassette, is there any reason to choose the CST over the cheaper BPM? Is it smoother or quieter, or more waterproof or more reliable or better looking or anything detectable without opening it?

If you have the choice between CST or not, always go CST. the sprockets on the outside of the regular motors have no bearings under them, which is very bad from an engineering standpoint. If you have the option to go CST and can affort CST, please don't think twice and go for it. I know some people who would sell their soul to the devil for a MAC-type motor with a CST.

 

[-dg]

If you have the choice between CST or not, always go CST. the sprockets on the outside of the regular motors have no bearings under them, which is very bad from an engineering standpoint.

For a normal bike hub I agree. But not so much for a hub motor. The difference is, the bearings that get moved outboard in the Shimano cassette design primarily support the axle. Freewheel bearings have never been a problem, they are large diameter and never spin under load, only when coasting.

The engineering problem with the traditional bike hub design that the Shimano cassette design addresses is the length of the unsupported axle between the bearing in the hub and the dropout on the drive side. This puts a lot of bending load on the axle which got worse when the spacing was lengthened from 120 to 126 to 130 to 135 to add more cogs (8,9 speed etc). Longer 10 mm axles tend to break right at the bearing. This is why there were never until recently 8, 9, 10 etc cog freewheels. The Shimano cassette design moves the bearing outboard which reduces the bending load on the axle eliminating the axle breakage on wide spaced rears.

A different solution is to strengthen the axle. Hub motors have done this by using 12 and 14 mm axles instead of 10 mm. I don't think that the engineering problem the cassette solves applies to hubs with large diameter axles.

What we are left with is not an engineering problem, but rather a marketing problem. The market for 8,9,10,11 speed freewheels is confined to hubs that don't use the Shimano cassette design but have strong enough axles to handle the extra length. Which is not very many, hence only one vendor of rather crappy freewheels (I suspect that the reason DNP 8,9,10 etc freewheels even exist is for hub motors). However good traditional 6 and 7 speed freewheels are still available as replacement parts for pre-cassette bikes, just not in very many sizes. There is no reason not to use them if the gearing works for the application.

 

[schwibsi]

The problem in my view lies in the connection of the freewheel to the sidecover of the motor.
I've seen a few that all broke in the same spot.

 

[-dg]

The problem in my view lies in the connection of the freewheel to the sidecover of the motor.
I've seen a few that all broke in the same spot.

Ah, that. Yes I agree that the sidecover needs to be strong enough. Regular hubs never break that way so they are doing something less good with the motors. I've heard of this failure on MACs, BMCs, BPM2, and early CST but this is all hearsay. Is this known to be common on the BPM?

I am actually pretty tempted by the CST. On the other hand, I can get gears, clutches, sidecovers for the BPM quite cheaply, or even just get a whole spare BPM for the price of the CST. So I dunno. There is a lot of mostly good experience out there with the BPM.

 

[-dg]

The problem in my view lies in the connection of the freewheel to the sidecover of the motor.
I've seen a few that all broke in the same spot.

Ah, that. Yes I agree that the sidecover needs to be strong enough. Regular hubs never break that way so they are doing something less good with the motors. I've heard of this failure on MACs, BMCs, BPM2, and early CST but this is all hearsay. Is this known to be common on the BPM? What models of motor have you seen break this way?

I am actually pretty tempted by the CST. On the other hand, I can get gears, clutches, sidecovers for the BPM quite cheaply, or even just get a whole spare BPM for the price of the CST. So I dunno. There is a lot of mostly good experience out there with the BPM.

 

[schwibsi]

I've seen it on MACs (quite a few), Bafangs and Q100.

You're making a good point about the cost. The price for the cassette version of the Bafang is ridiculous.
Xofo might be the solution here.
But I've not had any first hand experience with them so I don't know, how they are in terms of power, efficiency and noise.

 

[schwibsi]

Does anyone have any mor information on the controller pins?

Especially the on/off pin and the three-speed switch?

 

[d8veh]

Does anyone have any mor information on the controller pins?

Especially the on/off pin and the three-speed switch?

If you look at my photo above, the thin red wire is the on/off, and connects to a pad marked VCC.
Those three wires on the left (green, blue and black) are the three speed switch wires.

 

[schwibsi]

Thanks, that helped.

You wouldn't happen to know, which pin the black one should be connected to for the highest power level?