KT Controller – still drawing power above cut-off speed?

[Jodel]

Having read all the positive comments and feedback about KT controllers, I recently fitted a KT sine wave controller to my solo bike – comments on that to follow.

I bought the 15 amp 6 MOSFET version which fits into a Hailong 1 battery base. The original Luishi controller was a 17 amp model, but as I’m only around 70Kg and I have a Shengyi 48 volt DWG22C rear hub motor, there’s still more than enough oomph for me.

I really like having a watt meter on the LCD 3 display as I now have an idea how much effort the motor is contributing. Which leads me to my query for the experts on the forum.

I’ve set the maximum speed to the legal limit of 25 kph / 15.5 mph. Even when comfortably above that speed (20mph plus) the watt meter shows that I’m still using between 15 – 25 watts when peddling. Naturally, if I stop peddling, the watts meter shows zero.

I would have expected the motor output to drop to zero above the cut-off speed, but is this a characteristic of the KT controller? This isn’t a major issue as everything works well, I just wondered if it was normal for the KT controller.

 

[saneagle]

The power should cut off completely. There are a couple of things that come to mind. One is the C4 setting.which I think can override the speed limit in one case. Secondly is the way the speed display and control work. It uses the speed sensor to display the speed, but it uses the hall sensors to get the speed for the cut-off. I did a test with a spoofed speed signal. The motor was always cutting off at the same rpm, but different speeds were showing on the LCD in accordance with the spoofed signal, so check that you have the right wheel size and have a play with the P1 setting.

 

[Jodel]

Thanks for the feedback. My C4 & P4 are both set to 0 - i.e. the throttle works all the time. I'll try different settings here. Setting C4 to 2 =
"Zero startup, throttle speed limited to rider's specification" which may be the setting in question.

Interesting to know that the speed display and cut-off speed are taken from different sensors. The speed and distance display are pretty much bang on when compared with my Cateye computer and Strava / Map my ride.

The wheel size is correct at 700C, but I understand from previous threads that the P1 setting can need a bit of 'tuning'. At the moment, it's set for 5.3 x 16 = 85 - but I'll mess about with this setting and report back in due course. The motor seems happy enough and is very smooth / quiet with this setting.

Oh, another anomaly I found was in the setting of the P5 Power monitoring. I set this to 0 initially, but after riding the bike and checking various parameters, I discovered this had changed itself to a value of 12. My battery is a 48 volt x 12AH. I couldn't reset this value back to 0, it would only go down as far as 1. At this stage I didn't want to do a full factory reset, but as an inveterate 'fiddler' this kind of stuff keeps me entertained!

As an aside, I've read that the response to the throttle on KT controllers can be a bit 'on / off'. Mine isn't - it's very smooth and progressive
.

 

[saneagle]

Thanks for the feedback. My C4 & P4 are both set to 0 - i.e. the throttle works all the time. I'll try different settings here. Setting C4 to 2 =
"Zero startup, throttle speed limited to rider's specification" which may be the setting in question.

Interesting to know that the speed display and cut-off speed are taken from different sensors. The speed and distance display are pretty much bang on when compared with my Cateye computer and Strava / Map my ride.

The wheel size is correct at 700C, but I understand from previous threads that the P1 setting can need a bit of 'tuning'. At the moment, it's set for 5.3 x 16 = 85 - but I'll mess about with this setting and report back in due course. The motor seems happy enough and is very smooth / quiet with this setting.

Oh, another anomaly I found was in the setting of the P5 Power monitoring. I set this to 0 initially, but after riding the bike and checking various parameters, I discovered this had changed itself to a value of 12. My battery is a 48 volt x 12AH. I couldn't reset this value back to 0, it would only go down as far as 1. At this stage I didn't want to do a full factory reset, but as an inveterate 'fiddler' this kind of stuff keeps me entertained!

As an aside, I've read that the response to the throttle on KT controllers can be a bit 'on / off'. Mine isn't - it's very smooth and progressive
.

P5 doesn't do anything to the motor. It's a simple damping coefficient for the battery display. Keep it round about 15 if you want to use the battery meter as a fuel guage.

 

[Jodel]

Well, I had a chance to try a few things today, but I still have the motor drawing watts above the cut off speed.

In addition to trying the bike out on the road, I also put it on a workstand and turned the cranks to see what was happening. I’ll try to keep this logical.

1) Changed the P1 value of 85 from double to half (and many small steps in between) with no effect other than to alter the speed the motor would spin.
2) Spinning the cranks to give speeds over 20mph, the motor was still engaged and showing a consistent drain of 13 watts.
3) Changed from current control to speed control – no difference.
4) Tried all values for P2 – no difference
5) Tried all values for C2 – no difference – I was quite surprised that the motor actually ran with all these different settings.
6) Tried all values for C4 – no difference

What I did notice was that if I used the throttle and kept the speed at or below 10mph, there were no watts showing on the LCD, although the motor was spinning.

I have another bike with the same hub motor, but with the Luishi speed type controller. I tried this on the workstand to confirm that the motor cuts out if the set speed is exceeded. It was quite clear that the motor stopped when the speed went over the set limit.

Although the Shengyi motor is quite quiet with the KT controller, my other bike is noticeably quieter. I appreciate that describing noise is not very helpful as one persons quiet sound is another persons deafening racket. There is an appreciable difference in sound / noise between the two motors though, with more gear type noise from the KT set-up.

I would like to get to the bottom of this as I don’t want to damage the hub motor with incorrect parameters. Any comments gratefully received!

 

[saneagle]

The gear type noise you hear is the commutation pulses at low speed. It's much more noticeable on a motor with a low internal reduction ratio, and it's amount also is proportional to current. As the motor speeds up, the individual pulses can no longer be resolved to something you can hear of feel.

If changing P1 affects the maximum speed that the motor will spin, why don't you set it to the speed you want it to cut off?

Don't forget that C4 goes with P4. The two settings work together.

 

[Jodel]

Thanks for the info.

The only way I could confirm if there is any difference in the noise of the gears / motors is to swap the wheels over from one bike to another - a task for another day I think.

The P1 setting did change the actual speed of the motor, but unfortunately did not make it cut out - it seems to stay running (albeit at very low power above the cut-off speed) whenever you are peddling. The cut-off does work after a fashion. When riding the bike, if the speed is oscillating just above / below the 15.5 mph mark, you can feel the power cutting in and out but there is always a small residual wattage being drawn for some reason and the motor stays powered constantly.

I'll try some different combinations of P4 and C4 to see if that resolves things.

Bit of a puzzler this one.

 

[Sturmey]

......... When riding the bike, if the speed is oscillating just above / below the 15.5 mph mark, you can feel the power cutting in and out but there is always a small residual wattage being drawn for some reason and the motor stays powered constantly........

I suspect that this may be done intentionally in some circumstances so you get a faster and smoother take up of power (and possibly save the clutch) when (for example) starting to climb a hill after descending.
The reason I think this may be the case is that the Pendleton Somerby had a well reported annoying habit of taking a few seconds to regain power in a similar circumstances as for examples reported below.
There might be a connection in that feeding a few watts to the motor to keep it spinning might help it kick in faster and smoother..

Pendleton Somerby Upgrade

Hi, My wife has a Pendleton Somerby E Bike, fitted with a Bafang Rear Hub RM 201.D 250w Motor, a 36V 8.7ah Battery and a King Meter 790 display. (brake cut offs also fitted) She has a big problem with pas delay when starting to climb a hill after descending. The power simply doesn't seem to kick...

www.pedelecs.co.uk

Bafang conversion kit : Recovery from over-speed is slow

Heyo, I have a 250 W Dillenger front hub kit on a 10-y-old hybrid commuter which has transformed my commute, but: When I sail down a hill which puts the road speed over the limit of 15.5 mph, the motor freewheels (great), but when I start up the next hill the controller waits until the road...

www.pedelecs.co.uk

 

[saneagle]

Thanks for the info.

The only way I could confirm if there is any difference in the noise of the gears / motors is to swap the wheels over from one bike to another - a task for another day I think.

The P1 setting did change the actual speed of the motor, but unfortunately did not make it cut out - it seems to stay running (albeit at very low power above the cut-off speed) whenever you are peddling. The cut-off does work after a fashion. When riding the bike, if the speed is oscillating just above / below the 15.5 mph mark, you can feel the power cutting in and out but there is always a small residual wattage being drawn for some reason and the motor stays powered constantly.

I'll try some different combinations of P4 and C4 to see if that resolves things.

Bit of a puzzler this one.

Have you checked that your clutch isn't stuck?

 

[Jodel]

Guys, thanks for the comments.

Sturmey - I understand where you're coming from, but I don't believe this is the issue. My Luishi controller is similar in behaviour to that described in your linked posts. For example, if you are close to the cut-off speed, the motor cuts off if you exceed the limit, but it takes slightly longer than I'd like to re-start when the speed drops back again. In that respect, the KT controller seems much more responsive to minor variations in speed. The Luishi controller seems to have a +/- 'window' of about 2 mph before it reacts, the KT seem to need only about 1 mph to respond.

saneagle - I doubt if the sprag clutch is stuck for a couple of reasons. The bike was fine when using the Luishi controller, there is no drag or motor gear noise when peddling at '0' assist with the KT and that motor has only done about 1,500 miles so still quite a youngster. Doesn't mean it isn't stuck, but I'm not sure its worth dismantling the motor (yet!) to check.

I'll probably try a bit of mixing and matching of components between my two DWG22C equipped bikes. If the KT controller causes the same issue with my other bike, then it may be that KT just doesn't work well with Shengyi motors. Hopefully I'll get to the bottom of it in due course.

I'll make further comment on my experience of current control versus speed control in due course. They are certainly different, with current control feeling more 'natural', but not necessarily better if you have old arthritic ankles and limited leg strength. As a sweeping generalisation, my initial impressions are that current control is good for solo bikes, speed control works better for the tandem.

 

[Sturmey]

Guys, thanks for the comments.

Sturmey - I understand where you're coming from, but I don't believe this is the issue. My Luishi controller is similar in behaviour to that described in your linked posts. For example, if you are close to the cut-off speed, the motor cuts off if you exceed the limit, but it takes slightly longer than I'd like to re-start when the speed drops back again. In that respect, the KT controller seems much more responsive to minor variations in speed. The Luishi controller seems to have a +/- 'window' of about 2 mph before it reacts, the KT seem to need only about 1 mph to respond.

saneagle - I doubt if the sprag clutch is stuck for a couple of reasons. The bike was fine when using the Luishi controller, there is no drag or motor gear noise when peddling at '0' assist with the KT and that motor has only done about 1,500 miles so still quite a youngster. Doesn't mean it isn't stuck, but I'm not sure its worth dismantling the motor (yet!) to check.

I'll probably try a bit of mixing and matching of components between my two DWG22C equipped bikes. If the KT controller causes the same issue with my other bike, then it may be that KT just doesn't work well with Shengyi motors. Hopefully I'll get to the bottom of it in due course.

I'll make further comment on my experience of current control versus speed control in due course. They are certainly different, with current control feeling more 'natural', but not necessarily better if you have old arthritic ankles and limited leg strength. As a sweeping generalisation, my initial impressions are that current control is good for solo bikes, speed control works better for the tandem.

Beautiful morning here so I brought the bike for a 54Km spin. The bike has a rear Yosepower 350W motor (2018) with its original 36v controller and LCD3. I set the speed restriction to 25 Km/hr. One section of the ride consisted of a 7Km steady descent. On this descent at about 30Km/hr, while freewheeling, the LCD reads 0 w motor power. On pedaling, my motor power reading jumped momentarily to 18w and then to 9w. The LCD continues to show this reading of 9w once I keep pedaling and this reading goes back to 0w immediately when I stop pedaling. I repeated this over and over and the LCD consistently reads 0w when freewheeling and 18w reducing to 9w while pedalling as described above.
I checked my clutch and the wheel freewheels freely in the forward direction and stiffly in reverse when lifted off the ground.
Anyone with a bike with a KT controller & LCD3 and a convenient hill can easily do the above check to verify if this is the case with there bike.

 

[saneagle]

Guys, thanks for the comments.

Sturmey - I understand where you're coming from, but I don't believe this is the issue. My Luishi controller is similar in behaviour to that described in your linked posts. For example, if you are close to the cut-off speed, the motor cuts off if you exceed the limit, but it takes slightly longer than I'd like to re-start when the speed drops back again. In that respect, the KT controller seems much more responsive to minor variations in speed. The Luishi controller seems to have a +/- 'window' of about 2 mph before it reacts, the KT seem to need only about 1 mph to respond.

saneagle - I doubt if the sprag clutch is stuck for a couple of reasons. The bike was fine when using the Luishi controller, there is no drag or motor gear noise when peddling at '0' assist with the KT and that motor has only done about 1,500 miles so still quite a youngster. Doesn't mean it isn't stuck, but I'm not sure its worth dismantling the motor (yet!) to check.

I'll probably try a bit of mixing and matching of components between my two DWG22C equipped bikes. If the KT controller causes the same issue with my other bike, then it may be that KT just doesn't work well with Shengyi motors. Hopefully I'll get to the bottom of it in due course.

I'll make further comment on my experience of current control versus speed control in due course. They are certainly different, with current control feeling more 'natural', but not necessarily better if you have old arthritic ankles and limited leg strength. As a sweeping generalisation, my initial impressions are that current control is good for solo bikes, speed control works better for the tandem.

You don't need to dismantle the motor to check the clutch. You just grab the wheel with your hand and move it forwards and backwards. You should feel the difference between the wheel turning the motor backwards and the clutch disengaged when moving forwards.

If anybody has got a KT controller with LCD3 and someone to help them, could you try this: Switch on the bike ready to go, then lift the wheel off the ground and turn the wheel backwards as fast as you can by hand. Ask your helper to look at the LCD3 to see if any watts are appearing in the display. I'm wondering what happens when the motor generates.

Many years ago, I had a Cycle Analyst on my bike with a normal geared hub-motor. At the end of each journey, it showed me how many watt-hours I'd generated, even though I didn't have a regen controller. The CA was between the battery and controller, so the charge must have been coming out of the controller somehow. I assumed that the charge came from the motor still rotating after power was cut because it takes a few seconds to slow down.

 

[Jodel]

Sturmey - that mimics my experience almost exactly. I was checking the display on a long descent yesterday and saw pretty much the same as you. I wasn't aware of any motor / gear noise when I was riding the bike as the wind noise masked it. At home with the bike on the workstand, I could tell the motor was still engaged when peddling above the cut-off speed.

Interesting too that the wattage figures you quote on your 36 volt system are almost the same as my 48 volt installation. I think mine were 18 - 13 watts - I'll need to check again to be sure.

saneagle - sorry, I thought I had to take the motor to bits to examine the clutch. It behaves just like Sturmey's - there is resistance when rotating in reverse but the wheel freewheels easily forwards. I like the idea of turning the wheel backwards to see if any watts are generated. I didn't know you could do that. Is this linked to the C13 setting? (Mine presently set to 0)

The Cycle Analyst looks like an excellent (if pricey) bit of kit. A bit too complex for a thicko like me though.

 

[saneagle]

Just thinking. It seems that the controller uses the motor rpm to cut the power (P1 setting), not the speed sensor. In that case, the motor will be held at its cut off rpm even when the bike is over the indicated cut-off speed, so it will run with slightly less power than when the wheel is in the air because it'll be helped by friction rather than hindered, and it won't suffer the affect of air resistance, so 15w sounds about right.

That will only happen if you keep the power on by pedalling or throttle. When freewheeling, the motor will stop, like normal, and no power will show on the LCD.

 

[Jodel]

saneagle, that makes sense in this case. With my P1 setting at 5.3 x 16 = 85, the speed on the display and the trip distance are pretty much bang on when compared to my cycle computer and phone apps. I expected the power indicator to drop to zero when above the cut off speed (which is what I get with the Luishi controller) but maybe KT just works in a slightly different way with the Shengyi motor and also Sturmey's Yosepower motor. The KT controller is listed as being compatible with Shengyi motors and it did power up with minimal fiddling from me.

If I'm understanding you correctly, the motor will just be continuing to spin the gears inside the hub at the cut-off rpm / speed, but not fast enough to engage the clutch mechanism.

I still intend to try swapping wheels / controllers around ('cos I like to fiddle with things) to see if there is any difference in gear type noise. I just didn't want to risk damaging the controller / motor with incorrect settings. My Luishi / Shengyi combo is slightly quieter than the KT / Shengyi, even though that motor has done about 5,000 miles on the tandem, which is a much greater load.

I do really like having a real-time display of watts and of battery voltage, something I think only the KT controllers / LCDs can offer.

 

[Sturmey]

.........If I'm understanding you correctly, the motor will just be continuing to spin the gears inside the hub at the cut-off rpm / speed, but not fast enough to engage the clutch mechanism........

I have seen discussions suggesting that by running the motor as above, the clutch will have a sort of gentle engagement so when the power ramps back up it will prevent a 'clunk'. I would like to know more. The only discussion I can readily find is below but refers to mid drives)

VESC configuration for mid-drives

I'm playing with a Makerbase 75200 VESC based controller with a Cyclone 3000 motor on a normal 10 speed bicycle drivetrain and I cannot smooth initial acceleration with ADC current control. I've got throttle mapping adjusted how I like but when I engage the throttle I get an aggressive "clunk"...

endless-sphere.com