Are you absolutely certain that its not the controller? My daughters Pendleton suffered that problem with the santroll controller and the problem disappeared when the controller was changed. Seajay and Bikes4two report the same below. I know this may be a different case but nevertheless there are reports of this type of problem been cured by replacing the controller.

PS. My LCD3 has L settings. (There is one setting for switching the controller to sensor-less working which I presume could cause a problem if incorrectly set).

 

https://www.pedelecs.co.uk/forum/threads/pendleton-somerby-upgrade.37785/

Yes, but Seajay solved that issue by changing to a KT controller. Benjamin already has a KT controller.

I seem to be at the point of having no further settings alterations or fettling (apart from adjusting the operation point of the brake switch) that I can do here. The controller seems to be doing what it should which would lead to consideration of it being a physical issue rather than electrical.

So, is it worth having a look in the motor to see if it is a grease/pawls issue?

If so, I'm going to need some help because I've never stripped an AKM.

Not knowing anything whatsoever about this utterly bizarre and ineffably incomprehensible hub motor WTF PLC setting infested stuff, I'm more inclined to believe it's a firmware befuddlement issue caused by some combination of incorrect speed and/or pedaling speed sensing - unexpected things can happen when the speed sensed is incorrect on my BBS01B. I seem to recall you couldn't get the external speed sensor magnet close enough, perhaps try gluing magnets on top? Also, one major difference betwen your KT setup and [mention=3847]saneagle[/mention] 's & [mention=9614]Nealh[/mention]'s, is they don't use 12 magnet PAS (for whatever mysterious reason, as far as I recall)... therefore personally, I'd certainly investigate those avenues before attempting motor disassembly searching for an apawling problem. I'm sure wiser hub motored minds will wade in soon rotating at high speed like their eyes.

 

When I think Wales, I think mid-drive.

 

BTW What did Topbikekit have to say about this? Or whoever you bought this conversion gear from?

From my recollection from when I opened my AKM128C many years ago, there is little to do or see by opening the hub the only pawls are the ones in the freehub body. The freehub is integrated to the hub side plate and isn't removable.

Clutch wise any there are in the sealed clutch unit there is little to do or see by opening the hub apart from a possible replacing a bearing ( if worn) or applying more grease if sparse.

There is little one can do by opening the hub bar some maintenance if the hub is noisy or parts are worn.

 

KT controllers can be a bit hit and miss with 12 pole PAS, esp non KT PAS.

Whether one uses 8 , 10 or 12 pole there is little difference in the rotational activation time . My go to is BZ10 as I use exterior bearing cup BB's, rotational, activation is approx. 1/8th rotation so very responsive.

Edited October 12, 2025Oct 12 by Nealh

I've been following this with much interest, but without anything to add. However (and this is probably a complete red herring) I can recall a couple of similar occurrences on a long ride last summer with my Yosepower AKM95 motor and KT controller.

I decided to "see how fast I could go while pedalling" down a long steep road with an equal up-slope from the bottom.

I got to about 25 mph and when I hit the rise speed dropped to about 8 mph before I got any obvious assistance from the motor, which then accelerated slowly back up to the 15.5mph speed limit. This was in PAS 5 with a 36v. battery. I didn't observe the watts on the display, but did notice that the voltage had dropped quite considerably before recovering. I just dismissed it as being due to the steep hill.

When I get the opportunity I will try it with my 48v. battery, although unfortunately it is some distance away and not a regular ride. Do your symptoms change with battery capacity?

If you change wheel size on your KT display and this "clutch engagement" failure doesn't occur at the same bike speed (cutoff speed adjusted to suit) downhill etc, I hazard it isn't an apawling problem - a separate bicycle computer could be used indicate if the alleged "clutch engagement" failure happens at the same speed and in the same circumstances.

Edited October 12, 2025Oct 12 by guerney

From my recollection from when I opened my AKM128C many years ago, there is little to do or see by opening the hub the only pawls are the ones in the freehub body. The freehub is integrated to the hub side plate and isn't removable.

Clutch wise any there are in the sealed clutch unit there is little to do or see by opening the hub apart from a possible replacing a bearing ( if worn) or applying more grease if sparse.

The clutch is a separate unit fixed to the side-plate by the three screws. It is easy to service, as the rollers are very accessible. I can't really see anything that could be wrong with it becaude its operation is simple. I'm pretty sure that any problem with grease would have sorted itself out by now.

 

To open the motor, you remove the three screws in the side, then use those holes to rotate the side-plate off. Ideally, you need a tool to do that, but there are many other ways to do it if you have a good imagination. I recommend heating up the area between the spokes until it's just about too hot to touch, then the side-plate unscrews easily. Without heating, it can be very stuck.

 

There are very tiny thin springs behind each roller. Make sure you don't lose any of them.

 

If it were my motor, I'd leave it for now. It might still sort itself out if it's a problem with grease or rollers.

I've been following this with much interest, but without anything to add. However (and this is probably a complete red herring) I can recall a couple of similar occurrences on a long ride last summer with my Yosepower AKM95 motor and KT controller.

I decided to "see how fast I could go while pedalling" down a long steep road with an equal up-slope from the bottom.

I got to about 25 mph and when I hit the rise speed dropped to about 8 mph before I got any obvious assistance from the motor, which then accelerated slowly back up to the 15.5mph speed limit. This was in PAS 5 with a 36v. battery. I didn't observe the watts on the display, but did notice that the voltage had dropped quite considerably before recovering. I just dismissed it as being due to the steep hill.

When I get the opportunity I will try it with my 48v. battery, although unfortunately it is some distance away and not a regular ride. Do your symptoms change with battery capacity?

I haven't noticed any effect of lower voltage. Otherwise it sounds exactly like what I'm observing. There are a lot of hills around here so I regularly get well over 20+mph.

As Saneagle and Neal don't seem to have these problems I will look at the pas, change it to a 10 pole. It's about all I have left.

The only other thing I can think of is a heat related issue. Is it possible that heat generated on the uphill at full power gets the opportunity to spread out from the MOSFETs to a temperature sensor during the downhill, and triggers a thermal shutdown?

 

Is the controller out in free air for best cooling? Can more heat sinking be added to its casing?

 

If you test ride not using throttle and instead manually change PAS levels, but only use 4, does that change anything?

 

Maybe a 12 MOSFET controller..,

I thought getting the 9 mosfet would be enough, it's being run at 20A max so 2A below it's capacity. I have checked after climbs and, as far as I can detect, it only gets warm not hot. However, with the previous motor, it did get very hot, hot enough to not be able to keep my finger on it. I do wonder if this has caused some damage.

I have it in a hard plastic junction box gut I've left some of the grommets out front and back to give it some air flow.

A simpler way to reduce heating and see if anything changes is to change max current.

 

Any change helps eliminate possible causes. 15A?

Another probably broken straw to clutch at...

 

I'm assuming that your speedometer is the LCD, and it always shows a sensible number?

 

No possibility of the cut-in speed being computed from the still rotating motor?

 

If I had one, I would try an alternative external speed sensor plus whatever settings are needed just to eliminate.

It is indeed the lcd showing speed. With the current settings it does seem to be giving sensible readouts though I haven't exhaustively checked with gps. I think I have an external sensor in the shed, I'll see if I can find it. I seem to remember though that, when first setting up, I did try it and couldn't get the lcd to respond. Probably settings somewhere.

 

I've now adjusted the brake switch operation point so that it's further away from the brake bite point. It's made it easier to activate brake cutoff but I still messed it up in an absent minded moment when out yesterday. It's this continual need for awareness to be right on top of it that's getting to me, when I just want to enjoy the ride and take in my surroundings.

It won't be due to thermal cut out, not down hill.

It is indeed the lcd showing speed. With the current settings it does seem to be giving sensible readouts though I haven't exhaustively checked with gps. I think I have an external sensor in the shed, I'll see if I can find it. I seem to remember though that, when first setting up, I did try it and couldn't get the lcd to respond. Probably settings somewhere.

 

I've now adjusted the brake switch operation point so that it's further away from the brake bite point. It's made it easier to activate brake cutoff but I still messed it up in an absent minded moment when out yesterday. It's this continual need for awareness to be right on top of it that's getting to me, when I just want to enjoy the ride and take in my surroundings.

You can't use two speed sensors at the same time. If you want to use an external one, you have to find a way to disable the one in the motor, which is easy if you have block connectors and virtually impossible if you have sealed ones.

It won't be due to thermal cut out, not down hill.

Shouldn't be, but heat soak can give a delayed reaction. Thermal cut-out might be just over the top of the hill, before most of descent.

You can't use two speed sensors at the same time. If you want to use an external one, you have to find a way to disable the one in the motor, which is easy if you have block connectors and virtually impossible if you have sealed ones.

Yes I have block connectors, what would I need to do to disable the internal sensor?

Some KT's have an ext spd sensor wire outlet.

One though has to disconnect the white/signal Hall wire at the controller.

 

If no ext spd sensor wire outlet then one will need to use the wire port for the ext spd sensor signal wire. Splice the ext spd sensor 5v red and Gnd Blk in to any 5v & ground wiring, one can splice into the throttle or PAS wiring or the Motor Hall wiring.

Edited October 14, 2025Oct 14 by Nealh

Just been out on a provisions run and, right from the off, the system was playing up, but here's the main weirdness.

Going into the local town I'm going down a 1:6 hill. So pas is off, obviously brakes are on, speed up to 30+.

The grocery shop is just at the bottom of the hill, so I parked up and went in to get my veg. By the time I came out ( about 10 minutes) the system had timed out and turned itself off - all good. I turned on, set pas at 3 and set off. There was an initial surge at the first half turn of the pedals, then it went to zero and stayed there. I went down through the gears to speed up the pedals but still nothing, I pressed the throttle (which the display acknowledged) but still nothing. By this time I'm doing about 5mph up the rise into the centre of town. Then it slowly kicked in, starting at 40/50watts it very slowly ramped up to the 300 or so of pas3. Only then did it start to respond to the throttle slowly ramping up to 650w.

All the previous pondering's about motor speed v wheel speed wouldn't seem to apply here. All plugs are in and haven't been disturbed (certainly not while I was in the shop) and the system was mostly behaving for the rest of the run, so I don't think there's an intermittent connection, at least not externally.

I am coming to the conclusion that there is something hooky about this controller. Whether it's programming, wandering component values, or damage caused by heat when running the previous motor - I don't know. But, in the face of others more knowledgeable than myself saying they have no such problem with this system and me not being able to establish reliably repeatable cause and effect, I don't know what other conclusion to draw.

I think the only option I've got now is, when I've got some spare cash, I will order a direct replacement and see how that behaves.

Just been out on a provisions run and, right from the off, the system was playing up, but here's the main weirdness.

Going into the local town I'm going down a 1:6 hill. So pas is off, obviously brakes are on, speed up to 30+.

The grocery shop is just at the bottom of the hill, so I parked up and went in to get my veg. By the time I came out ( about 10 minutes) the system had timed out and turned itself off - all good. I turned on, set pas at 3 and set off. There was an initial surge at the first half turn of the pedals, then it went to zero and stayed there. I went down through the gears to speed up the pedals but still nothing, I pressed the throttle (which the display acknowledged) but still nothing. By this time I'm doing about 5mph up the rise into the centre of town. Then it slowly kicked in, starting at 40/50watts it very slowly ramped up to the 300 or so of pas3. Only then did it start to respond to the throttle slowly ramping up to 650w.

All the previous pondering's about motor speed v wheel speed wouldn't seem to apply here. All plugs are in and haven't been disturbed (certainly not while I was in the shop) and the system was mostly behaving for the rest of the run, so I don't think there's an intermittent connection, at least not externally.

I am coming to the conclusion that there is something hooky about this controller. Whether it's programming, wandering component values, or damage caused by heat when running the previous motor - I don't know. But, in the face of others more knowledgeable than myself saying they have no such problem with this system and me not being able to establish reliably repeatable cause and effect, I don't know what other conclusion to draw.

I think the only option I've got now is, when I've got some spare cash, I will order a direct replacement and see how that behaves.

Are you using sealed connectors or the block type?

Block type.

I can send you this one to try. It seems to to be 36v only. It's the larger 22A size 125 x 65 x 40mm. If you want to try it, send me a PM with your address.

I've sent you a pm - I think, I've never done one before.

Got the controller this morning. Family are visiting at the moment, so I will get to it as soon as I can.