Should we bail on the new Swytch conversion kit

[jdunphy]

Hello everyone,

I am really enjoying reading and learning about e-bikes from these forums. I certainly feel that I could buy any kit and do it myself now including building a battery pack thanks to some really great posted video's from here on how to safely do this. I already maintain my own bikes so have most of the specific bike tools and do fix electronics at the component level so not afraid of a multi meter and a soldering iron. I am a retired systems programmer by trade.

My wife received her swytch kit a few weeks ago after 7 months waiting for it. It worked perfectly for 1 and 1/2 rides (approx 18 miles). She has been working with Swytch over the past 10 days to find resolution and I am not sure I buy their conclusion as I see multiple unexplained failure modes. They will ship a new handle bar battery pack assembly (see picture) but don't have it in stock and then it will be 2-4 weeks after that. We are in Victoria, BC Canada so double that time it always feels like. ;--) The problem I have is that her bike was on its kickstand for about 30-45 seconds with her off of the bike and behind it. The front wheel motor suddenly lurched forward throws the bike on the ground with enough force (wrestling pile driver comes to mind) to skew her handbars sideways. We were discussing the motor cutting in/out so I witnessed this first hand. Prior to that we saw this behavior when moving the crank by hand with no result only to see it happen 10-15 seconds later or during coasting at times when she is on it. Normally there is a small delay about 1-2 seconds which is normal for this controller which she likes.

The other symptoms are that she isn't able to keep the motor running for more than 5 mins with the average about 10-20 seconds as the motor cuts off. Sometimes it cuts back in fairly soon which is ok but a lot of time she can't get it to start again for minutes. The only thing that is reliable is that the motor will only operate when you don't need it. ;-) ;-) That means almost never completely up a hill. It works almost always down hill but can fail there too. I estimate her cadence is around 50-60 but I have tested her PAS by turning the peddle with my hand and observed the sensor picking each and every pass of the magnets. Another odd thing which we only saw once was a fully charged battery mysteriously lost 2 bars of charge in 1.5 miles on flat roads??? Fortunately, we ride mostly flat roads or 1-2 degrees gradients here and bought the swytch so she could ride with me. She loves her bike and was hesitant to go electric so it seemed like a good solution and now is hooked on the concept of a front wheel hub motor. I am not sold on her kit but her small hub motor is more than good enough for her style of riding. She never has it past assist level 1. We generally go about 22 miles per ride so her swytch eco kit could be perfect if it would function as advertised.

I have completely done and redone the PAS. I have strapped her battery + controller + BMS + display in that swytch bag so that there is no vertical play with the handle bar connector. I have cleaned the contacts. I have reset the battery pack and re-inserted it into the connector after it begins to fail and she still can't get the hub motor to engage at times. We will pull over, I will verify the PAS and then pull and re-insert the battery+controller+etc and off we go. Motor will not engage at times even after that doing that. We get better results during the first 1-2 miles of our ride where it works more often than it doesn't but can still cut out. The longer she rides during the trip, the more frequent the motor cutting in/out she says.

She has also experienced what felt like riding over a washboard as the front hub motor/wheel started to vibrate is the best way to describe it. It only happened once (very short duration) and we have probably tested this conversion kit about 7 rides now. Every day with a full charge on the battery. She hasn't had the battery below 50%. Sounds very quiet to me and I generally can't tell if the motor is operating when I ride beside her.

Short of pulling everything inside her battery case (all the components) which is where I am almost at and beginning over, I have 2 questions.

1) Do you think that replacing the handle bar connector will fix the problems as Swytch has concluded?
2) Should we cut our losses, keep the front hub wheel and order replacement parts? She has a rear rack on her bike that is empty. Recommendation for similar lightweight low range kits appreciated too. Wouldn't mind BT for the BMS for example.

I don't really care about the money (we still have 1 year warranty) but waiting for a company that took 7 months to ship the kit and have it only work for 18 miles has me thinking, enough of our time and lets make this work reliably and get back to doing what the purpose of this kit - to assist in riding a bike. Although I do enjoy the diagnostic and problem solving like the next guy, it is not as fulfilling if we are playing SWAP-A-MOLE for months and months until something eventually works. Why can't we verify it versus guess and ship one part at a time slowly over the next year. Feels like we are following some flow chart of solutions.

I'll put some dialectic grease on it next and probably do a stress test of the lateral/vertical movements of the connector assemble while in walk mode on my bike stand. I guess I could also put some cardboard shims to tightened the seating should lateral movement be causing the motor to stop/start. The big question is why doesn't it start again if it stop so readily based on the connector movement premise unless it's a wire component in the harness? Really an odd system. If anyone has some swytch work around ideas, I am eager to learn.


Thanks for any advice

Jim

 

[jdunphy]

I have started to document this kit. Might be helpful to others. This is the eco kit. See pictures of controllers and pin outs. Believe the motor could be a AKM 85sx (current guess).

Note: these are not my pictures but found them in other forums, etc.

 

[matthewslack]

Hello everyone,

I am really enjoying reading and learning about e-bikes from these forums. I certainly feel that I could buy any kit and do it myself now including building a battery pack thanks to some really great posted video's from here on how to safely do this. I already maintain my own bikes so have most of the specific bike tools and do fix electronics at the component level so not afraid of a multi meter and a soldering iron. I am a retired systems programmer by trade.

My wife received her swytch kit a few weeks ago after 7 months waiting for it. It worked perfectly for 1 and 1/2 rides (approx 18 miles). She has been working with Swytch over the past 10 days to find resolution and I am not sure I buy their conclusion as I see multiple unexplained failure modes. They will ship a new handle bar battery pack assembly (see picture) but don't have it in stock and then it will be 2-4 weeks after that. We are in Victoria, BC Canada so double that time it always feels like. ;--) The problem I have is that her bike was on its kickstand for about 30-45 seconds with her off of the bike and behind it. The front wheel motor suddenly lurched forward throws the bike on the ground with enough force (wrestling pile driver comes to mind) to skew her handbars sideways. We were discussing the motor cutting in/out so I witnessed this first hand. Prior to that we saw this behavior when moving the crank by hand with no result only to see it happen 10-15 seconds later or during coasting at times when she is on it. Normally there is a small delay about 1-2 seconds which is normal for this controller which she likes.

The other symptoms are that she isn't able to keep the motor running for more than 5 mins with the average about 10-20 seconds as the motor cuts off. Sometimes it cuts back in fairly soon which is ok but a lot of time she can't get it to start again for minutes. The only thing that is reliable is that the motor will only operate when you don't need it. ;-) ;-) That means almost never completely up a hill. It works almost always down hill but can fail there too. I estimate her cadence is around 50-60 but I have tested her PAS by turning the peddle with my hand and observed the sensor picking each and every pass of the magnets. Another odd thing which we only saw once was a fully charged battery mysteriously lost 2 bars of charge in 1.5 miles on flat roads??? Fortunately, we ride mostly flat roads or 1-2 degrees gradients here and bought the swytch so she could ride with me. She loves her bike and was hesitant to go electric so it seemed like a good solution and now is hooked on the concept of a front wheel hub motor. I am not sold on her kit but her small hub motor is more than good enough for her style of riding. She never has it past assist level 1. We generally go about 22 miles per ride so her swytch eco kit could be perfect if it would function as advertised.

I have completely done and redone the PAS. I have strapped her battery + controller + BMS + display in that swytch bag so that there is no vertical play with the handle bar connector. I have cleaned the contacts. I have reset the battery pack and re-inserted it into the connector after it begins to fail and she still can't get the hub motor to engage at times. We will pull over, I will verify the PAS and then pull and re-insert the battery+controller+etc and off we go. Motor will not engage at times even after that doing that. We get better results during the first 1-2 miles of our ride where it works more often than it doesn't but can still cut out. The longer she rides during the trip, the more frequent the motor cutting in/out she says.

She has also experienced what felt like riding over a washboard as the front hub motor/wheel started to vibrate is the best way to describe it. It only happened once (very short duration) and we have probably tested this conversion kit about 7 rides now. Every day with a full charge on the battery. She hasn't had the battery below 50%. Sounds very quiet to me and I generally can't tell if the motor is operating when I ride beside her.

Short of pulling everything inside her battery case (all the components) which is where I am almost at and beginning over, I have 2 questions.

1) Do you think that replacing the handle bar connector will fix the problems as Swytch has concluded?
2) Should we cut our losses, keep the front hub wheel and order replacement parts? She has a rear rack on her bike that is empty. Recommendation for similar lightweight low range kits appreciated too. Wouldn't mind BT for the BMS for example.

I don't really care about the money (we still have 1 year warranty) but waiting for a company that took 7 months to ship the kit and have it only work for 18 miles has me thinking, enough of our time and lets make this work reliably and get back to doing what the purpose of this kit - to assist in riding a bike. Although I do enjoy the diagnostic and problem solving like the next guy, it is not as fulfilling if we are playing SWAP-A-MOLE for months and months until something eventually works. Why can't we verify it versus guess and ship one part at a time slowly over the next year. Feels like we are following some flow chart of solutions.

I'll put some dialectic grease on it next and probably do a stress test of the lateral/vertical movements of the connector assemble while in walk mode on my bike stand. I guess I could also put some cardboard shims to tightened the seating should lateral movement be causing the motor to stop/start. The big question is why doesn't it start again if it stop so readily based on the connector movement premise unless it's a wire component in the harness? Really an odd system. If anyone has some swytch work around ideas, I am eager to learn.


Thanks for any advice

Jim

The connector is a weakness because so many contacts have to be disturbed each time the pack is removed and replaced. Must be 8 for the motor and three for the cadence sensor as a minimum? Compared to just two for a normal generic battery.

The taking off from a stand indicates multiple detected 'pulses from the cadence sensor, which if the software is unsophisticated could be a dodgy connection. Better s/w would examine the waveform for clean edges and realistic on/off periods, and signal error if outside reasonable limits.

I think packing the bag contents tightly will make it worse, not better. Now all the inertia of the relatively heavy battery is tightly bound to the connector, and every little bump will be transmitted to it, increasing the likelihood of glitches

Ideally the physical restraint and electrical connections need to be separately implemented..

So yes, I think a new connector might work, but 7 months is a long wait.

 

[jdunphy]

Thanks for your reply. Those are excellent points. It doesn't seem like the new connector solution will last if they are shipping the same part without some sort of major structural modification. We got 18 miles on a recent 2022 build; how many miles for the next/same connector? I have read about similar problems as far back as 2018 so it isn't something new.

We have also found similar reports of the bike just starting on it's own at rest. The reported work around is to power off the unit especially at more dangerous rest scenarios such as being stopped at intersections.

That the PAS could send some random voltage at rest (having been stopped for 30-45 seconds) or a controller thought it observed/registered voltage/spurious noise or even some other fault anomaly gives me a lot to think about on how to isolate this further.

I tried grease that others found was a fix for but it didn't help here. We got 1/2 block down the road and it cut out briefly and we came back. I cleaned again with isopropyl to get back to pristine clean surfaces. I believe the PAS signal line is on blade/pin 10, blade/pin 2 (hot) and blade/pin 11(ground). Still thinking on how to isolate and find a way to test whether it could be a controller problem vs a PAS problem vs other.

I pulled the battery pack apart to look for any loose connectors that might be unseated. Also verified the controller which which was built 10/2021 but otherwise identical to the picture I posted from someone that received a replacement controller for similar motor cut off problem. A casual observation looked like everything is firmly attached and they use plenty of conforming glue/coating or re-enforced the ends of tehir wires to hold things in place and provide additional mechanical stress relief.

I have not looked too closely at the part they plan on replacing - only from the outside. The few pictures I could take of it and magnify any external wires showed no evidence of stress on wires. Seems to be common enough that everyone has a few workarounds that sort of help for some. As for being at rest and motor starting, It also plausible that something has triggered the controllers walk assist mode which doesn't require PAS signalling and only needs a signal or spurious signal to initiate?

Update: We queried Swytch support again Apr 25, 2025 and they have stated that the new connector will fix both types of faults (motor cutting on/off and suddenly starting at rest without PAS on a kickstand).

A few more pictures from our kit (eco) to help others identify parts/cables that might be needed in the future.

 

[Nixtoo]

Those images are really useful, thanks. I’m waiting for an eco kit to arrive for my son’s bike. But they’re messing me about with delivery dates. I was sold on the light weight motor but am starting to question the quality of the kit from what I’ve read!

 

[jdunphy]

I have read a lot of good things too but I would guess when something doesn't work, the owners frustration can generate more noise in contrast. I have had a lot of that recently! ;-) ;-)

What I didn't say is this:

The instructions are super clear (colour pictures in 65 page instruction booklet, how to videos, a large user community, packed really well (Apple would be proud), and there is Sywtch support to help should you run into any problems during your install + it's warranted for a year (they will ship you parts at no cost). So very high probability that any bike will be an e-bike with the kit once your package arrives.

I don't know enough what is good or bad quality out there other than the stuff I read from Grin Technologies but my wife's use is different from some of their customers that use there ebikes for 5000mile treks over the Rocky Mountains without having the motor burn upon climbs. My wife wanted some assist but not at the expense of adding too much weight to her bike as exercise is still her primary reason for riding. We tend to ride around 15-17 mph on level ground and she keeps it at level 1 assist... but that extra boost does make a big difference for her. She gets to keep her own bike and its gearing and picking a solution for her was fairly easy given the marketing done by Swytch.

I am no expert but it appears they took some off the shelf parts and packaged them so that most if not all bikes can be converted with a universal kit. Added great instructions, how to videos, and support to make sure everyone that buys has success. For example, not having to pull the crank to install the PAS eliminated any specialty tools some bikes might require.

One of the things I am toying with is purchasing something like Cycle Analyst V3 from Grin, plugging all the devices into that and have it manage the motor controller. That should allow us to add features like cruise control, better assist modes for her cadence, keep parts and eventually bypass and replace things that are out of warranty or update when our needs change. Still in the research phase on this. That would allow us to work around the weaknesses of the product like the harness should it fail and put the pieces where it makes the most sense for her bike. Once you open that battery bag, you can unconnect all the parts and use them how or where you want with cable extenders, racks, bags, etc.

 

[harrys]

I believe the Cycle Analyst sends a voltage into the throttle input of your controller. One of the things it can do is read a torque sensor driven off your pedals to do that. Does the Swytch have a throttle?

Two suggestions. I believe the motor has a standard 9 pin Bafang cable?



You could buy a battery with a built in controller. Add the display, brake levers, throttle, PAS sensor, etc. About $375 before shipping. Might need an extension motor cord. So you put the battery on the water bottle mounts. Add the controls, Plug into motor,

2022 Polly Battery DP-4C 36V15AH Samsung- 21700 50E Li-ion Battery Pack for Ebike [Polly DP4C 36V15AH Samsung 21700] - $232.96 : Zen Cart!, The Art of E-commerce

Zen Cart! 2022 Polly Battery DP-4C 36V15AH Samsung- 21700 50E Li-ion Battery Pack for Ebike [Polly DP4C 36V15AH Samsung 21700] - ★★★★★ 2022 Polly Battery DP-4C 36V15AH Samsung-21700 50E Li-ion Battery Pack for Ebike SPECIFICATION: 1-Normal voltage:36V 2-Capacity:20AH 3-Cells in series:10S3P...

www.topbikekit.com

A second approach is to look at the hilltopper kits. They are similar as the above, except you have to buy a motor, and they ship out of North AMerica.

 

[jdunphy]

I believe the Cycle Analyst sends a voltage into the throttle input of your controller. One of the things it can do is read a torque sensor driven off your pedals to do that. Does the Swytch have a throttle?

I haven't done enough reading on Cycle Analyst at this point but the swytch controller does support a throttle which I installed last week to allow my wife to have access to the motor in a more predictable fashion. They did swap the signal and ground with their cabling coming out of the connector so off the shelf can be a problem unless you do this: ref: https://www.pedelecs.co.uk/forum/threads/2-swytch-conversions-and-a-kit-from-amazon.43224/page-2#post-650669

Two suggestions. I believe the motor has a standard 9 pin Bafang cable?

View attachment 46809

You could buy a battery with a built in controller. Add the display, brake levers, throttle, PAS sensor, etc. About $375 before shipping. Might need an extension motor cord. So you put the battery on the water bottle mounts. Add the controls, Plug into motor,

2022 Polly Battery DP-4C 36V15AH Samsung- 21700 50E Li-ion Battery Pack for Ebike [Polly DP4C 36V15AH Samsung 21700] - $232.96 : Zen Cart!, The Art of E-commerce

Zen Cart! 2022 Polly Battery DP-4C 36V15AH Samsung- 21700 50E Li-ion Battery Pack for Ebike [Polly DP4C 36V15AH Samsung 21700] - ★★★★★ 2022 Polly Battery DP-4C 36V15AH Samsung-21700 50E Li-ion Battery Pack for Ebike SPECIFICATION: 1-Normal voltage:36V 2-Capacity:20AH 3-Cells in series:10S3P...

www.topbikekit.com

A second approach is to look at the hilltopper kits. They are similar as the above, except you have to buy a motor, and they ship out of North AMerica.

Thank you, Those are excellent options also and I will look into that further. That would be an excellent method to unwind any proprietary decisions that Swytch may have created. Currently the hub motor is working perfectly with the throttle so the PAS/harness seems to be the best guess of it's random behavior. Adding additional dampening to the mount makes it happen less often in addition to switching the angle the pack is mounted. Not enough data to recommend this yet. We did a 22 mile trip on Friday and it behaved very well.

 

[afia2k8]

@Nealh

The new Swytch pocket size kit is being released tomorrow. I've got an early bird link for 60% off. Do you think if the price is around £300 it will be worth buying?

 

[Nixtoo]

@Nealh

The new Swytch pocket size kit is being released tomorrow. I've got an early bird link for 60% off. Do you think if the price is around £300 it will be worth buying?

The £300 kit only has a 2.5ah battery

 

[StuartsProjects]

@Nealh

The new Swytch pocket size kit is being released tomorrow. I've got an early bird link for 60% off. Do you think if the price is around £300 it will be worth buying?

The small battery would only be of (limited) value for someone who is really light and wants to go on short flat rides.

One reason they may have made such a small battery is so that the can claim 'the lightest electric bike conversion' or similar.

 

[matthewslack]

I was expecting the 700g pocket sized battery to be 10s1p 21700 cells, so small but not 2.5Ah small. Perhaps it is 18650s and only 2.5Ah...

Who will be first to look inside and post a picture ?

 

[Nealh]

At 2500mah then the cells will be 18650, 90wh isn't going to get you more then 6 - 7 miles depending on the current draw. 2500mah is the rating but for sure one won't see much more then 1750 - 2000mah form a 1p cell and it would have to be a top cell for 1p none of your flaky mediocre 5a cells, it would need to be a 20 or 25a cell of which there are only 4 or 5 that would fit the bill.

£300 for a kit sounds good but one is going to need to buy a battery still as 2.5ah isn't going to cut it. The least one can get away with is 2p of top notch cells in 18650 form and that means a 3000mah cell at 15/25a.

 

[Nealh]

@Nealh

The new Swytch pocket size kit is being released tomorrow. I've got an early bird link for 60% off. Do you think if the price is around £300 it will be worth buying?

Only if you like pushing the bike or it can ride ok with no power, range will be poor less then 6 or 7 miles with power so ok to pootle around very local.
To know exactly one would need to know the cell specs.

 

[Nixtoo]

This is the info they sent me the AIR (2.6Ah, 15km range, 90Wh) and the MAX (5Ah, 50km range, 180Wh.) Both motor wheels are 250W and 36V. I'm currently recieving terrible customer service and will probably request a refund by the end of the week!

 

[matthewslack]

This is the info they sent me the AIR (2.6Ah, 15km range, 90Wh) and the MAX (5Ah, 50km range, 180Wh.) Both motor wheels are 250W and 36V. I'm currently recieving terrible customer service and will probably request a refund by the end of the week!

At least the range claim of 15km is realistically achievable on 90Wh. Substantially limiting though for riders who use assist all the time, but at the same time for riders who only want help up the hills it might do very nicely.

I await reviews with interest!

 

[Nealh]

50km form 5ah is pie in the sky talk unless one plans on using no assist.
They are known for terrible CS, I think previously they were called Panda Bikes.

They are using simple maths to come by meaningless ranges. 90 wh / 10wh per mile = 9 miles. One needs to be a good regular cyclist to use 10wh or less per mile and at the lowest power level. Bing in some hills and one will be walking.

 

[Nealh]

The 90wh though is at 0a or 0.2a current draw, once one starts to draw current even a lowly 2a the wh usage in real world terms is a lot less. Until the spec of the battery cell is known then we can only speculate.

 

[Nixtoo]

They also said this 'The cells inside that are 5000 mAh capacity'

 

[matthewslack]

They also said this 'The cells inside that are 5000 mAh capacity'

Just as I thought it had become clear!

5000mAh cells are going to be 21700, and 10 of those would give 36V 5Ah = 180Wh. I.e the MAX. The AIR's 90Wh is either 18650 cells or (rather unlikely) fewer than 10 21700s and a DC-DC converter...