PEEK Gear for the Ultra Needed

For what they're charging for their bikes, they better be! 😁
ha ha.. yeah. I had been following up with their battery and charging unit providers for our bikes. It didnt work out but got a peek into what high quality production looks like. Very well done. They do have some glithches on the software side though, that is typically an afterthought, but manufacturing is quite good.

Onwards :)
 
So I spoke with a couple of you guys here 1-1. Thanks @brake034.

I wanted to follow with a slightly deeper background based on some testing we did a few months ago (and continue to do). Please note that these issues apply to all motors - not just Bafang.

1. The way the gears sit in a mid drive e-bike motor (typically) is based on the motor casing and positioning of the stator / core etc.
2. The tolerance for these motor casings is low. Across different production batches, you may see ±0.25-0.5mm change. That is significant.
3. All machined gears are also made with loose tolerance - sometimes deviation is as high ±0.1-0.3mm

Note: For this discussion I am ignoring the thermal sink issues (motor heating , lack of effective cooling).

With loose tolerance, riders will notice the following
1. Louder noises - as gears rotate, the higher the effective gap, the louder the noise. I have tried Bafang and bosch Gen 4- both motors have the "noise".
2. As you ramp up to higher speeds (and higher RPM), because the Ultra is all steel internally, there is a rapid increase in temp due to these tolerance issues. The motor doesnt get too hot because it has a fairly large heat sink. This heat issue is present in Bosch though- if you go turbo, thermals are comparable to the bafang m600 - too much power in a smaller enclosure.
3. If we put 30 amps through bosch / yamaha / etc, the gears will absolutely start to fail (IIRC most are actually nylon!)

With that background, a Single PEEK gear for the bafang ultra IMO is not a sufficient solution. It may work for lower power motors (say 300-500W nominal) , but with higher wattage, high manufacturing tolerances, we should absolutely expect a higher failure rate with PEEK.

So why are the motors made with loose tolerances?
1. Cost. tightening the tolerance adds (anecdotally) 25-50% increase in production cost.
2. Manufactuers may also need to re-do molds and tooling. It is a one time cost but significant enough investment upfront.

Which are the quiet motors then ?
1. Brose and Yamaha motors appear to have better tolerances than Bosch, Shimano, Bafang (Anecdotally).
2. Lower wattage will allow any manufacturer to do all PEEK without addressing the tolerance issues.

The ideal way to tackle noise / weight / heat is to take a holistic approach
1. Better tolerances on casing.
2. Better tolerances on gears.
3. PEEK for ALL gears, not just the primary gear.

This allows for reduced noise, better gear inter-locking, and noticeable reduction in heat generation. This may not be the answer everyone is looking for, and we may have short term success with a single PEEK gear . However it is not an effective long term /scalable solution. The problem is much deeper, and it is not limited to Bafang. All ebike motors have (in general) high tolerances, most just don't run into it because they have lower power motors.


If we want to look at mid-drives with nice motors, look at the new Zero motors drive train - IIRC, they actually reduced the motor shell size by a little in the new versions, as their quality has improved. Also, TDCM (the hub motors on stromer) - the higher end versions - very very tight tolerances. OOh.. my fav is Pinion gear box. I know its a gear box- but it is beautiful. It helps to have a fluid bath :)
I appreciate the really in depth analysis.....here is the comment from another thread when I inquired if anybody out there was still running the original OEM Nylon gear from when Bafang first came out with the Ultra.

"I still have the nylon gears in my 2017 Biktrix Ultra. No problems and dead quiet. If it fails, I have the metals gears on the shelf. Roshan sent them to me free of charge because he loves me and wants me to be happy !"

I hear comments like that and it makes me think we could at least try an exact copy of that mechanism with the one larger gear in PEEK the one that used to be Nylon.
It was Luna that convinced Bafang to switch their nylon gears to steel, they had a picture of 8 dead motors they sent them with shredded gears, problem was they were tweaking everything to the max with those Ebikes.
I am not surprised they failed, the limitations on that gear had not been pushed by Bafang when they brought that to market, there are so many variables, maybe it was fine as long as it stayed cold, so winter riding on fat Ebike would be ok, but riding in 100 degree heat would be a death sentence for it. As the heat would weaken it's strength making it shear.
 
Last edited:
E-Mountianbike had this to say about current motors from all brands.

"What are motor manufacturers doing wrong? The new Shimano EP8 also makes a metallic knocking noise when rolling over obstacles, just like the Bosch Performance Line CX motor introduced last year. All the ends bike manufacturers are going to to make bikes quieter seem to be in vain."

The point is whoever can bring this to market on the ultra will have achieved something all the big players have not, a very powerful motor with a throttle that is quieter, and many companies will want in on this.
 
There still appears to be a lot of stigma associated with “bafang”. By default a lot of people associate it with “lower quality”. There is always an argument to be made that the motor is so bad that it needs custom controller and quiet gears to compete. 😂

It is true that there is a premium performance space , but it adds significant cost that IMO few are willing to pay. Back of the envelope, it’s cost of motor+controller + gears = $14-1500 for just the motor. That’s fairly steep IMO. The better way is to just build it right from the ground up with the right parts, and target the 500-700 sweet spot.
 
There still appears to be a lot of stigma associated with “bafang”. By default a lot of people associate it with “lower quality”. There is always an argument to be made that the motor is so bad that it needs custom controller and quiet gears to compete. 😂
You can find Youtubers saying the same about Ferraris and Lamborghinis. The brand new car that cost half a million needs to be mapped, new exhaust, new wheels etc :p
 
E-Mountianbike had this to say about current motors from all brands.

"What are motor manufacturers doing wrong? The new Shimano EP8 also makes a metallic knocking noise when rolling over obstacles, just like the Bosch Performance Line CX motor introduced last year. All the ends bike manufacturers are going to to make bikes quieter seem to be in vain."

The point is whoever can bring this to market on the ultra will have achieved something all the big players have not, a very powerful motor with a throttle that is quieter, and many companies will want in on this.
Yes. Completely agree.
 
E-Mountianbike had this to say about current motors from all brands.

"What are motor manufacturers doing wrong? The new Shimano EP8 also makes a metallic knocking noise when rolling over obstacles, just like the Bosch Performance Line CX motor introduced last year. All the ends bike manufacturers are going to to make bikes quieter seem to be in vain."

The point is whoever can bring this to market on the ultra will have achieved something all the big players have not, a very powerful motor with a throttle that is quieter, and many companies will want in on this.
Spot on.
 
There still appears to be a lot of stigma associated with “bafang”. By default a lot of people associate it with “lower quality”. There is always an argument to be made that the motor is so bad that it needs custom controller and quiet gears to compete. 😂

It is true that there is a premium performance space , but it adds significant cost that IMO few are willing to pay. Back of the envelope, it’s cost of motor+controller + gears = $14-1500 for just the motor. That’s fairly steep IMO. The better way is to just build it right from the ground up with the right parts, and target the 500-700 sweet spot.
Public opinion: Generally derived from a lot of loose lips blabbing what they themselves have no understanding of. 🤔
 
There still appears to be a lot of stigma associated with “bafang”. By default a lot of people associate it with “lower quality”. There is always an argument to be made that the motor is so bad that it needs custom controller and quiet gears to compete. 😂

It is true that there is a premium performance space , but it adds significant cost that IMO few are willing to pay. Back of the envelope, it’s cost of motor+controller + gears = $14-1500 for just the motor. That’s fairly steep IMO. The better way is to just build it right from the ground up with the right parts, and target the 500-700 sweet spot.
Interesting comment with the top controller already in hand are you actually consider building you own motor in house ?
 
Interesting comment with the top controller already in hand are you actually consider building you own motor in house ?

Long term, we need our own for sure. No doubt.

Current manufacturers move too slowly and design for 250-400W first. We have looked at atleast 20-30 different variations from various manufacturers from China, Taiwan, Germany, Portugal, Italy, etc etc. Most motors are commodities - like literally I spoke to the bofelli and fazua OE guys.

The development is not trivial, but given that we have figured out the controller part, we can actually apply this to pretty much any motor out there. Ideally we dont need to develop everythign from scratch, and take an "open" design and update it. We have a couple of promising candidates - but it is a super long term project 24months. Motor at small scale is no problem, but scaling means we need to consider partnerships, OE vendors etc.
 
Long term, we need our own for sure. No doubt.

Current manufacturers move too slowly and design for 250-400W first. We have looked at atleast 20-30 different variations from various manufacturers from China, Taiwan, Germany, Portugal, Italy, etc etc. Most motors are commodities - like literally I spoke to the bofelli and fazua OE guys.

The development is not trivial, but given that we have figured out the controller part, we can actually apply this to pretty much any motor out there. Ideally we dont need to develop everythign from scratch, and take an "open" design and update it. We have a couple of promising candidates - but it is a super long term project 24months. Motor at small scale is no problem, but scaling means we need to consider partnerships, OE vendors etc.
That is really impressive, I am sure you will come up with something awesome, would you be looking at a patent for only Watt Wagon, or would you make it so you could sell to anybody to fit on most Ebikes.

On another note the rider with the nylon gear, removed his throttle on his Ebike, I am pretty sure this is what contributed to the survival of his nylon gear all this time.
 
Thanks. Lot of future work, balancing with what we have to roll out now and rest of the year :) Pushing the boundaries of what is possible within our limited $$ pool, basically. :)

Yeah, throttle is a big issue with mid drives. I would love to know if there is a better way to message this ?
 
Long term, we need our own for sure. No doubt.

We have a couple of promising candidates - but it is a super long term project 24months. Motor at small scale is no problem, but scaling means we need to consider partnerships, OE vendors etc.
I hope that the design of the motor casing/housing would make it retrofit into the current Helios/Hydra frames! And maybe halve the weight....
 
I hope that the design of the motor casing/housing would make it retrofit into the current Helios/Hydra frames! And maybe halve the weight....
Retrofit is definitely the goal.

No point in trying to come up with a new mount design - the 3 bolt version is the simplest I have seen.

Weight will not be halved tbh - we can shave off a little but not more than 1lb total.
 
Revisiting this topic. I’m looking for some guidance on gears and motors.

Is there an expert here I could reach out to for some technical questions? We have a couple of questions about tolerances and heat dissipation for PEEK vs Metal.
 
Created a new thread here - in case people want to try stuff.
 
All,

Quick update on the PEEK gear.

1. We have experienced minor to catastrophic failures of the gear across all our test gears. We have experimented with all variants -

- PEEK, - initial failure at around 650 miles. Complete failure at around 1000 miles.
- PEEK with a different carbon polymer material - complete failure at 500 miles. This is unusual- were hoping to be better than PEEK, so this may also be a manufacturing defect. Regardless, it failed.
- Carbon infused PEEK. Initial failure at 1350-1400 miles. Complete failure at 2000 miles. <-- this was what we were hopeful for, and had ordered. Unfortunately it failed as well.
2. All gears were on motors tuned to Archon X1 @1800 Watts (around 180Nm of torque)

As of right now, we DO NOT recommend using PEEK in the bafang ultra, or any motor over 500W nominal / 100Nm of torque. The only realistic way of reducing noise is to have better tolerance on the motor. We are working on that and will post progress over the coming months.
 
Back