Bosch CX gen 4 chainring nut

Timmoh

Well-known member
May 18, 2020
248
217
Wales
I come back to the topic
TECH NERD ALERT: very long and boring post ?

So, the lockring is threaded ccw. We must take into consideration that these components are designed in such a way that their own inertia acts in favor of tightening.

Let me explain and try to analyze the phenomenon.

Working with such a component, the designer determines the direction of the thread on the basis of the direction of the main stress, which will generate a reaction of the component itself (given by its own inertia) in the opposite direction.
In a system with a spider and its lockring, if the system itself forces the component to turn clockwise, its own inertia will generate an opposite, counterclockwise reaction.
The spider (of course) turns clockwise when pedaling.
I can assume (and I want to clarify it, mine are assumptions as a "nerd" designer) that the Bosch designers have taken into account the accelerations impressed by the rider (raises, starts, etc ...) to the chainring via the pedals, action reinforced by the motor also. Then we have the motor shaft that lets the spider move in a CLOCKWISE direction, dragging the lockring with it, making it snap in the same direction; the lockring is therefore forced to sudden start rotating clockwise. Its inertial reaction will generate an anti-clockwise force (i.e. the inertia of the lockring will tend to oppose the movement that was imposed by the spider, making it rotate counterclockwise). All this is represented in the image below.
Blue disk = spider + chainring
Orange disk = lockring
Blue arrow = rotation force impressed by the spider
Orange arrow = lockring intertial reaction

View attachment 33692

With this consideration, I agree with the Bosch designers: the ccw thread is a correct choice because its inertia works to tighten the lockring threaded connection with the spider.

Off-road, however, there is another stress that can become much more predominant: the kick-back generated by the rear suspension.
The stresses that comes from the rear swingarm in a high speed descent on rockgarden or woods create a reverse chain pull force that pulls the chainring back. These stresses are very impulsive, therefore they create a very strong acceleration (in short, a sudden jerk) that works in the OPPOSITE direction given by pedaling action. The consequence is that the inertia of the lockring works this time in reverse, leading it to unscrew.
It makes me believe (I repeat, they are my assumptions) that these stresses are more violent and above all much more frequent in AM, Freeride, Enduro, etc ... compared to those of pedaling (in Ednuro, AM, freeride, etc .. we're used to pedal quiet till the top then descend "full throttle"). We will therefore have a lockring exposed to a reverse stress, the inertia of which will then push it clockwise, thus tending to unscrew it. Jerk, after jerk, after jerk, the lockring becomes loose.

Having said that, were the Bosch designers wrong? NO. They simply imagined that most of the users with Gen4 will do easy and little off-road rides; in this case, I consider the choice correct.

Based on this, I tried to invent a solution to substantially prevent the lockring from moving clockwise because of the kickback
First, a small 2mm hole

View attachment 33693

... then a tap of a grinding wheel to create a small "pocket" on the inner face

View attachment 33694
View attachment 33695

... and then a nice band of safety iron wire. And since security is never too much, even a bit of threadlocker.

View attachment 33696

Now I just have to screw the ring nut again and practice the safety tie using one of the spider spokes. In this way the lockring is prevented to rotate cw (the one that causes the ring nut to be unscrewed) by the wire. It is the same system that I used on the motorbike on the track, where oil caps, brake screws and other components that could be unscrewed had to have the safety tie for regulation.

View attachment 33697
View attachment 33698
View attachment 33700

Unfortunately, job agenda and bad weather do not allow me to perform tests
I'll come back to you in the next weeks with the first feedbacks from the field
Just about to get one of these motors and will do this. It takes me back my days working in Aerospace-wirelocked anything that moved, and if its good enough for things we fly in with our Famlies, that's good enough for me. Plus, I like the look of it, and it adds a bit of Pro looking fettling to the ride.
Great little fix and if you do manage to break it/get it to fail, they will probably be scraping you off the scenery anyway due to the speed/abuse/height you were giving it just prior ??.
 

b45her

Member
Dec 1, 2019
94
87
wales
has no one read any bike manual ever?

the first thing they all say is to check nuts and bolts regularly, every bike i have ever owned (and there have been a lot) has had nuts and bolts work loose over time. spend 10 minute every couple of weeks with a few basic tools and problem solved.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
has no one read any bike manual ever?

the first thing they all say is to check nuts and bolts regularly, every bike i have ever owned (and there have been a lot) has had nuts and bolts work loose over time. spend 10 minute every couple of weeks with a few basic tools and problem solved.
The point you make is valid, but not 100% applicable in this case. The lockring tightens counterclockwise, and can only be tightened to 30nm, so its always going to come loose without thread lock compound, which is not applied when the bike is assembled. It also requires a special tool, which most people don't have, and have had to buy especially for this purpose. I am lucky in that I read about the issue while I was waiting for my bike, otherwise I would have had no idea, and the lockring would have come loose on a ride, which has been most people's experience
 

Timmoh

Well-known member
May 18, 2020
248
217
Wales
has no one read any bike manual ever?

the first thing they all say is to check nuts and bolts regularly, every bike i have ever owned (and there have been a lot) has had nuts and bolts work loose over time. spend 10 minute every couple of weeks with a few basic tools and problem solved.
I popped a spot of paint (model enamel) on both surfaces to make for an easy visual check pre ride (and on the ride).
Always have done a cursory tighten and check every ride, it's an old school habit from when there was no suspension and things just came loose after every ride anyway back in the day ?.
 

HORSPWR

E*POWAH Master
May 23, 2019
853
680
Alice Springs, Australia
Bosch specifies an o-ring between the chainring and chainring nut....bosch part number 1270016119. It seems to be for one-time use. I removed and replaced the chainring and now the o-ring wants to bunch up and squirm out of place. There is no groove for the o-ring. But the technical support line from Bosch says it must be there. I'm not impressed with this design, nor the fact that the nut tightens in an opposite direction of pedaling, nor the fact that the chainring nut is aluminum and seems delicate (there will be aluminum shavings as you remove and replace the chainring nut). I have heard that some shops in the USA are using loctite, and one mechanic admitted that they were not using the o-ring. Trek does not have the o-rings, and other than some European online sites, I can't find the o-rings in the USA.
Pretty sure left hand thread is exactly what it should be, you force the cranks clockwise and the load forces the chainring and chain counter-clockwise (because of the resistance of the rear wheel against the ground) and hence this action tries to tighten the locknut. My guess is there isn't enough NM torque and vibration and lateral movement make it come loose.
 

Timmoh

Well-known member
May 18, 2020
248
217
Wales
Pretty sure left hand thread is exactly what it should be, you force the cranks clockwise and the load forces the chainring and chain counter-clockwise (because of the resistance of the rear wheel against the ground) and hence this action tries to tighten the locknut. My guess is there isn't enough NM torque and vibration and lateral movement make it come loose.
Just watched Le Mans 66 and it got me thinking. The cars use centre wheel spinner nut to secure (older racing cars and Cobras/Ferrari etc. Did a bit of looking on a few sites and found this on a site for people with posh cars;
"On Ferraris with knock-off wheels, the spinners on the right side of the car tighten counterclockwise. On the left side of a car, they tighten clockwise. The same is true for Ferraris with a single, center nut on each wheel. This is intended to prevent the spinners or center nuts from loosening".
They go on to explain that bicycle pedals are a little different due to the pedal/crank directions. We are only concerned with the direction on the right side crank and Motor spindle Union, so the Ferrari explanation (?) fits.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
Just watched Le Mans 66 and it got me thinking. The cars use centre wheel spinner nut to secure (older racing cars and Cobras/Ferrari etc. Did a bit of looking on a few sites and found this on a site for people with posh cars;
"On Ferraris with knock-off wheels, the spinners on the right side of the car tighten counterclockwise. On the left side of a car, they tighten clockwise. The same is true for Ferraris with a single, center nut on each wheel. This is intended to prevent the spinners or center nuts from loosening".
They go on to explain that bicycle pedals are a little different due to the pedal/crank directions. We are only concerned with the direction on the right side crank and Motor spindle Union, so the Ferrari explanation (?) fits.
I think Dr Ergal explains it best. But whatever the engineering theory behind Bosch's design, the fact remains that the chainring lockring comes undone , in some instances during people's first ride. So maybe the design is sound, but poorly executed? If thread lock was applied when the bikes were assembled, there would be no issue
 

HORSPWR

E*POWAH Master
May 23, 2019
853
680
Alice Springs, Australia
Just watched Le Mans 66 and it got me thinking. The cars use centre wheel spinner nut to secure (older racing cars and Cobras/Ferrari etc. Did a bit of looking on a few sites and found this on a site for people with posh cars;
"On Ferraris with knock-off wheels, the spinners on the right side of the car tighten counterclockwise. On the left side of a car, they tighten clockwise. The same is true for Ferraris with a single, center nut on each wheel. This is intended to prevent the spinners or center nuts from loosening".
They go on to explain that bicycle pedals are a little different due to the pedal/crank directions. We are only concerned with the direction on the right side crank and Motor spindle Union, so the Ferrari explanation (?) fits.

The picture below probably explains it best, F1 is the force applied to the pedal and F2 is the force acting on the chainring, F3 is the force acting on the rear wheel sprocket. This is why the Bosch chainring nut is a left hand thread to tighten counter clockwise, Bosch would have been better to have done away with the locknut and actually have the spider itself left hand threaded to spin on counter clockwise so its always tightening under load.
bicycle_physics_4.png
 

Dr. Ergal

Active member
Mar 4, 2020
46
73
Italy
It conferms my thoughts: Bosche engineers made the best move to design the nut left hand, so that the normal direction of the chainring (clockwise) and the related solicitation could naturally tend to tighten the nut

So, engineers Bosch were right, and that's what I wrote in my post

But I also pointed out another fact.
I wrote that the solicitation coming from another phenomenom (chain kick-back given by the chainstay movement in hard and fast descents) is the cause of the loosening, because it creates a force having opposite force direction into the chain, and to the chainring.

Pedal action: nut inertia tends to tighten itself
Kick-back action: nut inertia tends to loosen itself

I'm pretty sure that Bosch engineers thought about it as well. Most probably, they counted how many city, cargo, XC and "easy" bike would have used their motor in comparison with the number of enduro bikes. Enduro bikes (afftected by the kick-back phenomenom) are surely the minority, so they probably thought that enduro issues are not so relevant in the entire Gen4 scenario.

And I agree with them. I would have made the same thought

Also... I agree with Philly: if a proper amount of threadlocker were applied on manufacturing, this issue would ne absent
But... how vintage and cool wirelocking is...? :cool: ?
 
Last edited:

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
It conferms my thoughts: Bosche engineers made the best move to design the nut left hand, so that the normal direction of the chainring (clockwise) and the related solicitation could naturally tend to tighten the nut

So, engineers Bosch were right, and that's what I wrote in my post

But I also pointed out another fact.
I wrote that the solicitation coming from another phenomenom (chain kick-back given by the chainstay movement in hard and fast descents) is the cause of the loosening, because it creates a force having opposite force direction into the chain, and to the chainring.

Pedal action: nut inertia tends to tighten itself
Kick-back action: nut inertia tends to loosen itself

I'm pretty sure that Bosch engineers thought about it as well. Most probably, they counted how many city, cargo, XC and "easy" bike would have used their motor in comparison with the number of enduro bikes. Enduro bikes (afftected by the kick-back phenomenom) are surely the minority, so they probably thought that enduro issues are not so relevant in the entire Gen4 scenario.

And I agree with them. I would have made the same thought

Also... I agree with Philly: if a proper amount of threadlocker were applied on manufacturing, this issue would ne absent
But... how vintage and cool wirelocking is...? :cool: ?
Dr. Ergal, your wirelocking is rad! ? but beyond my meagre skills - it's threadlocker for me! :D
 

Shy Ted

Member
Aug 20, 2019
95
76
Inbed
All this theorising is great, lovely mechanical based explanations but in the real world, it’s tosh!
my spider nut came undone on today’s ride, I tightened it up with the aid of a rock and a tyre lever, it lasted about half a mile before it came undone again. On a 15 mile off-road ride I had to stop at least six times to tighten the bugger up again, and I mean really tightened it, but no amount of forward pedalling helped to tighten it. I’m hoping the loctite fix fixes it, because that’s what I did when I got back from the ride. Just a bad design by Bosch in my opinion.
 

HORSPWR

E*POWAH Master
May 23, 2019
853
680
Alice Springs, Australia
All this theorising is great, lovely mechanical based explanations but in the real world, it’s tosh!
my spider nut came undone on today’s ride, I tightened it up with the aid of a rock and a tyre lever, it lasted about half a mile before it came undone again. On a 15 mile off-road ride I had to stop at least six times to tighten the bugger up again, and I mean really tightened it, but no amount of forward pedalling helped to tighten it. I’m hoping the loctite fix fixes it, because that’s what I did when I got back from the ride. Just a bad design by Bosch in my opinion.
Yes I agree, its a very bad design enhanced by the fact that the spider can't rotate far enough counter clockwise to actually enact enough force to tighten the lock nut.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
All this theorising is great, lovely mechanical based explanations but in the real world, it’s tosh
The purpose of the theorising has been to understand why Bosch designed it like that. I have learned from reading these posts that there is a sound engineering principle behind the design. However, we can all agree that, as you say, in a real world off-road MTB application, the design is flawed. Loctite and torquing to 30nm seems to do the trick though, so far so good. Also ditched the ridiculous o-ring, can't understand what purpose that was meant to serve.
 

HORSPWR

E*POWAH Master
May 23, 2019
853
680
Alice Springs, Australia
I come back to the topic
TECH NERD ALERT: very long and boring post ?

So, the lockring is threaded ccw. We must take into consideration that these components are designed in such a way that their own inertia acts in favor of tightening.

Let me explain and try to analyze the phenomenon.

Working with such a component, the designer determines the direction of the thread on the basis of the direction of the main stress, which will generate a reaction of the component itself (given by its own inertia) in the opposite direction.
In a system with a spider and its lockring, if the system itself forces the component to turn clockwise, its own inertia will generate an opposite, counterclockwise reaction.
The spider (of course) turns clockwise when pedaling.
I can assume (and I want to clarify it, mine are assumptions as a "nerd" designer) that the Bosch designers have taken into account the accelerations impressed by the rider (raises, starts, etc ...) to the chainring via the pedals, action reinforced by the motor also. Then we have the motor shaft that lets the spider move in a CLOCKWISE direction, dragging the lockring with it, making it snap in the same direction; the lockring is therefore forced to sudden start rotating clockwise. Its inertial reaction will generate an anti-clockwise force (i.e. the inertia of the lockring will tend to oppose the movement that was imposed by the spider, making it rotate counterclockwise). All this is represented in the image below.
Blue disk = spider + chainring
Orange disk = lockring
Blue arrow = rotation force impressed by the spider
Orange arrow = lockring intertial reaction

View attachment 33692

With this consideration, I agree with the Bosch designers: the ccw thread is a correct choice because its inertia works to tighten the lockring threaded connection with the spider.

Off-road, however, there is another stress that can become much more predominant: the kick-back generated by the rear suspension.
The stresses that comes from the rear swingarm in a high speed descent on rockgarden or woods create a reverse chain pull force that pulls the chainring back. These stresses are very impulsive, therefore they create a very strong acceleration (in short, a sudden jerk) that works in the OPPOSITE direction given by pedaling action. The consequence is that the inertia of the lockring works this time in reverse, leading it to unscrew.
It makes me believe (I repeat, they are my assumptions) that these stresses are more violent and above all much more frequent in AM, Freeride, Enduro, etc ... compared to those of pedaling (in Ednuro, AM, freeride, etc .. we're used to pedal quiet till the top then descend "full throttle"). We will therefore have a lockring exposed to a reverse stress, the inertia of which will then push it clockwise, thus tending to unscrew it. Jerk, after jerk, after jerk, the lockring becomes loose.

Having said that, were the Bosch designers wrong? NO. They simply imagined that most of the users with Gen4 will do easy and little off-road rides; in this case, I consider the choice correct.

Based on this, I tried to invent a solution to substantially prevent the lockring from moving clockwise because of the kickback
First, a small 2mm hole

View attachment 33693

... then a tap of a grinding wheel to create a small "pocket" on the inner face

View attachment 33694
View attachment 33695

... and then a nice band of safety iron wire. And since security is never too much, even a bit of threadlocker.

View attachment 33696

Now I just have to screw the ring nut again and practice the safety tie using one of the spider spokes. In this way the lockring is prevented to rotate cw (the one that causes the ring nut to be unscrewed) by the wire. It is the same system that I used on the motorbike on the track, where oil caps, brake screws and other components that could be unscrewed had to have the safety tie for regulation.

View attachment 33697
View attachment 33698
View attachment 33700

Unfortunately, job agenda and bad weather do not allow me to perform tests
I'll come back to you in the next weeks with the first feedbacks from the field
Nice work, that'll fix it. I perhaps would've gone a tad simpler. I would've drilled and tapped a small hole (3mm) in the spider itself in between one of the lock nut cutouts and then threaded a socket head cap screw into it so the head of the cap screw sits in the lock nut cut out, preventing it from rotating in either direction unless the cap screw is removed first. Just looks like it would be a cleaner look rather than lock wire.
 

Mike a769

Member
Sep 30, 2020
7
0
Norcal
Chainring nut / spider nut or other names...
Has anyone experienced their chainring nut coming loose?
Today is the first day my Rail has let me down, I had to cut my ride short today when I felt a vibration at my feet, checked the chainring and it was wobbling and the spider nut had come completely undone... anyone experienced this?
I made myself a tool to tighten it back on today and will check it again next time out. Has anyone else had this happen on a Gen 4 motor or van I put it down to a build issue? View attachment 27677
Same Thing Happened on the same bike crank arm as well came loose Evidently the same Loose tightening procedure
Mine also has loose nut behind the steering tube
 

dixie600mhz

Active member
Oct 13, 2020
174
159
Austin, Texas
Interesting, thanks.

Another thread says the Unior 1671.2/4 tool fits, at £19 not too painful.
Just have to source a replacement chain ring now.
I just ordered a 2021 rail 9.8XT and see it comes with a 32 tooth ethirteen chain ring. I want to go immediately to a 38T chainring. Is the Shimano xt chainring a direct replacement? Not sure of the standard used for that crank and chainring attachment.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
I just ordered a 2021 rail 9.8XT and see it comes with a 32 tooth ethirteen chain ring. I want to go immediately to a 38T chainring. Is the Shimano xt chainring a direct replacement? Not sure of the standard used for that crank and chainring attachment.
I doubt you'll be able to fit a 38t! Not enough clearance on the chainstay
 

dixie600mhz

Active member
Oct 13, 2020
174
159
Austin, Texas
I doubt you'll be able to fit a 38t! Not enough clearance on the chainstay
Really!!! That would suck. Even with the 53mm offset? My wife's Luna X1 has a 38T and probably needs bigger. Her bike does have the 27.5" wheels tho.

I see e*thirteen are doing espec chainrings and are linked to from the rail 9.8 Trek official page. Sizes are 34T, 36T and 38T. I'm running a 36T oval on my canyon spectral carbon (non ebike) and hardly ever use 1st gear, even on the most technical climbs.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
Really!!! That would suck. Even with the 53mm offset? My wife's Luna X1 has a 38T and probably needs bigger. Her bike does have the 27.5" wheels tho.

I see e*thirteen are doing espec chainrings and are linked to from the rail 9.8 Trek official page. Sizes are 34T, 36T and 38T. I'm running a 36T oval on my canyon spectral carbon (non ebike) and hardly ever use 1st gear, even on the most technical climbs.
I've never tried a bigger chainring but the clearance is pretty minimal

20201018_151941.jpg
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
Thanks for the pic. What size chainring is shown? Doesn't look like much room like you said.
It's a 34 tooth. I wish the motor completely decoupled after the 32 km/h assistance threshold. I don't mind that the assistance cuts out at this speed, but it also resists pedalling beyond this speed, like engine braking (even though it's apparently much better than the gen 3 motor) For this reason I wouldn't bother changing the gearing on the bike, I think there's no point.
 

dixie600mhz

Active member
Oct 13, 2020
174
159
Austin, Texas
I thought the 4th gen with firmware update completely disengaged? I test drove a 2020 rail 7 and it felt like throwing out a boat anchor at 32km/h so I know what you mean. I see there's an inline connector that bypasses the limit. It's available online! Pretty pricey tho ~$170.
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
I thought the 4th gen with firmware update completely disengaged? I test drove a 2020 rail 7 and it felt like throwing out a boat anchor at 32km/h so I know what you mean. I see there's an inline connector that bypasses the limit. It's available online! Pretty pricey tho ~$170.
The 2020 Rail 7 is what I have! Yeah the 4th gen motor is much better than the 3rd gen that was on the Powerfly, but the firmware update (which I have) doesn't change the boat anchor effect at 32km/h. I bought the Rail for helping me up steep climbs, and it's the best thing ever for doing that, but with the right gearing I could achieve a higher top speed on the flat riding my accoustic. Doesn't bother me too much since most of my riding is just climbing up to bomb back down.
 

hannu

Member
Aug 9, 2019
26
4
FIN
It's strange that the first 500km with the motor I didn't have any problems with the nut, but after the first maintenance where they had removed the nut, it started to get loose just after one ride. Noticed that they've either forgotten the rubber O ring or used a damaged one when putting the bike back together so it had worn out completely. I put a new O ring there (and no loctite or anything) and now the nut has been alright for the last 300km at least. So maybe the O ring really plays some role there.
 

Conti2020

New Member
Jul 22, 2020
15
0
Brasil
Chainring nut / spider nut or other names...
Has anyone experienced their chainring nut coming loose?
Today is the first day my Rail has let me down, I had to cut my ride short today when I felt a vibration at my feet, checked the chainring and it was wobbling and the spider nut had come completely undone... anyone experienced this?
I made myself a tool to tighten it back on today and will check it again next time out. Has anyone else had this happen on a Gen 4 motor or van I put it down to a build issue? View attachment 27677
Look how close is the chainring to the chainstay. My is a rail 7 XL and the chainring is damaging de chainstay. Anyone has this problem too? Is there a solution for this? Oring, new chainring with offset?

Screenshot_20201105-200709_Files_by_Google[1].jpg
 
Last edited:

rbmusica

New Member
Mar 22, 2021
1
0
Beautiful Italy
Hi Conti, I have a rail 9.8 (sram, medium frame), and I have exactly the same problem, bike is new with just 200km and the frame is already scratched in the same spot....will replace ASAP the chainring with a 30t, or 32-34 with offset....anyone knows if the standard 34t chainring has a +3 offset, or an offset at all?
 

dixie600mhz

Active member
Oct 13, 2020
174
159
Austin, Texas
I tried a 36T front chainring and everyone is right. There's no way it fits my 9.8XT in large. Even the ethirteen e-spect 34T is super close.

Has anyone found an online place that sells the gen 4 performance lock ring and o-ring? Mine felt like it would strip the threads just getting it up to 20Nm.

36T_2.jpg
tool.jpg
 

Philly G

Well-known member
Jun 29, 2020
692
517
New Zealand
@dixie600mhz In my experience the o-ring isn't necessary, I found it impossible to tighten the lock ring nut without tearing the o-ring in any case. Curious as to why you feel you' need a replacement lock ring nut, if the threads are not damaged in any way you should be fine to re-use it
 

dixie600mhz

Active member
Oct 13, 2020
174
159
Austin, Texas
@dixie600mhz In my experience the o-ring isn't necessary, I found it impossible to tighten the lock ring nut without tearing the o-ring in any case. Curious as to why you feel you' need a replacement lock ring nut, if the threads are not damaged in any way you should be fine to re-use it
I did end up re-using it and getting it to 30Nm. It felt like it wasn't getting tight but it must have been the chainring pressing into place. It's working. Can't believe how close the 34T chain ring is to the chain stay. It's not rubbing but it's super close. I checked all the motor mount bolts and everything seems ok.

It would be nice if the lock ring was easier to source. For now I'm good.
 

EMTB Forums

Since 2018

The World's largest electric mountain bike community.

554K
Messages
27,985
Members
Join Our Community

Latest articles


Top