Rail (750Wh) Bosch Smart V5 update Mileage/Speed error

[Zimmerframe]

It will be interesting to see how Trek is dealing with this Bosch/Magnet issue considering they just dropped all their new 2023 Rails

I've heard they're going twice as far with the latest models .......

 

[Arminius]

I think we've proposed this as being the issue from the start (even before it became it's own thread...)



I don't think so. I still think it's purely down to how the trek sensor is placed in a different position/angle, combined with a larger magnet to obtain a reading. This doesn't always work and sometimes the magnet needs to be manually adjusted as proposed earlier in the thread as a possible solution to this problem.

Swapping to the Trek sensor/magnet doesn't necessarily cure the problem if we don't know how they've mounted it and in what combination and spacing - ie, done in such a way that the more de-restriction sensitive firmware, still thinks somethings not right so it doubles up.

For all the electronic de-restrictors, it won't matter what sensor feeds into the unit as the unit will output it's own signal which will only be based on pulse frequency, not strength/time/echo's. Though with the Gen4 smart, some will need to vary that signal output to reflect what's expected at given rotation speeds.

For the mechanical de-restrictors, it would be interesting to know if anyone with a planet3 has this problem, or if because of how the magnet is mounted, everything actually works just as before - which as @Paul Mac said earlier, it would be ironic to use a de-restriction device to actually fix this problem, but even more ironic if the firmware mods made to limit de-restriction, actually don't affect some of the bikes which are derestricted whilst affecting the bikes which are standard :-D

Either way, I can't imagine someone at Bosch fecking something up on purpose out of spite for one of their largest customers. Naivety, lack of thought, inexperience - yes, more likely.

If related to prevent from tuning the position of Trek magnet with possibly two signals at a time could be a reason. For example Wiesel is blocking counts and Bosch might have implemented a check for half and double or inconsistent counts :
„Wheel revolutions missed due to tuning are added at the end of the ride. The speedometer then shows the exact distance travelled.“

PS: Did one of the Trek owners have not only Bosch magnet but also sensor replaced as a try with improvised fixing of the sensor? Or was it magnet only?

 

[blob]

Whilst this might be a great solution, people should generally use someone elses computer if they're searching for Rim Magnets .. otherwise all sorts of weird results come up ..

View attachment 107983

I've ordered this magnet, but can you provide some mounting instructions?

 

[Arminius]

I've ordered this magnet, but can you provide some mounting instructions?

Text only! 🫣

 

[Jeremybub]

If related to prevent from tuning the position of Trek magnet with possibly two signals at a time could be a reason. For example Wiesel is blocking counts and Bosch might have implemented a check for half and double or inconsistent counts :
„Wheel revolutions missed due to tuning are added at the end of the ride. The speedometer then shows the exact distance travelled.“

PS: Did one of the Trek owners have not only Bosch magnet but also sensor replaced as a try with improvised fixing of the sensor? Or was it magnet only?

Trek swapped the speed sensor cable and magnet last friday to the Bosch speed sensor cable and it did not work, and did not fit in the rear triangle either.
Trek ordered the wheel magnet Monday and it will arrive tomorrow morning, fingers crossed Should be fixed.

 

[JoeJ]

Got my one fixed today with the new rim magnet, all good now

What exactly have they done to fix the issues and the walk mode? So I know what to ask the dealer.

 

[Paul Mac]

What exactly have they done to fix the issues and the walk mode? So I know what to ask the dealer.

All that has been done, is rim magnet attached and Bosch software at dealer changing from the wired sensor to rim sensor, that's it, nothing else.

 

[Jeremybub]

FIXED, however not perfect but not doubling the miles anymore!
Bocsh and Trek enabled the internal motor magnet senso, installed the wheel magne And presto
see image,
The Folsom Trek Factory shop offered to work with any USA Trek dealer to solve the issue, contact number below. have them ask for the service manager Mike
9168363890

 

[Martin2610]

Hello. I don't know why I should put such a strange thing on my bike when my bike worked normally before the update. Let Bosch and Trek take care of it.

 

[Arminius]

I am glad the issue can be solved and wonder if there will need to balance wheels with such device on the rim?

By this occasion I would like to say that I enjoy this forum a lot! Because of its culture, i.e. how problems and postings are addressed as well as the respectfully conversations!

 

[billium]

Wow,
I write SW on a daily basis (some for embedded systems) and this smells like a real Cluster F!

Can't blame anyone here for speculating, but there are blaringly obvious troubleshooting and process issues in the way that this has been handled by Trek.


How it should have gone once it got validated and escalated as a real issue:

Step 1 - Did the previous FW report properly or not?
Ok, test it at Trek Engineering on one of the many bikes lying around that have not yet been updated - lets be generous and give them 1/2 a day to find out if existing sensor setup works and gives correct numbers.

Next ( following day) call the person or very small team at Bosch that writes the FW and let them know the answer.
If YES then someone in SW needs to find the new bug. 1 day to find it , 1-2 days to test

If No then Trek requests a Debug App from Bosch that displays the raw magnet readings on a display and they spin wheels and count revolutions. If you spin the wheel 100 times does the sw count to 100?
Testing could take several days but the problem would likely be obvious in first 5 minutes with the debug app. The fix could be involved but telling customers 'it might be this ' or 'try that' suggest their troubleshooting process management is pretty poor.

These are big companies but they really should know when and how to spin up a 'tiger team' to address issues like this.

Their 'fix' looks like a bodge and we should ask the following:

• How heavy is that big magnet?
• Will it imbalance the wheel?
• How does it work on metal vs carbon rims & different rim profiles?
• Will it stay secure or will fast rocky descents move it around and cause the valve to leak at the rim?
• If it does move, might it wear holes in carbon rim?
• What if owner wants to Mullet with smaller wheel - will it still work?
• TyreWiz compatibility?
• Since nobody else is doing this, are you really sure this solution is better than stock sensor system? Prove it!

 

[Shark58]

On a German emtb forum someone complained about double mileage on his bike after the last Bosch motor update.

When I saw this thread here I had asked him whether he owned a Trek bike. No he said it’s a Cube from 2022 but came back later to report that he had mounted his second wheel set for the winter and checking that found a Trek magnet on the disk. All he needed to do to fix the issue was remove the Trek magnet and put an original Cube magnet in its place. Problem solved - mileage back to normal and all else working as supposed.

AFAIK Cube uses stock Bosch components for magnets and speed sensors. You might want to try this.

The user provided some measurements for the magnet and the Cube and Trek ones differ significantly.
Bosch (Cube) magnet is flat and round, mounted with 10mm more distance (32mm) to the axle. Trek magnet is cylindrical and sits only 22mm from the axle.

 

[paquo]

i just bought that bosch slim speed magnet to use on my ep8 bike so i should switch it out and see

 

[Mikerb]

On a German emtb forum someone complained about double mileage on his bike after the last Bosch motor update.

When I saw this thread here I had asked him whether he owned a Trek bike. No he said it’s a Cube from 2022 but came back later to report that he had mounted his second wheel set for the winter and checking that found a Trek magnet on the disk. All he needed to do to fix the issue was remove the Trek magnet and put an original Cube magnet in its place. Problem solved - mileage back to normal and all else working as supposed.

AFAIK Cube uses stock Bosch components for magnets and speed sensors. You might want to try this.

The user provided some measurements for the magnet and the Cube and Trek ones differ significantly.
Bosch (Cube) magnet is flat and round, mounted with 10mm more distance (32mm) to the axle. Trek magnet is cylindrical and sits only 22mm from the axle.
View attachment 108386

View attachment 108387

the difference being the Bosch magnet has North pole magnetic field facing the sensor only. The Trek magnet has both North and South pole magnetic field visible to the sensor. The Trek sensor therefore sends two signals per revolution whilst the Bosch magnet sends one?

 

[billium]

the difference being the Bosch magnet has North pole magnetic field facing the sensor only. The Trek magnet has both North and South pole magnetic field visible to the sensor. The Trek sensor therefore sends two signals per revolution whilst the Bosch magnet sends one?

Makes sense since 'bad' readings should be both fewer(slower) And variable ( not a precise multiple)
If you are correct then the fix could be as simple as two lines of code with no need for Heath Robinson magnet!

if sensor_mode == SensorMode.Trek:
# divide speed by two
speed = speed/2

 

[Chris Manuel]

I've just had a reply back from Trek customer services regarding this issue who say that Bosch are working on an updated update to resolve the problem.

 

[Tmack1960]

Hi - I noticed my last three rides were double the distance they should have been and found this thread when trying to find out what was going on. I just called Trek NZ and they indicated that Bosch have said there is a patch coming so I imagine a bunch of software geeks frantically coding away in a dark room with lots of coca cola!! Hopefully, they will have a way to 'wind' back the 115 phantom kms my bike is now reporting as well.

 

[juice09]

The shop I bought my bike from said according to their contacts at bosch a fix is "imminent ". I'm sure it says that on this thread somewhere a few weeks ago so I'm not holding my breath.

It also says somewhere on here that mileage cannot be rolled back. I did ask bosch about that but they didn't reply. I also asked how they managed to release an update without testing it first...... they're a bit sensitive about that one

 

[irie]

Hi - I noticed my last three rides were double the distance they should have been and found this thread when trying to find out what was going on. I just called Trek NZ and they indicated that Bosch have said there is a patch coming so I imagine a bunch of software geeks frantically coding away in a dark room with lots of coca cola!! Hopefully, they will have a way to 'wind' back the 115 phantom kms my bike is now reporting as well.

It's a cunning marketing plot to more rapidly age your bike so that you upgrade it sooner. Maybe.

 

[Ark]

I really want to buy a trek5 or 7 but this whole thing has made me want to wait..

2months now? and still no fix?
also If I google the trek speed sensor I get a result on the trek site for writing a review Bosch G4 Speed Sensor | Trek Bikes (GB)
But clicking to go back to the actual part
just gives a 404 error.

https://www.trekbikes.com/gb/en_GB/404/

So in the UK Trek don't even sell the sensor anymore on their own website.
it's the same sensor thats on the 2023 rail5/7 right?
Weird it's seems to be obsolete within months of the bikes release date...
Do trek realise this whole ordeal is losing them sales?
Seems like you'd be crazy to buy a bike that doesn't use actual Bosch sensors since manufactures own ones appear to be no longer manufactured within months of a bikes release.
What you going to do in 1-2 years time if the speed sensor fails? you have a 5-6k EBike worth nothing?

Yea sure I guess you can revert to the motors sensor but there must be a reason they don't use that one in the first place?
oh the motors sensor fails and it's costing you like 900 for a whole new motor....
You'd think these people don't want to sell bikes
Literally have the money waiting to buy something, really wanna go Bosch, seems Trek is the best option but everything I read is telling me don't do it

 

[JP-NZ]

Literally have the money waiting to buy something, really wanna go Bosch, seems Trek is the best option but everything I read is telling me don't do it

Man that was a rant, two things...

1. If your looking at a 22 Rail 5 or Rail 7 you have nothing to worry about as its only Bosch smart systems that are affected.
2. If your wanting to buy a 2023 Rail 5 or 7, do you really think brand new bikes on the shop floor will have this problem? Take one around the carpark, you'll soon see if there is any issue. If there is I'm very sure the shop will fix it before you buy.

As far as I'm aware this is a trek smart system problem for MY22 bikes (9.8 & 9.9's)

 

[juice09]

Man that was a rant, two things...

1. If your looking at a 22 Rail 5 or Rail 7 you have nothing to worry about as its only Bosch smart systems that are affected.
2. If your wanting to buy a 2023 Rail 5 or 7, do you really think brand new bikes on the shop floor will have this problem? Take one around the carpark, you'll soon see if there is any issue. If there is I'm very sure the shop will fix it before you buy.

As far as I'm aware this is a trek smart system problem for MY22 bikes (9.8 & 9.9's)

I have a 23 rail 7 with this problem on my first ride and apparently its not just trek

 

[JP-NZ]

I have a 23 rail 7 with this problem on my first ride and apparently its not just trek

Stand to be corrected then.. Your the first person post with a brand new MY23 Trek Rail

 

[paquo]

Literally have the money waiting to buy something, really wanna go Bosch, seems Trek is the best option but everything I read is telling me don't do it

The rail is an amazing bike, especially as specialized are truly horrible re the brose motor and shimano way better but still sketchy. The bosch is a step above ,i would wait for the fix

 

[juice09]

Stand to be corrected then.. Your the first person post with a brand new MY23 Trek Rail

Unless they track their rides and check their mileage against the flow app its entirely possible some people don't even realise there is a problem.

 

[JP-NZ]

I have a 23 rail 7 with this problem on my first ride and apparently its not just trek

Did you update the software within the flow app after buying it or was it incorrect from the shop?

 

[juice09]

Did you update the software within the flow app after buying it or was it incorrect from the shop?

I updated the software with the app as soon as I got the bike home. I have no idea whether it was ever working normally. It shows 2 miles were added to the bike before I got it. If I had known there was an issue I probably wouldn't have bought the bike at all. Its now been dragging on nearly 2 months and a fix has been "imminent" for most of that time.

 

[blob]

According to a rep from Trek, they stopped pushing this update when they realized there's a problem. So the number of us lucky ones who got the update should be limited.

 

[Shark58]

As I mentioned in #192 just remove the Trek magnet and mount an original Bosch Magnet. That should fix the issue. No need to wait for an imminent solution from Trek.

Not only the mileage is wrong, this also leads to incorrect average speed. I saw a ride statistic where someone managed to reach a higher average than max speed on a ride.

 

[juice09]

As I mentioned in #192 just remove the Trek magnet and mount an original Bosch Magnet. That should fix the issue. No need to wait for an imminent solution from Trek.

Not only the mileage is wrong, this also leads to incorrect average speed. I saw a ride statistic where someone managed to reach a higher average than max speed on a ride.

It is bosch that are providing the "imminent " solution. I think a magnet change may have been tried earlier in this thread and it didn't work?