New "TSDZ2 Torque Sensor Central Motor"

[dameri]

Nice thoughts and pictures Wapous.

 

[ornias]

Most of these issues have been known for a few years by now. The issues is not lack of knowhow, the problem is bascially that no one bothered to actually work on getting that shitty axle fixed.

..... and Wapous is one of the people who helped this community over these years to build that knowhow.
So whats your point other than being negative about other peoples contributions??
Also ‚getting the axle fixed‘ is a matter of personal priorities: I will not spend around 100€ on a preventive maintenance spare part for a motor that costs only a little over 200€ - and were the failure rate in many (i.e. my) use cases is pretty low. Because when you do that you will only stumble into the next weakest link, i.e. the known crack sensitivity of the chainring shaft.

Also: Replacing the seal in the crank with a bearing, nullifies a lot of offset issues between spindle-axle and main-drive-gear.

That is only true for the right side. Exzessive play on the left side is caused by the dimensions of quoted needle bearing and the chainring shaft (ontop of the axle) and can not really be compensated by adding an outer bearing on that side

The point was, that while those initial drawings where great, some people where reacting like there was something new in those initial drawings. It's important to note it's a rehash, otherwise we again get a flood of new people, adding a lot of repeated "blablabla" helping absolutely no one.

Anyway:
The issue with the axle is not it breaking, the issue is the fact the tolerances are so loose, it breaks other parts. Like: 10x too-loose. It's not a preventative maintainance thing that is of relevance, it's the fact it's a badly manufactured part. Which actually increases the chances of failure of other parts.

For example: When you add play to an axle, inside another axle (which is basically the case here) which has a needle bearing with too-much play already, those forces are not just additive, they are often expodential. Leading to expodential increases in failure rates.

Also: The motor is about 400-450 euro's incl. Tax (Which is always the case when prices are in euro's)

 

[ornias]

Here is how I would machine a new BB shaft using 3 roller bearings.

Interesting design, i'm willing to put some time into this as a second prototype, if you're willing to forward me the CAD drawing

In general I like how the thick Boss endpiece, limits sideways play while also providing a more rigid connection between the torque-sensor and axle. It also removes the need of gluing (or circlip-weakspot) bearings on the non-drive side.

Only issue I have with this, is the fact it still retains the weekspot circlip on the drive side

Though I do note that you're using 2 needle bearings on the torque sensor (besides the new 1 additional on the axle boss), which I think is still based on the older torque sensor design?

 

[tosiwic227]

How long can stock 2 needle bearings last without maintenance? 10,000 km will last at least? And how should they be lubricated during maintenance?

 

[ornias]

is this worthy of time and money ?

I found the money isn't that much, it's mostly the time it takes getting everything ready to have things machined.

How long can stock 2 needle bearings last without maintenance? 10,000 km will last at least? And how should they be lubricated during maintenance?

Basically many other parts will likely fail before those needle bearings.


---
On another note:
Can anyone send me drafts and pictures of the TSDZ2 main clutch bearing and/or has anyone figured out what (if any) the aftermarket bearing numbers for it are?

 

[tosiwic227]

And how long can a torque sensor work before it breaks? How many kilometers will the rubbing parts withstand?

 

[ornias]

And how long can a torque sensor work before it breaks? How many kilometers will the rubbing parts withstand?

Breaking: Unknown, depends on the tolerances within your personal parts

The rubbing is irrelevant, you should put it into your maintenance schedule to check the low-friction material for wear and replace accordingly

 

[tosiwic227]

At what mileage low-friction material usually wear out?

 

[ornias]

is this worthy of time and money ?

Absolutely not. And I hope there are no penalties for doing this.

I wonder why you're saying this?
Like: The money isn't thát bad and the time... well... for each their own hobby I guess.

In terms of panelties:
- Having a third roller-bearing non-drive side, does mean the tolerances are much more important (In comparison to cardridge bearings + bearing glue), however as the seal can also still be replaced by a (glued-in) bearing, I doubt this is a significant issue.
- Weight distrubution has changed, but considering the huge gear on the other side, I guess that's an improvement

Still would like all the CAD drawings we can get (primarily the axle by you Wapous), as i'm planning another axle prototype in januari/februari

 

[pxl666]

honestly , i've put a lot stress to this motor - i weight 89kg and did some serious climbing (live close to the mountains) and apart from temperature sensor and some better heat dissipation with thermopads or aluminium pieces i see no point in this .
the key is 34 / 36 chainring and 50t casette and rider who uses them with reason ...i never broke nylon gear and never broke motor (only got water inside by using karcher to get bike clean...)and for this price i see no point in such a heavy modding ...

 

[ornias]

honestly , i've put a lot stress to this motor - i weight 89kg and did some serious climbing (live close to the mountains) and apart from temperature sensor and some better heat dissipation with thermopads or aluminium pieces i see no point in this .
the key is 34 / 36 chainring and 50t casette and rider who uses them with reason ...i never broke nylon gear and never broke motor (only got water inside by using karcher to get bike clean...)and for this price i see no point in such a heavy modding ...

Hacking hardware is not ónly about it having a point though...

 

[pxl666]

from my point of view the key is to mod as little as needed to have fun with finished product like ebike as a whole . I was riding tsdz2 for a season with no modding at all and was happy with it. I found in need for mod when my torq sensor lost some sensitivity . of course now I have somewhat better motor but thanks to mbrusa's hard work it was easy like a fart...I can't imagine machining new parts for this motor however i understand it's fun for fun's sake ...

 

[endlessolli]

Also: The motor is about 400-450 euro's incl. Tax (Which is always the case when prices are in euro's)

Nope - 246,64EUR incl free shipping from germany:
https://www.pswpower.com/products/europe-or-china-stock-36v-250w-or-350w-42t-vlcd5-tsdz2-electric-bicycle-central-mid-motor-with-throttle-e-brake-lever-79

 

[sqeeezy]

48v 500w 15s3p LiFePO4 diy 32650 battery, 8 months in now, still on OEM firmware. Did heat mods, with diy ally shim and heat foam. Using 34T front and rebuilt rear cassette to give me a 34T granny. In the south of Spain, my use-case is shopping 3 times a week with a 500m climb over 9km from town to home. All good so far, taking care to travel as early as possible to avoid the heat in summer....but a clonking noise today on pedalling prompted me to wonder about the feasible lifetime for this motor, given the doubts as to its quality, manufacturing tolerances. Is there a best estimate for a reasonable expectation for how many km I can clock before the thing falls to bits or the torque sensor fails. Who's done the max mileage so far? Is the TSDZ2 a "keeper"?

 

[beemac]

48v 500w 15s3p LiFePO4 diy 32650 battery, 8 months in now, still on OEM firmware. Did heat mods, with diy ally shim and heat foam. Using 34T front and rebuilt rear cassette to give me a 34T granny. In the south of Spain, my use-case is shopping 3 times a week with a 500m climb over 9km from town to home. All good so far, taking care to travel as early as possible to avoid the heat in summer....but a clonking noise today on pedalling prompted me to wonder about the feasible lifetime for this motor, given the doubts as to its quality, manufacturing tolerances. Is there a best estimate for a reasonable expectation for how many km I can clock before the thing falls to bits or the torque sensor fails. Who's done the max mileage so far? Is the TSDZ2 a "keeper"?

It's so hard to give a single answer to that - it depends so much on a number of factors. How you ride, luck with the particular motor you bought, the mods you do, the terrain etc. etc. I love the tsdz2 but I wouldn't wholeheartedly recommend it to someone who wanted a fit and forget solution. If you like to fiddle I think it's great - cheap, easy to repair, lots of software/display options, good community yada yada.

You've done the same mods as me pretty much and I've had no issues in two years. I have two tsdz2's but the one I ride daily is an fsr with knobbly tires and Bristol is pretty hilly.

When my bike started clunking when I pedalled it was because the hex bolts holding the chainring to the spider had loosened, might be worth a check esp if you've changed chainrings.

 

[Rikstaker]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

I have the osf kit, with 42t solid ebike chainring which has a 10mm offset. It was not an easy install, I had to cut the eccentric bottom bracket adapters at a machine shop as the Grail BB has a groove which wont let the adapters go all the way in, Then had to pay a bike shop to externally route the RD cable as it was in the way. After all that was done and I installed it, chain kept coming off frame-side when I shifted to higher than 5th. Took it to lbs, yupp, bad news cant do anything as the angle of chain at anything above the 5th smallest cog is not doable.

I looked up a few options, I thought I will buy a bafang ring with a much higher offset like 14 or even 30 and frankenfit it using a spider adaptor or something but when I looked under the chainring, I realized thats out of question as there is barely any clearance between the housing and chainring.

See pics.

Maybe if I go a crazy big chainring like 52-54 I can clear the housing and a dished shape would bring the chainring back in, but at what cost. Il probably melt the blue gear going up any moderate hills.

Any suggestion. I am pretty close to binning it and maybe get a photon motor made by cyc.

 

[tosiwic227]

I'm also looking at CYC PHOTON. And the main question that arises is: will the CYC PHOTON be more reliable than the 4 TSDZ2B bought for this price?

 

[Elinx]

...will the CYC PHOTON be more reliable than the 4 TSDZ2B bought for this price?...

You have to buy and try a Cyc Photon for a longer time before you can judge that.
But if this is not the case, you will miss the simplicity of the Tsdz2.

 

[beemac]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

I have the osf kit, with 42t solid ebike chainring which has a 10mm offset. It was not an easy install, I had to cut the eccentric bottom bracket adapters at a machine shop as the Grail BB has a groove which wont let the adapters go all the way in, Then had to pay a bike shop to externally route the RD cable as it was in the way. After all that was done and I installed it, chain kept coming off frame-side when I shifted to higher than 5th. Took it to lbs, yupp, bad news cant do anything as the angle of chain at anything above the 5th smallest cog is not doable.

Is the chainline particularly bad to begin with? I run the same 42t solid ebike chainrings - one with a 12-34T 11 speed cassette - and another with an 10-51T 12 speed cassette - both have no issue shifting across the range. I even have a 2nd 50T solid ebike chainring and a standard front mech on the 11 speed and that shifts fine across all gears...

my only issue is that I chew through the smaller cassette sprockets... i go through them quicker than brake pads

edit: I've just realised that both my bikes had triple chainrings originally - that probably makes all the difference...

 

[seattlesockey]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

Would internal gear hub work?

 

[Rikstaker]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

I have the osf kit, with 42t solid ebike chainring which has a 10mm offset. It was not an easy install, I had to cut the eccentric bottom bracket adapters at a machine shop as the Grail BB has a groove which wont let the adapters go all the way in, Then had to pay a bike shop to externally route the RD cable as it was in the way. After all that was done and I installed it, chain kept coming off frame-side when I shifted to higher than 5th. Took it to lbs, yupp, bad news cant do anything as the angle of chain at anything above the 5th smallest cog is not doable.

Is the chainline particularly bad to begin with? I run the same 42t solid ebike chainrings - one with a 12-34T 11 speed cassette - and another with an 10-51T 12 speed cassette - both have no issue shifting across the range. I even have a 2nd 50T solid ebike chainring and a standard front mech on the 11 speed and that shifts fine across all gears...

my only issue is that I chew through the smaller cassette sprockets... i go through them quicker than brake pads

edit: I've just realised that both my bikes had triple chainrings originally - that probably makes all the difference...

My bike was a 2x originally. The problem lies with the frame, the BB area, notice how the BB sticks out past the welds, this causes the motor to not go in enough to get a decent chainline even with the offset chainring.

 

[dameri]

I have a Trek X-caliber 9. The TSDZ2 frame hits the chainstay and the three largest sprockets cannot be used.

 

[sqeeezy]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

I have the osf kit, with 42t solid ebike chainring which has a 10mm offset. It was not an easy install, I had to cut the eccentric bottom bracket adapters at a machine shop as the Grail BB has a groove which wont let the adapters go all the way in, Then had to pay a bike shop to externally route the RD cable as it was in the way. After all that was done and I installed it, chain kept coming off frame-side when I shifted to higher than 5th. Took it to lbs, yupp, bad news cant do anything as the angle of chain at anything above the 5th smallest cog is not doable.

I looked up a few options, I thought I will buy a bafang ring with a much higher offset like 14 or even 30 and frankenfit it using a spider adaptor or something but when I looked under the chainring, I realized thats out of question as there is barely any clearance between the housing and chainring.

See pics.

Maybe if I go a crazy big chainring like 52-54 I can clear the housing and a dished shape would bring the chainring back in, but at what cost. Il probably melt the blue gear going up any moderate hills.

Any suggestion. I am pretty close to binning it and maybe get a photon motor made by cyc.

What I did was rebuild my cassette to put the granny gear in the furthest reachable inboard position (5th in) and miss out 3, which I installed inboard of the granny just to keep the spacing/indexing right. I set my low limit so it doesn't pass the granny and it's workable although it would be nice to have less of a jump across the 5 that I'm using instead of 8. I bought a SunRace as directed by this guy https://youtu.be/1WBMpZv-pqE but afterwards I ground out the rivets on my Shimano cassette and it wasn't at all difficult to shuffle the sprockets to how I wanted them.[EDIT: I've been given an Alfine 11 so that should make everything good, as long as I fill it with grease and not oil]

 

[sqeeezy]

48v 500w 15s3p LiFePO4 diy 32650 battery, 8 months in now, still on OEM firmware. Did heat mods, with diy ally shim and heat foam. Using 34T front and rebuilt rear cassette to give me a 34T granny. In the south of Spain, my use-case is shopping 3 times a week with a 500m climb over 9km from town to home. All good so far, taking care to travel as early as possible to avoid the heat in summer....but a clonking noise today on pedalling prompted me to wonder about the feasible lifetime for this motor, given the doubts as to its quality, manufacturing tolerances. Is there a best estimate for a reasonable expectation for how many km I can clock before the thing falls to bits or the torque sensor fails. Who's done the max mileage so far? Is the TSDZ2 a "keeper"?

It's so hard to give a single answer to that - it depends so much on a number of factors. How you ride, luck with the particular motor you bought, the mods you do, the terrain etc. etc. I love the tsdz2 but I wouldn't wholeheartedly recommend it to someone who wanted a fit and forget solution. If you like to fiddle I think it's great - cheap, easy to repair, lots of software/display options, good community yada yada.

You've done the same mods as me pretty much and I've had no issues in two years. I have two tsdz2's but the one I ride daily is an fsr with knobbly tires and Bristol is pretty hilly.

When my bike started clunking when I pedalled it was because the hex bolts holding the chainring to the spider had loosened, might be worth a check esp if you've changed chainrings.

Cheers for that. So you're the only taker in the longevity stakes so far, congrats on your two years. (Raises voice) "Any advance on two years..." I'll give those chainset bolts a coat of looking at, ta.

 

[Rikstaker]

PSA

This motor wont work with Canyon Grail. Well to elaborate you cant use more than 5 highest gears if you install this on a Grail.

I have the osf kit, with 42t solid ebike chainring which has a 10mm offset. It was not an easy install, I had to cut the eccentric bottom bracket adapters at a machine shop as the Grail BB has a groove which wont let the adapters go all the way in, Then had to pay a bike shop to externally route the RD cable as it was in the way. After all that was done and I installed it, chain kept coming off frame-side when I shifted to higher than 5th. Took it to lbs, yupp, bad news cant do anything as the angle of chain at anything above the 5th smallest cog is not doable.

I looked up a few options, I thought I will buy a bafang ring with a much higher offset like 14 or even 30 and frankenfit it using a spider adaptor or something but when I looked under the chainring, I realized thats out of question as there is barely any clearance between the housing and chainring.

See pics.

Maybe if I go a crazy big chainring like 52-54 I can clear the housing and a dished shape would bring the chainring back in, but at what cost. Il probably melt the blue gear going up any moderate hills.

Any suggestion. I am pretty close to binning it and maybe get a photon motor made by cyc.

What I did was rebuild my cassette to put the granny gear in the furthest reachable inboard position (5th in) and miss out 3, which I installed inboard of the granny just to keep the spacing/indexing right. I set my low limit so it doesn't pass the granny and it's workable although it would be nice to have less of a jump across the 5 that I'm using instead of 8. I bought a SunRace as directed by this guy https://youtu.be/1WBMpZv-pqE but afterwards I ground out the rivets on my Shimano cassette and it wasn't at all difficult to shuffle the sprockets to how I wanted them.[EDIT: I've been given an Alfine 11 so that should make everything good, as long as I fill it with grease and not oil]

Hi thanks, thats a good suggestion. I took it for a spin and its definitely doable with half the gears and rearranging the cassette should help. I was wondering if I should get a cheap bike for this like a hybrid with suspension and fewer gears to begin with and a threaded BB and just use my Grail for quick 1 hr high intensity rides. My main purpose for electrification was to enable me to go on longer rides where I would encounter some steep hills. I was wondering if I can just swap the 100mm spindle out from the motor and use a stock one on it or do I need a whole new motor to work with the new bike.