New "TSDZ2 Torque Sensor Central Motor"

[izeman]

Shame. Didn't look at the screw it self. Not that I could have read it without magnifying glasses anyway. Thanks!
I'm not using the supplied cranks, as I had to shorten mine because of the 20" rims I use in my former 26" bike. Could be that the Tongsheng taper is not exactly the same as the original Shimano one. Didn't think of that before. I'm totally aware how tapers are supposed to work, but neglected the fact that improper production could be the issue.
Maybe I'll try Helicoil once the thread gets really striped. I may also get longer bolts/screws as the threads on my axle are really deep. Still don't you think that it will weaken the axle even more? There is not much material and the hole and thread you need to drill for an M8 Helicoil is quite large.

 

[Huguenot]

Shame. Didn't look at the screw it self. Not that I could have read it without magnifying glasses anyway. Thanks!
I'm not using the supplied cranks, as I had to shorten mine because of the 20" rims I use in my former 26" bike. Could be that the Tongsheng taper is not exactly the same as the original Shimano one. Didn't think of that before. I'm totally aware how tapers are supposed to work, but neglected the fact that improper production could be the issue.
Maybe I'll try Helicoil once the thread gets really striped. I may also get longer bolts/screws as the threads on my axle are really deep. Still don't you think that it will weaken the axle even more? There is not much material and the hole and thread you need to drill for an M8 Helicoil is quite large.

Yes, it takes a magnifying glass to read (unless your wife just bought herself an iPhone 11 pro -amazing camera!) and I had originally assumed it was just the makers logo. The tapers should be the same as that is an industry standard. (There have been variations over the years but things appear to have settled down). Worth checking, as retrorocket mentions above, with dye or even a thick coat of felt pen ink if you are concerned about it.
Your idea of using a longer bolt is probably the best solution if the internal threads are bad, but the hole may not be tapped to the full depth it was drilled to so check first. Internal threads do not end just suddenly, the deepest ones will be tapered so a bolt running into this can give a false torque reading.
I would avoid a helicoil for this application due to limited material here as you mention.
A damaged taper can ...sometimes... be cleaned up with the proper file and great care. One thing I always do when installing a crank arm that has seen previous use is to clean up the ridge which was formed by the end of the original axle taper so it does not interfere with the new one. If present, I will bevel the sharp edges often present at the very end of the square with a file to minimize the shaving and ridge formation mentioned above.
Also, and especially when dealing with chinesium cranks, I mark the orientation of the crank arms so they can be reinstalled exactly the same way.

 

[Retrorockit]

https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/raw/master/TSDZ2_different_versions_axles_comparition.jpg
There are 2 different shafts and the taper appears to end sooner on the narrow version.
Some cranks may seat up against the end of the taper instead of seating on the flat.
There are "bottoming" taps that cut threads to the end of the hole if that will help you.

 

[barrettrussell]

Hello all. I have a motor that is not working.

There is no torque on pedal, and the throttle only barely allows it to move. Was working fine last night and then this. The display also cuts out oddly quick. My first thought was one of the legs of power wasn't going to the motor. This is also compounded by the fact that after a few seconds of grinding slow moving, the throttle wont have function at all if let off and then pressed again, but only if the wheel is turned backwards slightly, possibly realigning something in the motor, will it work again? The motor has a measly 2 weeks and 200km on it.
Thanks!

https://www.youtube.com/watch?v=R-bNCbKeK9c

 

[izeman]

https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/raw/master/TSDZ2_different_versions_axles_comparition.jpg
There are 2 different shafts and the taper appears to end sooner on the narrow version.
Some cranks may seat up against the end of the taper instead of seating on the flat.
There are "bottoming" taps that cut threads to the end of the hole if that will help you.

Thanks for letting me know. I will take some blueing fluid and a file to get a perfect fit. Then look for a longer screw and tighten it with the correct torque

 

[MaartenL]

Hi all!

I have about 2300km on my motor on my fatbike. So at 1000km I did take the motor apart and gave it a clean / grease. Now at 2300km I got this noise:

[youtube]XWrn7EUKT_k[/youtube]

I have the metal gear installed since the beginning.

Before I take it apart, any thoughts? PS. what sort of grease (link?) should I consider for the engine. I'm from Haarlem The Netherlands, so if anyone knows a local webshop even better.

 

[bergerandfries]

Hello all. I have a motor that is not working.

There is no torque on pedal, and the throttle only barely allows it to move. Was working fine last night and then this. The display also cuts out oddly quick. My first thought was one of the legs of power wasn't going to the motor. This is also compounded by the fact that after a few seconds of grinding slow moving, the throttle wont have function at all if let off and then pressed again, but only if the wheel is turned backwards slightly, possibly realigning something in the motor, will it work again? The motor has a measly 2 weeks and 200km on it.
Thanks!

https://www.youtube.com/watch?v=R-bNCbKeK9c

Do you know what voltage the battery is charged to? What's the voltage of the kit you bought?

 

[barrettrussell]

Hello all. I have a motor that is not working.

There is no torque on pedal, and the throttle only barely allows it to move. Was working fine last night and then this. The display also cuts out oddly quick. My first thought was one of the legs of power wasn't going to the motor. This is also compounded by the fact that after a few seconds of grinding slow moving, the throttle wont have function at all if let off and then pressed again, but only if the wheel is turned backwards slightly, possibly realigning something in the motor, will it work again? The motor has a measly 2 weeks and 200km on it.
Thanks!

https://www.youtube.com/watch?v=R-bNCbKeK9c

Do you know what voltage the battery is charged to? What's the voltage of the kit you bought?

Thanks for the reply!

The motor has worked flawless for about 200km of trail. 36v with 36v motor. I took the motor assembly somewhat apart today and found the blue gear is fine, and the motor controller (?speed control) which is potted, doesn't appear to have any burned spots. Tonight after supper I will take apart the other side to assess the toque sensing rings. I have read these can fail due to hard riding, but I am unsure how that happens as I don't really understand how they work at this poin, that is until i see them.

Thanks again.

 

[Waynemarlow]

If you have the standard VLCD5 screen, there's a sub menu which you can see the readings from the torque converter to check its correct operation.

https://wooshbikes.co.uk/manuals/VLCD5.pdf

 

[barrettrussell]

If you have the standard VLCD5 screen, there's a sub menu which you can see the readings from the torque converter to check its correct operation.

https://wooshbikes.co.uk/manuals/VLCD5.pdf

Thanks! I will check the numbers when i get it back together. Everything appears new inside. gears, bearings, and torque sensor.

Can the motor controller partially fail? I am having difficulty getting it out of the body.

 

[rosenbaaron]

... The issue I'm having is that if i charge my battery past full then it won't engage when I turn it on. ...

This is a frequent issue with the TSDZ2. The thought is that the part within the controller that reports the voltage has a very loose tolerance, and sometimes it's fine, and, sometimes it's not. With the factory firmware, my 48V unit works fine as long as the battery is only charged to <= 54.0V (full charge ~54.6V). This is one of the issues that the Open Source Firmware gets past, and I have tested it to do so, but I presently prefer the factory firmware for other reasons. The only recourse, without swapping parts yourself, is to return it for adjustment/replacement, and I have heard of a few cases (pre-pandemic) where this was successfully done, but, it takes lots of time for the round trip. My controller reports 1.5 volts higher than actual. Sorry the news isn't better, but I am now well adjusted to charging less than 100%, which I am told provides a longer life from the battery pack anyway.

Thank you for the reply. This issue is with the firmware on the TSDZ2 so without swapping the tsdz2 I can't permanently fix it, have you tried the luna chargers that charge to 80%?

 

[romelec]

barrettrussell: everything is possible, if you have to put the wheel (and I guess the motor) in a specific position it can be an issue with the hall sensors.

 

[windburner]

..., have you tried the luna chargers that charge to 80%?

No, I have a voltmeter hooked up when charging, and stop the charging when the voltage reaches the target range. Once the charge approaches the 70% range, it is slow enough to provide plenty of time for a loose watch of the voltage. But, automatic is very convenient, so, I'll take a look at them.

 

[Waynemarlow]

Can the motor controller partially fail? I am having difficulty getting it out of the body.

Yes we have had two units now that the controller fails to read the torque sensor. One was a very early unit and the other had been literally drowned in a river so we have had our fair wear and tear out of them.

Don't forget the two small screws that hold the controller into the body.

 

[barrettrussell]

Can the motor controller partially fail? I am having difficulty getting it out of the body.

Yes we have had two units now that the controller fails to read the torque sensor. One was a very early unit and the other had been literally drowned in a river so we have had our fair wear and tear out of them.

Don't forget the two small screws that hold the controller into the body.

thanks for your input.

I am wondering if it may be the hall sensors or a fet because if it was a torque sens I would expect the throttle to behave normally. Motor is loud, ALMOST cogging back and forth, but rather, moves slowly and loudly forward on throttle and only for a partial revolution before stopping, and, no discernable power on pedalling.

i have everything apart. Will test phases. FETS will be hard because they are buried in the potting compound, and im not sure how to test hall sensors.

Barrett

 

[romelec]

I've created the pedal axle in Fusion 360, you can view/download it here: https://a360.co/3gwHR9I
The dimensions should be pretty good but an error is always possible.

I set the axle diameter to 15mm instead of the 14.9mm they did. If it is made with the popper tolerance (h7 in Europe, maybe US have different notations) the bearings should just slide without any play.

I don't have any tools to make it but it shouldn't be that hard to get a few manufactured, the only hard part is the spline that require a custom tool, unfortunately I don't have time right now to ask around.

 

[romelec]

I am wondering if it may be the hall sensors or a fet because if it was a torque sens I would expect the throttle to behave normally. Motor is loud, ALMOST cogging back and forth, but rather, moves slowly and loudly forward on throttle and only for a partial revolution before stopping, and, no discernable power on pedalling.

i have everything apart. Will test phases. FETS will be hard because they are buried in the potting compound, and im not sure how to test hall sensors.

You can test the halls by checking their voltage while the controller is ON (or use a 5V supply) and manually rotate the motor, the tricky part is to access the signal.

 

[bergerandfries]

36v with 36v motor.

This doesn't make sense to me. I understand 36v motor kit, however, a 36v battery usually charges to 42.0v as measured with a voltmeter. Did you measure the battery voltage with a voltmeter?

 

[barrettrussell]

I am wondering if it may be the hall sensors or a fet because if it was a torque sens I would expect the throttle to behave normally. Motor is loud, ALMOST cogging back and forth, but rather, moves slowly and loudly forward on throttle and only for a partial revolution before stopping, and, no discernable power on pedalling.

i have everything apart. Will test phases. FETS will be hard because they are buried in the potting compound, and im not sure how to test hall sensors.

You can test the halls by checking their voltage while the controller is ON (or use a 5V supply) and manually rotate the motor, the tricky part is to access the signal.

thanks for the info.!

 

[barrettrussell]

36v with 36v motor.

This doesn't make sense to me. I understand 36v motor kit, however, a 36v battery usually charges to 42.0v as measured with a voltmeter. Did you measure the battery voltage with a voltmeter?

no problems with battery, i was being too literal in my reply - giving you nominal rather than fully charged. 10s, shows correct volt on display, charged to ~42v. Something broke...worked for 200k or so over a few days. 15k into a ride and it just stopped.

 

[Retrorockit]

I've created the pedal axle in Fusion 360, you can view/download it here: https://a360.co/3gwHR9I
The dimensions should be pretty good but an error is always possible.

I set the axle diameter to 15mm instead of the 14.9mm they did. If it is made with the popper tolerance (h7 in Europe, maybe US have different notations) the bearings should just slide without any play.

I don't have any tools to make it but it shouldn't be that hard to get a few manufactured, the only hard part is the spline that require a custom tool, unfortunately I don't have time right now to ask around.

That's very nice. Here are my thoughts on a modification. I will give the reasons for my thinking on this.
If it were necessary to repair one of these that broke at the snap ring groove there would be two issues.
One is the torque load which could probably be adressed by brazing which in thin sections is very strong in shear.
The second is the bending cycles from the crank arm that put the snap ring area in tension and caused the breakage in the first place. If it were just brazed it would be weaker in tension than before and would fail again. The obvious solution would be to drill in from the end through the existing hole and thread the inner piece so the bolt would hold the repair in compression. The compression load would depend on the quality of the bolt and the torque applied. But the compression force would have to be overcome before the repair would be loaded in tension.
It seems to me that this compression load could be applied to the modified or even adapted to the original part by using a longer bolt at each end.If the drilling made the shaft more flexible this would also reduce the force at the snap ring groove. It might be possible to fix it before it breaks.

 

[barrettrussell]

***update***

It appears one of the hall sensors has failed. Voltage never changes, remains on .04v while the others fluctuate between 0-5v when the shaft is turned. I also tests the mosfet's by continuity - no shorts and around 40.0k ohms resistance. Sound good?

Anyone know what the inside of this motor looks like and what hall sensor to purchase? Thanks.
Barrett

vid of hall sensor voltages
https://www.youtube.com/watch?v=ceu565xA1Dk
vid of what happens when throttle is activated
https://www.youtube.com/watch?v=OsB5RSY7RpU

 

[Bobbybob]

Stealth Build...

- 2013 Motobecane LeChampion Titanium Road Bike
- 36v TSDZ2 w/ 7ah Sanyo GA 'H2GO' Water Bottle Battery
- Maroq/Open Source Firmware w/ VLCD-5 Display
- 52T w/ Chain Guard
- Total Weight w/ Everything: 31 lbs / 14 kg

This thing FLIES!!! ...but to be honest, I am not a road bike guy... and man is it a different ball game holding this bars on the inside with a skinny grip... but I can make better time with this than my folding full suspension 52v! That also has to do with suspension and tires but still... oh yeah, and you see the spare water bottle cage... I can easily bring a 2nd and switch...but you can get surprising range from a 36v7ah on this thing as is...

Anyway - had to share!

This is what I’m planning on doing on a specialised Allez.
IV been searching for the answer for a few days now, so I thought I’d post the question in the end.
I want a bottle style battery. They are 36v 10.5ah and 48v 7ah. Which is the recommended tsdz2? The 36v 500w motor or the 48v 500w motor. I don’t want to go fast. I just want to take the dread from going up some hills on a ride. If everywhere was flat I could go for miles just hills kill me so I tend not to venture out of my comfort zone. The added boost would really help me venture out and about.

 

[izeman]

I've created the pedal axle in Fusion 360, you can view/download it here: https://a360.co/3gwHR9I
The dimensions should be pretty good but an error is always possible.

Thanks for modelling the axle. So I can have look at it without having it in hands :thumb:
Overall it should be fairly easy to make, only problem I see is the splines. And I guess those splines are power loaded and deliver the human power and motor power to the wheel? Or maybe just human power. But still some serious torque.
So the need to me made very precise, correct? And looking at the model the spline sidewalls are parallel to each other, as well as the top surfaces are not curved but also plain.
May I share that file in another forum and ask for tips for fabrication?

 

[romelec]

The spline connects to the torque sensor which transfers the human power.
The motor drives the large gear where the chainring is screwed on.

Yes the bearing supports and spline dimensions must be very precise.

Of course you can share that's why I did it.
There can be small mistakes (<0.1mm) because I only have regular calipers.

Edit: yesterday I added a second bearing (61902-2RS) on the left size, it almost didn't change the axle play.
Then I added some Loctite 648 that is designed to glue bearings with up to 0.15mm play and reinstalled it after it cured (12 hours) and the play is gone.

I recommend to only put loctite on the second bearing because the first one is in contact with the torque sensor and it is better to let it move a little bit sideways.