New "TSDZ2 Torque Sensor Central Motor"

[casainho]

Just to give an update on the TSDZ2 EBike wireless project, with some good news because my Garmin GPS cycling computer now detects and connects to the TSDZ2 wireless board!! -- all the information on the project page: https://github.com/OpenSource-EBike-firmware/TSDZ2_wireless

Very first testing firmware that works: the wireless board nRF52840 is implementing the ANT+ LEV Ebike profile and the Garmin Edge cycling computer is detecting and connection to the EBike wireless board. Also, when starting an EBike activity, there is available a screen specifically to see EBike data and, is possible to select the assist level.

[youtube]YYaG6Zjt1xo[/youtube]

 

[szchz]

I've seen some references to this generally.

What constant power output is sustainable without overheating in ~ 20 degree ambient temp (my motor gives ~800 wats peak) ?

I'm anxious about the motor on hills without a thermometer installed.

I was thinking about dialling down the amperage, any advice ? I irrationally dropped it to 14, I don't know if that will really have an impact.

 

[Mackiezn]

Any idea what might of caused my torque sensor to disintegrate? It had been working well for about a year but then one day it would only assist full power and then not stop immediately when I stopped pedaling. I then reset it by unplugging the battery and plugging it in again and it would then only give me probably the equivalent to two bars of power. So I then took it home and opened it up and found it had been completely destroyed

 

[Balmorhea]

Awesome! It’s engineered like... something that wasn’t engineered at all.

 

[Waynemarlow]

Price wise its an awesome bit of engineering. Both units I've seen damaged were very old units and the design of the free floating sensor is different and less substantial than the latter ones. Both units also were regularly used in muddy and wet conditions and had water ingress stains and both were also known to have excess float ( again later ones have shims fitted to prevent this sideways float on the crank ). It all added up to failure.

The later motors have been pretty reliable.

 

[Retrorockit]

Awesome! It’s engineered like... something that wasn’t engineered at all.

New meaning to the term "reverse engineering".

 

[szchz]

Price wise its an awesome bit of engineering. Both units I've seen damaged were very old units and the design of the free floating sensor is different and less substantial than the latter ones. Both units also were regularly used in muddy and wet conditions and had water ingress stains and both were also known to have excess float ( again later ones have shims fitted to prevent this sideways float on the crank ). It all added up to failure.

The later motors have been pretty reliable.

Do you have any recommendations for reducing this risk? I live in the Pacific Northwest. Maybe some sealant around the engine case ?

 

[Waynemarlow]

Sadly there's not much you can do to limit water and dust ingress in severe operational conditions other than what the factory already does, simply the motor was not designed for deep water nor deep muddy conditions, but then nor were the early Bosch, Brose, Shimano engines and one doesn't have to look far to see large numbers of failures from these reputed manufacturers in the same conditions.

Where the main problem is, is the small hole which the torque sensor wires enter the cavity where the motor and electronics sit. Its very high up and in normal situations the outer chain ring side cover and its seal should deal with the normal road water and grime the motor was designed for. However in real muddy conditions we have seen that cover worn away and the bearing seals failing with water and dirt eventually getting behind the cover and then up through that hole with the inevitable eventual electronics failure.

We have so far simply greased the seals on the cover with a liberal coating of silicon or waterproof grease before assembly and filling the area around where the main instrument and speed sensor wires enter that cavity with the same silicon grease as although that area is siliconed at the factory, its quite poorly done and to fill around the area with grease simply prevents water and dirt getting to that silicon.

Also just make sure you change that outer chain ring side bearing more regularly than needed, the seals need minimum movement to work correctly.

 

[Doohickey]

I've seen some references to this generally.

What constant power output is sustainable without overheating in ~ 20 degree ambient temp (my motor gives ~800 wats peak) ?

I'm anxious about the motor on hills without a thermometer installed.

I was thinking about dialling down the amperage, any advice ? I irrationally dropped it to 14, I don't know if that will really have an impact.

AFAIK (someone correct me if I'm wrong) cnanging the amps setting on the stock display does not do anything (you could test by changing it to something really low and taking it up a hill).

This motor is definitely a 250w motor (that can briefly burst to 750..). In my experience, I would not take it over 250w continuous.

I think the primary market for Tongsheng is ebike resellers and manufacturers, e.g. I have seen cheapish bikes in supermarkets here with tsdz2 on them, those are limited to 250w (euro-legal). I think Tongsheng may have decided to up the power in the software, and then "hey, we can sell a few more of these on Aliexpress and the like". It will definitely cook itself at 750w i think. Even though it should be possible for the stock software to limit the temperature even without temp sensor (just needs to keep track of how many watt-hours have been used over the last x minutes, then reduce power when it exceeds some threshold.) I don't think it does that though.

As an example: on a 900m hill, with 60m of climb (i.e. quite steep), it gets to 70c going up, after that I can only use 150-200w or the temperature goes up more. It also takes quite a while to cool down again.

Ambient temperature makes a big difference - much more continuous power at 5 - 10 degress ambient, than now at 30.

I installed the temp sensor (along with the open source firmware), and having it there is quite reassuring. I often hit the limit that I've set, and it reduces the power (lots of hills here too).

The cooling modification described in the wiki also helps.

 

[izeman]

As an example: on a 900m hill, with 60m of climb (i.e. quite steep), it gets to 70c going up, after that I can only use 150-200w or the temperature goes up more. It also takes quite a while to cool down again.

This is how 250W nominal (aka EU-legal regulation) is defined. It will do 250W all day. The maximum is not defined. It's up to you to set the upper limit to a value where it will survive, if you set it too high it will over heat though - which is expected for a 250W motor. It's totally legal to have 700-800W peak in EU-pedelec-regulation, as long as the motor won't survive it, and you obey the 25km/h motor assisted limit.

 

[szchz]

waynemarlow: those are all awesome and actionable tips that are simple enough to do, I'll mainly be doing commuting (intention is for it to be a family grocery/day care/ day trips bike), at this point I don't plan on taking it off road or in waterlogged conditions so hopefully it lasts.

I read this article:
https://www.electricbike.com/tsdz2-750w-mid-drive-torque-sensing/
and it was recommended to drill a 1/8th diameter hole at the center of the bottom of the cover to allow any condensation and humidity to escape, would that be useful?

doohikey.. I have the 860c display. I'll try to reduce it to a really low setting and report back if it works, Injured my back this morning lifting some heavy boxes so taking a rest day . Yeah, I saw those recommendations, I'll have to get around to it eventually. Yeah I have to port the open source firware, looking forward to it. Thanks for the info on temp, from what I infer running the bike continuous at 300 - 450 watts for say 15 min would be highly unrecommended (48v, 500 watt version) and you'd recommend to keep it below 250 watts for continuous riding.

izeman.. thanks.

 

[Waynemarlow]

waynemarlow: those are all awesome and actionable tips that are simple enough to do, I'll mainly be doing commuting (intention is for it to be a family grocery/day care/ day trips bike), at this point I don't plan on taking it off road or in waterlogged conditions so hopefully it lasts.

I read this article:
https://www.electricbike.com/tsdz2-750w-mid-drive-torque-sensing/
and it was recommended to drill a 1/8th diameter hole at the center of the bottom of the cover to allow any condensation and humidity to escape, would that be useful?

doohikey.. I have the 860c display. I'll try to reduce it to a really low setting and report back if it works, Injured my back this morning lifting some heavy boxes so taking a rest day . Yeah, I saw those recommendations, I'll have to get around to it eventually. Yeah I have to port the open source firware, looking forward to it. Thanks for the info on temp, from what I infer running the bike continuous at 300 - 450 watts for say 15 min would be highly unrecommended (48v, 500 watt version) and you'd recommend to keep it below 250 watts for continuous riding.

izeman.. thanks.

For what you want to do the motor will do just fine, I would avoid drilling any hole in any cover as all it will do is promote condensation, the unit has enough volume in cavities to not be pressurised and the heat of the motor should eliminate most condensation.

Mmmm there's a lot of misconception on heat and over heating the units here, 36 volt units do get hot running them on 36 volts, try running them on 48 volts and even better on 52 volts, it transforms a pretty unremarkable road sub 80 cadence engine into a real little bargain gem that you can push along at 450 - 500 W constant for some time.

48 volt units, are just grunt torquey engines until you run them at 52 volts or on OSF where they come alive a bit. We find the 36 volt engine on 52 volts the nicest of the bunch if you are a regular rider and the 48 volts makes your partner who is an occasional rider, feel like Lance Armstrong.

Guys you have to put things into context, you can by 3 TSDZ2's for the price of a Brose, just accept that at times you are going to trash a motor.

 

[szchz]

waynemarlow: those are all awesome and actionable tips that are simple enough to do, I'll mainly be doing commuting (intention is for it to be a family grocery/day care/ day trips bike), at this point I don't plan on taking it off road or in waterlogged conditions so hopefully it lasts.

I read this article:
https://www.electricbike.com/tsdz2-750w-mid-drive-torque-sensing/
and it was recommended to drill a 1/8th diameter hole at the center of the bottom of the cover to allow any condensation and humidity to escape, would that be useful?

doohikey.. I have the 860c display. I'll try to reduce it to a really low setting and report back if it works, Injured my back this morning lifting some heavy boxes so taking a rest day . Yeah, I saw those recommendations, I'll have to get around to it eventually. Yeah I have to port the open source firware, looking forward to it. Thanks for the info on temp, from what I infer running the bike continuous at 300 - 450 watts for say 15 min would be highly unrecommended (48v, 500 watt version) and you'd recommend to keep it below 250 watts for continuous riding.

izeman.. thanks.

For what you want to do the motor will do just fine, I would avoid drilling any hole in any cover as all it will do is promote condensation, the unit has enough volume in cavities to not be pressurised and the heat of the motor should eliminate most condensation.

Mmmm there's a lot of misconception on heat and over heating the units here, 36 volt units do get hot running them on 36 volts, try running them on 48 volts and even better on 52 volts, it transforms a pretty unremarkable road sub 80 cadence engine into a real little bargain gem that you can push along at 450 - 500 W constant for some time.

48 volt units, are just grunt torquey engines until you run them at 52 volts or on OSF where they come alive a bit. We find the 36 volt engine on 52 volts the nicest of the bunch if you are a regular rider and the 48 volts makes your partner who is an occasional rider, feel like Lance Armstrong.

Guys you have to put things into context, you can by 3 TSDZ2's for the price of a Brose, just accept that at times you are going to trash a motor.

Cheers. Yup, great value, its been a treat thus far.

 

[sysrq]

waynemarlow: those are all awesome and actionable tips that are simple enough to do, I'll mainly be doing commuting (intention is for it to be a family grocery/day care/ day trips bike), at this point I don't plan on taking it off road or in waterlogged conditions so hopefully it lasts.

I read this article:
https://www.electricbike.com/tsdz2-750w-mid-drive-torque-sensing/
and it was recommended to drill a 1/8th diameter hole at the center of the bottom of the cover to allow any condensation and humidity to escape, would that be useful?

doohikey.. I have the 860c display. I'll try to reduce it to a really low setting and report back if it works, Injured my back this morning lifting some heavy boxes so taking a rest day . Yeah, I saw those recommendations, I'll have to get around to it eventually. Yeah I have to port the open source firware, looking forward to it. Thanks for the info on temp, from what I infer running the bike continuous at 300 - 450 watts for say 15 min would be highly unrecommended (48v, 500 watt version) and you'd recommend to keep it below 250 watts for continuous riding.

izeman.. thanks.

For what you want to do the motor will do just fine, I would avoid drilling any hole in any cover as all it will do is promote condensation, the unit has enough volume in cavities to not be pressurised and the heat of the motor should eliminate most condensation.

Mmmm there's a lot of misconception on heat and over heating the units here, 36 volt units do get hot running them on 36 volts, try running them on 48 volts and even better on 52 volts, it transforms a pretty unremarkable road sub 80 cadence engine into a real little bargain gem that you can push along at 450 - 500 W constant for some time.

48 volt units, are just grunt torquey engines until you run them at 52 volts or on OSF where they come alive a bit. We find the 36 volt engine on 52 volts the nicest of the bunch if you are a regular rider...

Isn't the best efficiency at rated motor speed (4000rpm)? Probably will gonna leave the 36V version as it is for now despite having access to ST programmer.

 

[knutselmaaster]

Brushless motors generally have the highest efficiency around 85% of max rotation speed.

 

[Waynemarlow]

Do we know what the max rpm for the motors are ?

 

[knutselmaaster]

As firmware changes can alter the max rpm, it seems that the speed is limited by the controller...
One would need to take the motor out of the TSDZ2 and run it with a controller that permits it to run freely without limitation, I suppose.
I'm just speculating here though.

 

[knutselmaaster]

As firmware changes can alter the max rpm, it seems that the speed is limited by the controller...
One would need to take the motor out of the TSDZ2 and run it with a controller that permits it to run freely without limitation, I suppose.
I'm just speculating here though.

 

[Mackiezn]

Also just make sure you change that outer chain ring side bearing more regularly than needed, the seals need minimum movement to work correctly.

Are you referring to the oneway bearings in the main gear?

 

[Waynemarlow]

Yes, the seals on them from the factory are just dust seals and the waterproofing of the whole engine relies on the small seal in the outer chain ring holder and the O ring on the bearing face itself. Once the bearing wears, the axle will move up and down a bit wearing that outer seal out pretty quickly if you start putting it in a muddy environment.

Early motors didn't actually have a full seal on the chainring carrier and only had a dust seal. Those with early motors, a 104bcd carrier is only 25 notes, it will prolong the life of that bearing enormously.

 

[Retrorockit]

Something from my truck maintenance background is that when a hot sealed housing encounters cold water the air inside contracts and causes a suction that draws water into the housing.
Maybe a vent tube that runs up high away from any water could help? Something to try.

 

[spinningmagnets]

This was one of the more interesting discoveries concerning moisture migration on a humid day, where rust would develop inside a hubmotor. As the motor cooled down, humid air would be drawn in through the cable bundle of motor phase wires and hall sensor wires. As the motor cooled even more overnight the trapped humid air would condense into liquid droplets.

Adding a vent to the sideplates and sealing the cable with goo seems to fix that. Also, spraying the insides of the motor and controller with a thin conformal dielectric spray has proven to be useful.

The next time the motor gets hot, the trapped liquid from the previous day vaporizes and flows out the vent.

 

[knutselmaaster]

Very useful information indeed. This might be helpful also for displays that show condensation on the inside.

 

[Waynemarlow]

This was one of the more interesting discoveries concerning moisture migration on a humid day, where rust would develop inside a hubmotor. As the motor cooled down, humid air would be drawn in through the cable bundle of motor phase wires and hall sensor wires. As the motor cooled even more overnight the trapped humid air would condense into liquid droplets.

Adding a vent to the sideplates and sealing the cable with goo seems to fix that. Also, spraying the insides of the motor and controller with a thin conformal dielectric spray has proven to be useful.

The next time the motor gets hot, the trapped liquid from the previous day vaporizes and flows out the vent.

Agree in an open vented system.

In an unvented, as the TSDZ2 should be, with plenty of volume of spare space inside the sealed system, heat should not have much effect on the atmospheric pressure within the motor and shouldn't be able to draw in moisture from the external air mass.

I would suggest that new the TSDZ2 engine is an unvented structure, some time and km's down the road it becomes a vented system, so do you drill a hole and hope that you do not get water ingress from the mud and rain under pressure from the outside as we hit puddles and the likes as we get in the UK, from new, or do you rely on the manufacturers design and take the unvented route until it becomes a vented system where it will vent back out through the same area that the moisture originally entered from.

Mmmm conundrum and until the latest Mahle engine used by Specialized where it has a Kevlar bladder to take expansion and contraction just for this purpose, all bike engines were experiencing the same problem, did Bosch, Shimano, Brose, Yamaha have a drain port on their engines, umm no. But then an awful lot of those failed with water inhalation. Mmmm even bigger conundrum.

 

[sysrq]

Very useful information indeed. This might be helpful also for displays that show condensation on the inside.

Changing desiccant every 12 months would be too much to ask for.