New "TSDZ2 Torque Sensor Central Motor"

Yes it's Not correct, any cyclist whith other ebike.
I will try to stay below 500w but when the temperature rises I don't think I can solve the problem. in off road it is also practically impossible unless you go very slow and for a longer period of time that inevitably leads to overheating of the engine. I will do some tests and publish the results
 
andrea_104kg said:
Yes it's Not correct, any cyclist whith other ebike.
I will try to stay below 500w but when the temperature rises I don't think I can solve the problem. in off road it is also practically impossible unless you go very slow and for a longer period of time that inevitably leads to overheating of the engine. I will do some tests and publish the results
Ah ok, if you are comparing your bicycle with TSDZ2 motor VERSUS any other big brand ebike, than yes, it would be like motor power wars, to see which one has more and more power.

If your ebikes stay on the 250W legal limit, then TSDZ2 is perfect capable.
 
andrea_104kg said:
Yes it's Not correct, any cyclist whith other ebike.
I will try to stay below 500w but when the temperature rises I don't think I can solve the problem. in off road it is also practically impossible unless you go very slow and for a longer period of time that inevitably leads to overheating of the engine. I will do some tests and publish the results
Ah ok, if you are comparing your bicycle with TSDZ2 motor VERSUS any other big brand ebike, than yes, it would be like motor power wars, to see which one has more and more power.

If your ebikes stay on the 250W legal limit, then TSDZ2 is perfect capable.
[/quote]
Completely agree. Stop trying to use an assist based product as a pedelec. Physics don't lie. If the TDSZ2 is pumping 250W into your ride this is the edge of what a fit rider can output by themselves. Who are these mithical 'fit' people you can't keep up with?
 
I also own a 2018 turbo levo, legal. I was very happy because the tsdz2 with the modified firmware has exactly the same performance. but if it overheats and needs to be used at half power it is no longer valid. the tsdz2 with original software is not even comparable to a modern mtb if you need to use the new firmware under the same conditions as the original where the advantage lies?
 
andrea_104kg said:
the tsdz2 with original software is not even comparable to a modern mtb if you need to use the new firmware under the same conditions as the original where the advantage lies?
If you know that TSDZ2 with original software is not even comparable to a modern mtb but if you are looking for that, I think you already know what you need to do -- use a different motor.

Here is a copy about some features of our flexible OpenSource firmware that may or may not be an advantage, depends on each user(I did copy the text from ElectrifyBike shop):

Features:
- Makes it possible to use more advanced displays with the TSDZ2 like the Bafang 850C color LCD or the KT-LCD3.
- Uses more sophisticated FOC algorithms allowing the TSDZ2 motor to run much more efficiently, making it more powerful while using less battery energy. The ebike will be also more responsive, fast and agile.
- Display shows more advanced information like battery voltage, rider pedal power in watts, motor temperature, cadence, trip time and distance.
- All options configurable directly on the LCD!! No programming cable required to make changes.
- Up to 9 assist levels configurable as a multiple of rider effort.
- Up to 9 boost levels to add extra power when starting and/or resuming pedaling.
- Multiple levels of walk assist for extra assist power on hills or with heavy loads or headwinds.
- Cruise control will maintain speed or accelerate to a preset speed.
- Allows use of batteries from 24 to 60 volts and allows batteries of different voltages to be changed on the fly.
- Detailed battery information including watts consumed and state of charge.
- Street legal and off-road modes for countries with restrictions. (This motor is legal at all levels in the USA)
 
My 36v 250w TSDZ2 fitted hybrid on the original firmware came up a climb yesterday in two-thirds of the time it takes me on my road bike; that's plenty quicker for me. How the motor copes with the modified firmware is surely dependent on how you set it up.
 
mctubster said:
andrea_104kg said:
the only thing is that the thermometer is broken, but it correctly reads the room temperature and also the body temperature. During the test I touched the body of the engine and I could hardly hold my hand, I think over 55 °, inside obviously the temperature is much higher. Unfortunately, in my hills, a 2 km climb is enough to reach and exceed 80 ° unless the sensor is broken ...
it is useless to control the current because I should do the whole tour with reduced power, since after 2km I am already above 80°, and with only 10° of ambience temperature :-(

Hi there.

I made this comment in an earlier post ... the motor is only small and not designed to run at high power other than short bursts.
Look at the size of it! If you are smashing up hills at the current/power limit yes it will overheat quickly.

Lower gears, higher cadence and lower speeds will solve the problem of course.

On high levels of assist you are not gaining much from the torque sensor and might as well go for a PAS motor that can handle your expectations better like the Bafang BBS02 or HD.

All the best
Agreed, seems like it's not disigned for heavy mountain biking and prolonged climbing, only for city biking with the height and weight of native people in mind.
 
Mike-P said:
Thanks for the info -is the rear light worked by the down button then? The one annoyance I find with the VLCD-5 is the lights being worked by the power on off button and consequently its seems too easy to switch off when wanting to switch the lights on.
I thought the latest version was 18.2 and the lights are very definitely working. There is a single cable that splits to power front and rear as well as the speed sensor - you have to buy it, though, unless ordered with the motor. I found that if you tried to take too much power off this cable eg by having both front and rear lights, the motor just refused and switched off. In any event I was getting fed up with wires running everywhere and there are some very cheap (a buck or two) rear lights running off AAA cells and hang off the seat post which work fine.
 
mctubster said:
andrea_104kg said:
the only thing is that the thermometer is broken, but it correctly reads the room temperature and also the body temperature. During the test I touched the body of the engine and I could hardly hold my hand, I think over 55 °, inside obviously the temperature is much higher. Unfortunately, in my hills, a 2 km climb is enough to reach and exceed 80 ° unless the sensor is broken ...
it is useless to control the current because I should do the whole tour with reduced power, since after 2km I am already above 80°, and with only 10° of ambience temperature :-(

Hi there.

I made this comment in an earlier post ... the motor is only small and not designed to run at high power other than short bursts.
Look at the size of it! If you are smashing up hills at the current/power limit yes it will overheat quickly.

Lower gears, higher cadence and lower speeds will solve the problem of course.

On high levels of assist you are not gaining much from the torque sensor and might as well go for a PAS motor that can handle your expectations better like the Bafang BBS02 or HD.

All the best

Found that it's easier to keep it on the lowest 28t gear and pedal it in short bursts while in turbo mode and then just let it coast. That way will have to worry a bit less about downshifting on average.
 
could someone please explain to me in plain english if i can just flash my tsdz2 36v motor 500w so it will run 48v batteries.

im happy with my controller (vlcd6)

every answer or thing ive read keeps talking about screen options and such.

i dont think it helps that the motor controller and the thing on the handle bars are both called controllers.

do i flash the open source firmware or the stock 48v firmwear?

i have a ts link a soldering iron and im willing to open up my speed sensor cable to make a flashing lead.

the big question is will this work with my totally standard vlcd6

am i going to flash the motor and then it not work because i need the ktlcd3 to make changes to the voltage or some other thing?

Edit

Started reading up on marqoc's version of the open source motor firmware specifically for the vlcd6

Looks very promising. :)
 
having a bad time trying to download st visual programmer to use with the st link

i go to the site fill in my name and email

open the email and click on the download now button

it sends me back to the site and nothing downloads.

help please.

Edit.

Sorted wouldnt download in google chrome. Works fine in explore..
1 more piece of the puzzle in place :)
 
sysrq said:
Does anybody know how the tsdz2 torque sensor rotary transformer can work only with two wires?
https://www.omega.com/literature/transactions/volume3/force3.html
http://www.sensorland.com/HowPage075.html
I showed it to a very smart electrical engineer at work, and he was impressed by the simplicity.

Here is what he said in non-technical language:

The two coils are, as you said, a transformer. An alternating current on the bike side induces an alternating current in the rotating coil. The controller feeds an alternating current to the fixed coil. It has to be alternating current because transformers only work with alternating current.

The fact one coil is rotating and the other is stationary makes no difference. The power across the transformer is unchanged with rotation.

The distance between the coils is important. That is why they are faced with low friction material and rub. Contact keeps the distance constant and is not a problem at crank rpm levels.

The torque is sensed by a hall effect sensor and a magnet, arranged and mounted so torque on the cranks flexes the steel in a way that moves the two with respect to each other. A hall sensor and magnetic can be very sensitive to very small movements.

The sensor sensor draws more and less power with magnet position. In effect the sensor is an electrical load that varies with torque. The load is powered by the current passed through the transformer.

The controller reads the torque by measuring the current going to the sensor. Since electrical load varies with torque current varies as well. So the signal is the power draw. Clever! No separate signal connection needed.

In the most basic terms:
1. Use a transformer to carry power across a moving interface without sliding contacts.
2. Make an AC torque sensor that varies electrical load with torque.
3. Measure current to the sensor.

It is well designed for an ebike, robust and inexpensive. It is not how an instrumentation grade sensor would be made. Hall effect sensors are susceptible to nearby magnetic fields, strain gages would be more linear and resistant to magnetic interference. The rubbing coils might bounce or generate heat at high rpm, precision guidance and an air gap could be added. The mechanical design of the axle could be improved to improve linearity. But these changes cost money and add weight and complexity. Accuracy is fine for ebike assist. Many ebike torque sensors make worse compromises.

Some only measure torque applied to one crank, the TSDZ2 sensors measures both. Some ebike sensors measure bending forces on the axle as well as torque. The TSDZ2 mechanical isolates torque pretty well. If it was being sold as a separate torque sensor I think it would be towards the high end.

A lot of good electrical and mechanical design went into the TSDZ2.
 
tomtom50 said:
sysrq said:
Does anybody know how the tsdz2 torque sensor rotary transformer can work only with two wires?
https://www.omega.com/literature/transactions/volume3/force3.html
http://www.sensorland.com/HowPage075.html
I showed it to a very smart electrical engineer at work, and he was impressed by the simplicity.

Here is what he said in non-technical language:

The two coils are, as you said, a transformer. An alternating current on the bike side induces an alternating current in the rotating coil. The controller feeds an alternating current to the fixed coil. It has to be alternating current because transformers only work with alternating current.

The fact one coil is rotating and the other is stationary makes no difference. The power across the transformer is unchanged with rotation.

The distance between the coils is important. That is why they are faced with low friction material and rub. Contact keeps the distance constant and is not a problem at crank rpm speeds.

The torque is sensed by a hall effect sensor and a magnet, arranged and mounted so torque on the cranks flexes the steel in a way that moves the two with respect to each other. A hall sensor and magnetic can be very sensitive to very small movements.

The sensor sensor draws more and less power with magnet position. In effect the sensor is an electrical load that varies with torque. The load is powered by the current passed through the transformer.

The controller reads the torque by measuring the current going to the sensor. Since electrical load varies with torque current varies as well. So the signal is the power draw. Clever! No separate signal connection needed.

In the most basic terms:
1. Use a transformer to carry power across a moving interface without sliding contacts.
2. Make an AC torque sensor that varies electrical load with torque.
3. Measure current to the sensor.

It is well designed for an ebike, robust and inexpensive. It is not how an instrumentation grade sensor would be made. Hall effect sensors are susceptible to nearby magnetic fields, strain gages would be more linear and resistant to magnetic interference. The rubbing coils might bounce or generate heat at high rpm, precision guidance and an air gap could be added. The mechanical design of the axle could be improved to improve linearity. But these changes cost money and add weight and complexity. Accuracy is fine for ebike assist. Many ebike torque sensors make worse compromises.

Some only measure torque applied to one crank, the TSDZ2 sensors measures both. Some ebike sensors measure bending forces on the axle as well as torque. The TSDZ2 mechanical isolates torque pretty well. If it was being sold as a separate torque sensor I think it would be towards the high end.

A lot of good electrical and mechanical design went into the TSDZ2.
 
Thanks for sharing. I just added to a wiki page: https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/How-TSDZ2-torque-sensor-works
 
casainho said:
Thanks for sharing. I just added to a wiki page: https://github.com/OpenSource-EBike-firmware/TSDZ2_wiki/wiki/How-TSDZ2-torque-sensor-works

Yes TomTom50, thanks for sharing "How the TSDZ2 Torque Sensor Works". Very interesting. You do plop down these real nuggets occasionally. :)
 
la8rat said:
could someone please explain to me in plain english if i can just flash my tsdz2 36v motor 500w so it will run 48v batteries.

im happy with my controller (vlcd6)

every answer or thing ive read keeps talking about screen options and such.

i dont think it helps that the motor controller and the thing on the handle bars are both called controllers.

do i flash the open source firmware or the stock 48v firmwear?

i have a ts link a soldering iron and im willing to open up my speed sensor cable to make a flashing lead.

the big question is will this work with my totally standard vlcd6

am i going to flash the motor and then it not work because i need the ktlcd3 to make changes to the voltage or some other thing?

Edit

Started reading up on marqoc's version of the open source motor firmware specifically for the vlcd6

Looks very promising. :)

First of all the thing on your handlebars is a display, not a controller. It has a microcontroller in it but that is just a small dedicated CPU.
The easiest way to use a 48v battery with a 36v TSDZ2 is to flash the TSDZ2 controller with 48v OEM firmware available here.

The other way to keep your VLCD6 display is to use the Marcoq version as you are investigating. This is more difficult to flash because you must set up a complete development environment to compile the firmware instead of just flashing precompiled firmware with the STLINK as you do with the OEM firmware or the FOS firmware with the KT-LCD3. I hope that was clear. :)
 
andrea_104kg said:
But My original questione was if there is a Way To cooling The motor Not a discussion about temp :)
I'm not sure if what you desire can be provided by the TSDZ2, but to reduce the temperature of the motor, the heat generated by the motor needs to be coupled to the (relatively) cool ambient air.

The air between the motor and the cover is not a great conductor of heat, so if you can fill some of this space with something that does conduct heat well (aluminium, copper, thermally conductive paste) that will transfer motor heat to the cover, which is exposed to the air passing the cover.
Ideally the cover will be close to the same temperature as the motor inside it. The hotter the cover is, the more heat energy from the motor can be dissipated to the environment.
Then you could also add more surface area (fins) to the cover in order to transfer even more heat to the environment. Just like a radiator the more surface area, temperature difference and air passing, the more heat you can transfer.

The best way is probably to coat the inside of the cover with some thermally conductive (electrically non-conducting) paste and fill the entire space between the motor and the cover.

The cheapest way would be to use aluminium foil to fill the space and avoid contact with the electrically conducting parts.

That said, given your description of the TSDZ2 performance, I do wonder if you haven't already destroyed the magnets within the motor, through overheating the motor magnets at some point?
That would cause a loss of efficiency, a lot of the energy going to the motor, to produce relatively little motor performance, thus causing the motor to consume more energy, while still not achieving the performance you want, so generating a lot of heat instead. A new motor might help more than anything?

Good luck and let us know how much difference it makes.
 
ThousandWax said:
That said, given your description of the TSDZ2 performance, I do wonder if you haven't already destroyed the magnets within the motor, through overheating the motor magnets at some point?
That would cause a loss of efficiency, a lot of the energy going to the motor, to produce relatively little motor performance, thus causing the motor to consume more energy, while still not achieving the performance you want, so generating a lot of heat instead. A new motor might help more than anything?
That is a possibility, as I documented on the wiki page, when I got about only about half of the motor torque with the same regular power applied. This is why it is really important to install the temperature sensor.
 
What is, in your experience of fried motors and temperature readings, the safe Amperage we can set for the TSDZ2 in order to keep the temperature under acceptable limits?
I've actually set 16A (with 36v means about 500w) and does not seem the motor heats up much. I use low gears when climbing, such as 30T chain wheel and 34 at the back, and I keep "cycling speed".. so I think I keep the motor in a good range but I would ask your valuable opinion on the matter.
Thanks
 
thineight said:
What is, in your experience of fried motors and temperature readings, the safe Amperage we can set for the TSDZ2 in order to keep the temperature under acceptable limits?
I've actually set 16A (with 36v means about 500w) and does not seem the motor heats up much. I use low gears when climbing, such as 30T chain wheel and 34 at the back, and I keep "cycling speed".. so I think I keep the motor in a good range but I would ask your valuable opinion on the matter.
Thanks
Unfortunately it isn't so simple as safe / unsafe amperage. Motor temperature is commonly controlled by watching I2t (Current squared with duration factored in).
http://doc.ingeniamc.com/emcl2/command-reference-manual/protections/i2t-protection

This type of watchdog function is built into industrial motor controllers. It isn't built into general purpose ebike controllers or the TSDZ2 (who knows, maybe integrated systems like Bosch or Shimano fold it in).

Long story short sustained high amperage can toast your motor where short bursts won't. So climbing a hill for five minutes at 18A might be fine where climbing a pass at 18A will Kil the motor.

And sustaining 32 mph for half an hour on the flat? I wouldn't do it without a temp sensor.
 
Thanks, sounds reasonable. I hope not to mess up anything with my usage.
 
For what I remember, my 48V motor, could sustain about 10 amps continuously but above that would pass the 85 degrees that I consider the limit.

When I demagnetized my motor, I was pushing at max current (like 18 amps) because I wanted to discharge my battery pack as fast as I could (I wanted to measure the amount of watts / hour and I was short of time to do that).
 
I hope to explain myself. I don't want a motorcycle or exceptional performance. however, I have the strong doubt that if the engine arrives quickly above 90 degrees in an uphill country road with 10 degrees outside temperature as soon as the temperature rises it will be unusable. I am therefore looking for solutions. since I don't use the bike in the rain I don't make river fords I was thinking of a small 5v fan. what do you think?
 
Back
Top