New "TSDZ2 Torque Sensor Central Motor"

Hi everyone, first message in the forum :)

Sorry for the noob questions, can´t manage to find the answer in the thread...

Just bought a 36v TSDZ2 version and I want to flash it with OSF firmware and connect to a 48v 13S5P battery.

- Should I change the motor setting to 48v or can I connect the 48v battery without issues?
- What "Battery low voltage cut off" value should I set?
 
Doohickey said:
Hudson said:
1. The back wheel will not free wheel at all, it feels very stiff, if i want to reverse I have to lift the back wheel. Got the bike on the stand now and can rotate the wheel backwards but it is not smooth. It rotates less than one revolution and then gets very stiff and then frees up, then rotates the same again before getting stiff.

I think you'll have to take apart the motor to to figure anything out. Since it doesn't spin properly, this sounds to me like something broken / snapped / stripped / stuck (possibly the blue gear?). I'd start off by taking off the non-drive side cover to access the inner motor and blue gear (IMO this is one of the easier things to check, depending on your bike this can be done without even taking the crank off, just take off the pedal, slide the locking tool over the crank arm and loosen the big nut); check that everything in there looks good.

Beyond that the main drive gear is also relatively easy to access to check, you need to take off the spider and chairing I believe.

Be careful since both the outer hex screws as well as the inner phillips ones holding the motor unit strip easily - use a good tool and try to avoid letting it slip out as you undo them, and don't overtighten when putting back together. (I replaced mine with torx)

Thanks for the advice, I dismantled the motor and found the nylon gear was damaged, i had a spare metal one (think I still prefer the nylon one but its what i have here to test with) This fixed the issue with the motor jamming when spinning the wheel backwards. I then tried the motor out and still getting the same issue, Error 3 torque sensor error. I am running the 1.1.0 code if i remember correctly this error may be incorrectly titled and the error may be something else? I have the bike on a stand and when i engage walk mode the motor engages but is very week and not smooth, then if i put the rear brake on very gently to simulate the bike operating in normal walk mode rather than being on a stand the E3 message appears.

casainho said it could be one of the following earler, so i guess i need to check the following but i am not sure which order to try them in

"Maybe fail hall sensor, then or a motor phase wire and or burned motor controller power mosfets... Please report later."
 
Hudson said:
Thanks for the advice, I dismantled the motor and found the nylon gear was damaged, i had a spare metal one (think I still prefer the nylon one but its what i have here to test with) This fixed the issue with the motor jamming when spinning the wheel backwards. I then tried the motor out and still getting the same issue, Error 3 torque sensor error. I am running the 1.1.0 code if i remember correctly this error may be incorrectly titled and the error may be something else? I have the bike on a stand and when i engage walk mode the motor engages but is very week and not smooth, then if i put the rear brake on very gently to simulate the bike operating in normal walk mode rather than being on a stand the E3 message appears.

casainho said it could be one of the following earler, so i guess i need to check the following but i am not sure which order to try them in

"Maybe fail hall sensor, then or a motor phase wire and or burned motor controller power mosfets... Please report later."

The hall sensors are quite easy to check wuth a multimeter. Measure the voltage from the hall connector between red or black and the rest of the colors, you should get consistent change of voltage from each one when you turn the motor.

If the phase wires were properly attached and show no signs of damage they are probably fine.

I don't know if theres any way to check the MOSFETs without removing the potting, you might need to get another controller.
 
Having weird torque sensor issues on a new motor. Using the 860c to calibrate - either the ADC value is really high (900+) and slowly falls. Or if I can get it to stabilize at, say 170 with no load on the pedals.. Then if I apply load - the value increases but very slightly - e.g. with me standing on the pedals (105kg) maybe rises to 180 - then takes a long time to fall back to the starting value...

Anyone seen similar? Any ideas other than to replace torque sensor/motor. Sod's law it's a motor i've already installed the temp sensor on - so no hope of any warranty...
 
beemac said:
Having weird torque sensor issues on a new motor. Using the 860c to calibrate - either the ADC value is really high (900+) and slowly falls. Or if I can get it to stabilize at, say 170 with no load on the pedals.. Then if I apply load - the value increases but very slightly - e.g. with me standing on the pedals (105kg) maybe rises to 180 - then takes a long time to fall back to the starting value...

Anyone seen similar? Any ideas other than to replace torque sensor/motor. Sod's law it's a motor i've already installed the temp sensor on - so no hope of any warranty...
You really need to check your torque sensor, I saw that kind of issues with a broquen one...
 
casainho said:
beemac said:
Having weird torque sensor issues on a new motor. Using the 860c to calibrate - either the ADC value is really high (900+) and slowly falls. Or if I can get it to stabilize at, say 170 with no load on the pedals.. Then if I apply load - the value increases but very slightly - e.g. with me standing on the pedals (105kg) maybe rises to 180 - then takes a long time to fall back to the starting value...

Anyone seen similar? Any ideas other than to replace torque sensor/motor. Sod's law it's a motor i've already installed the temp sensor on - so no hope of any warranty...
You really need to check your torque sensor, I saw that kind of issues with a broquen one...

Thanks, yes I think sadly that's looking like it's the case - I've tried lots of things and it's still the same... have another motor on order then I'll dig into this one and see if I can find the problem.

What's interesting though is how the motor behaves.... Turning the pedals briefly the motor runs (fast) and stops - but turning for more than about 3/4 of a turn it runs away at a crazy rate only stopping when you use the brakes - on the 860 the motor rpm and pwm values go into the red - riding it is quite alarming, the acceleration is pretty dramatic. Motor heats up pretty quick too...

I wonder if there's a filter/threshold on the torque adc that's not working properly - as if you didn't have good brakes/brake sensors - it would be quite a rollercoaster :)
 
Hudson said:
"Maybe fail hall sensor, then or a motor phase wire and or burned motor controller power mosfets... Please report later."

If you disconnect the inner motor and check the three phase wires with multimeter, all combinations should show the same resistance relative to each other (if I'm correct..). If you spin the motor by hand, it should have a "magnetic" feeling, which goes away and it turns smoothly when you short the three wires together.

Otherwise it could be the controller, or hall sensor, (or something else...)
 
beemac said:
casainho said:
beemac said:
Having weird torque sensor issues on a new motor. Using the 860c to calibrate - either the ADC value is really high (900+) and slowly falls. Or if I can get it to stabilize at, say 170 with no load on the pedals.. Then if I apply load - the value increases but very slightly - e.g. with me standing on the pedals (105kg) maybe rises to 180 - then takes a long time to fall back to the starting value...

Anyone seen similar? Any ideas other than to replace torque sensor/motor. Sod's law it's a motor i've already installed the temp sensor on - so no hope of any warranty...
You really need to check your torque sensor, I saw that kind of issues with a broquen one...

Thanks, yes I think sadly that's looking like it's the case - I've tried lots of things and it's still the same... have another motor on order then I'll dig into this one and see if I can find the problem.

What's interesting though is how the motor behaves.... Turning the pedals briefly the motor runs (fast) and stops - but turning for more than about 3/4 of a turn it runs away at a crazy rate only stopping when you use the brakes - on the 860 the motor rpm and pwm values go into the red - riding it is quite alarming, the acceleration is pretty dramatic. Motor heats up pretty quick too...

I wonder if there's a filter/threshold on the torque adc that's not working properly - as if you didn't have good brakes/brake sensors - it would be quite a rollercoaster :)

I've replaced the motor controller and now the torque ADC is reading sensible values - so looks like sensor was ok..

However - I now cannot get any assist whatsoever in any level. Walk assist works ok - so not a motor problem.

Looks like a software issue - but I've flashed back to release versions - no change...

I have mucked about with settings a lot whilst I was trying to work out what was going on - so that's my primary suspect - but i've put those back (and cleared app cache to boot). Brake sensors all working properly - otherwise walk assist wouldn't work...

What setting/combo of settings would cause no assist in any level...

This is running the wireless controller - but didn't want to jump threads :)

edit: I'll move this to teh wireless thread - think it's to do with the wireless controller rather than a hardware issue...
 
Is there a graph of torque vs RPM available for the TSDZ2 ?
I would like to know at which RPM do we have peak power and peak efficiency. I assume the field weakening introduced in the latest versions of the firmware hasn't changed theses values.
 
On some units when reprogramming and I 've had no assist in any levels after progamming, I've had to reprogram the Options Byte page. Not sure why but it seems to work.
 
Waynemarlow said:
On some units when reprogramming and I 've had no assist in any levels after progamming, I've had to reprogram the Options Byte page. Not sure why but it seems to work.

Cheers - i think it's probably related to stuff I was testing with street mode... need to roll back some things.

Zeroing option a good idea - better than mine; to troubleshoot I copied the data/program/option bytes from my working bike and as soon as I flashed them the motor ran at 100% and could only be stopped with power off... was very glad it was up on the stand! :)
 
Yes as a routine I now copy 0's to both the program and memory pages to ensure that there are no bits in there which are not supposed to be. I then copy over the new program.

I have raised this issue before but really do not understand why we at times need to reprogram the Options page to get assist back.
 
Waynemarlow said:
Yes as a routine I now copy 0's to both the program and memory pages to ensure that there are no bits in there which are not supposed to be. I then copy over the new program.

I have raised this issue before but really do not understand why we at times need to reprogram the Options page to get assist back.

Turns out this was my issue - https://github.com/OpenSourceEBike/TSDZ2_wiki/wiki/FAQ#PAS_sensor_fail

Torque sensor working fine - but cadence sensor not reading anything... pushed down on the connector for the PAS sensor under all the rubber and started working... :)
 
Waynemarlow said:
Yes as a routine I now copy 0's to both the program and memory pages to ensure that there are no bits in there which are not supposed to be. I then copy over the new program.

I have raised this issue before but really do not understand why we at times need to reprogram the Options page to get assist back.

You should upgrade to the motor software 1.1.1 (https://github.com/OpenSourceEBike/TSDZ2_wireless/releases/download/v0.7.0/TSDZ2-v1.1.1.hex) - That contains some changes to the way serial is handled as as result of the epic 'assist level 3' issue I worked on - since then I've never had an issue where I had to reflash to get the motor working - whereas I had to do that a few times before. In earlier versions corrupt serial data - or incomplete data e.g. when you connected/reconnected the 6/8 pin cable powered on could potentially overwrite some of the program memory and cause unexpected effects.
 
Yay finally got my FSR build to a working state... Still waiting on a new battery than hopefully fits in the triangle - and I've yet to work out where to place the 1200W DC booster I'm using - have it set to 58.5v at the moment - not sure if I should push it and run 60v all the time - only used it on a couple of short rides with both the booster and battery in a rucksack - but can definitely notice the difference from a standard 48v battery. Looking forward to no voltage fade on longer rides :)

Wireless controller is on the handlebar - needs a better case but is nice to have the led where I can see it.

Build is currently 17.1kg without battery - not as light as I hoped it would be but reckon I can get it down a bit... was hoping for 18kg with battery but I think it's going to be more like 20 - maybe 19 if I can get some lighter tyres and change the handlebar+crank arms...

File_013.jpegFile_015.jpegFile_017.jpegFile_021.jpegFile_024.jpeg
 
beemac said:
Wireless controller is on the handlebar - needs a better case but is nice to have the led where I can see it.

Build is currently 17.1kg without battery - not as light as I hoped it would be but reckon I can get it down a bit... was hoping for 18kg with battery but I think it's going to be more like 20 - maybe 19 if I can get some lighter tyres and change the handlebar+crank arms...
Why don't you 3D print the wireless remote enclosure and use it? you would not use the coin battery and install the DC-DC converter near the TSDZ2 display connector.

That beautiful ebike deserves a small battery as his frame, to keep being beautiful.
For battery savings, you can make your own and save weight and volume by choosing the materials: a smaller BMS, enclosure materials and other parts.
Also very important is to have like one small / lighter and other big battery, so you can use the lighter one daily and the bigger one for some special day like for long ride. There are way to make a quick install / remove fix for the battery.
 
casainho said:
beemac said:
Wireless controller is on the handlebar - needs a better case but is nice to have the led where I can see it.

Build is currently 17.1kg without battery - not as light as I hoped it would be but reckon I can get it down a bit... was hoping for 18kg with battery but I think it's going to be more like 20 - maybe 19 if I can get some lighter tyres and change the handlebar+crank arms...
Why don't you 3D print the wireless remote enclosure and use it? you would not use the coin battery and install the DC-DC converter near the TSDZ2 display connector.

That beautiful ebike deserves a small battery as his frame, to keep being beautiful.
For battery savings, you can make your own and save weight and volume by choosing the materials: a smaller BMS, enclosure materials and other parts.
Also very important is to have like one small / lighter and other big battery, so you can use the lighter one daily and the bigger one for some special day like for long ride. There are way to make a quick install / remove fix for the battery.

I did think about that - but my smallest board is quite a bit larger than just the nrf - so I don't think it will fit your design as-is unfortunately - and after building four of these boards I'm not in a hurry to solder another :)

So I'll probably just print a simple case with a lens for the led for now and when I have to build another board will try to fit in the remote case.

Thanks I am very pleased with it - took a lot of searching to find the right frame!

Yea going to be interesting what the battery is like when it arrives - went for a wallen power one in a new slightly slimmer case than the standard shark. 14s4p - hoping for just over 3kg. Making my own batteries is definitely something that's on my list - is why I got the a30t printer with the larger print size but for now I'm using off the shelf. If I end up rarely using the capacity even when only charging 80% i might make some smaller batteries for shorter rides.
 
beemac said:
casainho said:
beemac said:
Wireless controller is on the handlebar - needs a better case but is nice to have the led where I can see it.

Build is currently 17.1kg without battery - not as light as I hoped it would be but reckon I can get it down a bit... was hoping for 18kg with battery but I think it's going to be more like 20 - maybe 19 if I can get some lighter tyres and change the handlebar+crank arms...
Why don't you 3D print the wireless remote enclosure and use it? you would not use the coin battery and install the DC-DC converter near the TSDZ2 display connector.

That beautiful ebike deserves a small battery as his frame, to keep being beautiful.
For battery savings, you can make your own and save weight and volume by choosing the materials: a smaller BMS, enclosure materials and other parts.
Also very important is to have like one small / lighter and other big battery, so you can use the lighter one daily and the bigger one for some special day like for long ride. There are way to make a quick install / remove fix for the battery.

I did think about that - but my smallest board is quite a bit larger than just the nrf - so I don't think it will fit your design as-is unfortunately - and after building four of these boards I'm not in a hurry to solder another :)

So I'll probably just print a simple case with a lens for the led for now and when I have to build another board will try to fit in the remote case.

Thanks I am very pleased with it - took a lot of searching to find the right frame!

Yea going to be interesting what the battery is like when it arrives - went for a wallen power one in a new slightly slimmer case than the standard shark. 14s4p - hoping for just over 3kg. Making my own batteries is definitely something that's on my list - is why I got the a30t printer with the larger print size but for now I'm using off the shelf. If I end up rarely using the capacity even when only charging 80% i might make some smaller batteries for shorter rides.
I mean you can put the DC-DC board outside the handle bar, on the TSDZ2 display connector. Then you could use the wireless remote enclosure.
I have batteries of 14S2P (300Wh), so, half of your battery. But you can go with like 12S2P for instance and also not use that battery DC-DC if you are ok to sacrifice a bit of battery power using the Field Weakening feature.
 
casainho said:
beemac said:
casainho said:
beemac said:
Wireless controller is on the handlebar - needs a better case but is nice to have the led where I can see it.

Build is currently 17.1kg without battery - not as light as I hoped it would be but reckon I can get it down a bit... was hoping for 18kg with battery but I think it's going to be more like 20 - maybe 19 if I can get some lighter tyres and change the handlebar+crank arms...
Why don't you 3D print the wireless remote enclosure and use it? you would not use the coin battery and install the DC-DC converter near the TSDZ2 display connector.

That beautiful ebike deserves a small battery as his frame, to keep being beautiful.
For battery savings, you can make your own and save weight and volume by choosing the materials: a smaller BMS, enclosure materials and other parts.
Also very important is to have like one small / lighter and other big battery, so you can use the lighter one daily and the bigger one for some special day like for long ride. There are way to make a quick install / remove fix for the battery.

I did think about that - but my smallest board is quite a bit larger than just the nrf - so I don't think it will fit your design as-is unfortunately - and after building four of these boards I'm not in a hurry to solder another :)

So I'll probably just print a simple case with a lens for the led for now and when I have to build another board will try to fit in the remote case.

Thanks I am very pleased with it - took a lot of searching to find the right frame!

Yea going to be interesting what the battery is like when it arrives - went for a wallen power one in a new slightly slimmer case than the standard shark. 14s4p - hoping for just over 3kg. Making my own batteries is definitely something that's on my list - is why I got the a30t printer with the larger print size but for now I'm using off the shelf. If I end up rarely using the capacity even when only charging 80% i might make some smaller batteries for shorter rides.
I mean you can put the DC-DC board outside the handle bar, on the TSDZ2 display connector. Then you could use the wireless remote enclosure.
I have batteries of 14S2P (300Wh), so, half of your battery. But you can go with like 12S2P for instance and also not use that battery DC-DC if you are ok to sacrifice a bit of battery power using the Field Weakening feature.

Ah ok - is there room for the mosfet board though? Or you mean put all the power stuff in a separate enclosure? I think if I did go that route I should probably mod your design slightly to accommodate the extra bits in one enclosure...

With the DC booster I can run any battery pack in theory and get the full 60v if I want it even with 14s2p (although 2p might not give enough current unless i choose very high current cells). I recommend it for a new thing to play with - i also bought 1500/1800w versions but they appear to be very similar only differing in fuse/cooling design. The 1200W heat sink is passively cooled and pretty compact and even in a rucksack for a short run only got slightly warm... outside with air flowing over it I reckon should be fine for most conditions.

https://www.banggood.com/1200W-20A-DC-Converter-Boost-Step-Up-Power-Supply-Module-Input-10-60V-Output-12-83V-p-1006423.html
 
beemac said:
With the DC booster I can run any battery pack in theory and get the full 60v if I want it even with 14s2p (although 2p might not give enough current unless i choose very high current cells). I recommend it for a new thing to play with - i also bought 1500/1800w versions but they appear to be very similar only differing in fuse/cooling design. The 1200W heat sink is passively cooled and pretty compact and even in a rucksack for a short run only got slightly warm... outside with air flowing over it I reckon should be fine for most conditions.

https://www.banggood.com/1200W-20A-DC-Converter-Boost-Step-Up-Power-Supply-Module-Input-10-60V-Output-12-83V-p-1006423.html
Very interesting. I bought a booster a while ago but was wary of the quality of the power coming from it. I have a scope now so I can put it under load and see what it looks like.

Certainly you want to keep the set output voltage closer to the battery voltage. Only the boost voltage different over the battery voltage is subject to notable losses. Also make sure you set you battery cutoff voltage on the booster well!

You have inspired me I will dig mine out!

Cheers
 
beemac said:
....
With the DC booster I can run any battery pack i..... The 1200W heat sink is passively cooled and pretty compact and even in a rucksack for a short run only got slightly warm... outside with air flowing over it I reckon should be fine for most conditions.....
:thumb: I was thinking about this too, to feed my 36V tsdz2 (and 36V battery) with 48V, but had doubts about the DC output and heat. It is a relative cheap solution to improve tsdz2, compared to another battery.
I had planned to integrate a small Voltage display on the input, to monitor the battery Voltage, because the controller can't measure this anymore. As I understand, you had no problems with the tsdz2.
 
that FSR is beautiful and the weight really is pretty decent with the dropper and heavy flat pedal. I am doing a weight weenie build right now and trying to keep it on a budget but its like an endless spiral thats exponentially more past a certain point.

The TS i have on the m2 stumpjumper didnt clear the carbon hardtail frame i picked so im going rear hub on this one with a g310, 7/15a controller and a 13s2p 21700 pack.

https://bicyclemotorworks.com/product/ebike-battery-pocket-rocket-ii-52v-8ah/

something like this stuffed in a seat bag would probably not be a bad setup for you for shorter rides if you are trying to keep the weight down
 
mctubster said:
beemac said:
With the DC booster I can run any battery pack in theory and get the full 60v if I want it even with 14s2p (although 2p might not give enough current unless i choose very high current cells). I recommend it for a new thing to play with - i also bought 1500/1800w versions but they appear to be very similar only differing in fuse/cooling design. The 1200W heat sink is passively cooled and pretty compact and even in a rucksack for a short run only got slightly warm... outside with air flowing over it I reckon should be fine for most conditions.

https://www.banggood.com/1200W-20A-DC-Converter-Boost-Step-Up-Power-Supply-Module-Input-10-60V-Output-12-83V-p-1006423.html
Very interesting. I bought a booster a while ago but was wary of the quality of the power coming from it. I have a scope now so I can put it under load and see what it looks like.

Certainly you want to keep the set output voltage closer to the battery voltage. Only the boost voltage different over the battery voltage is subject to notable losses. Also make sure you set you battery cutoff voltage on the booster well!

You have inspired me I will dig mine out!

Cheers

Yea I can't vouch for the quality of the power as not looked at it under a scope - hopefully isn't too bad! My new battery will be 52v so I keep the boost to minimum to keep losses low.

I need to do a bit more setup - haven't changed the cutoff and would have probably forgotten had you not mentioned - thanks! I also twiddled the constant current pot before I realised what it was - i think it's set to max now but I need to find a test load to check it's not going to limit current...

Elinx said:
beemac said:
....
With the DC booster I can run any battery pack i..... The 1200W heat sink is passively cooled and pretty compact and even in a rucksack for a short run only got slightly warm... outside with air flowing over it I reckon should be fine for most conditions.....
:thumb: I was thinking about this too, to feed my 36V tsdz2 (and 36V battery) with 48V, but had doubts about the DC output and heat. It is a relative cheap solution to improve tsdz2, compared to another battery.
I had planned to integrate a small Voltage display on the input, to monitor the battery Voltage, because the controller can't measure this anymore. As I understand, you had no problems with the tsdz2.

So far no issues but I've only gone a few miles as a test - the voltage/soc reading is a bit of a pain - as you say the controller has no idea of battery state any more. If I start to use the DC booster regularly I'm planning to add support for a 2nd SOC sensor on the wireless controller so you can tell the software to use either the built in SOC sensor - or the alternative that can be placed between the battery and DC booster - either as a patch or hopefully included in the main code if casainho agrees - but tbh there are many other changes that are more important than that so might be a while.
 
Back
Top