TSDZ2 not working after automatic shutdown

[Eggie]

So first thing to try is to change or adjust the speed sensor and secondly the wiring of the motor/torque sensor?
If I’m not mistaking this should be it? I can check for humidity when checking the cabling/insulation. Mechanical I don’t think there are any issues. It must be something sensor or wiring or maybe software (but that seems far fetched).
Thanks for all the help and input. This forum is a godsend for diy’ers like myself!

 

[beemac]

So first thing to try is to change or adjust the speed sensor and secondly the wiring of the motor/torque sensor?
If I’m not mistaking this should be it? I can check for humidity when checking the cabling/insulation. Mechanical I don’t think there are any issues. It must be something sensor or wiring or maybe software (but that seems far fetched).
Thanks for all the help and input. This forum is a godsend for diy’ers like myself!

I can't work out from the thread if you have the issue with the stock fw or only osf? If just osf - do you have the temp sensor installed? I've thought my battery had run low before when actually it was just the osf throttling back the power due to the motor being too warm.

If you've not done any shims/thermal paste mods - the case might be cool but the motor could still be hot because it's effectively insulated in the case (one of Chalo's favourite TSZD2 design choices!)

 

[Eggie]

Actually I had problems with both but the symptoms I explained are with stock FW.

 

[ornias]

That's not a job for TSDZ2, then.

I suggest having two e-bikes. One with a hub motor for getting around, and another one with TSDZ2 for troubleshooting and repairing.

That’s a strange idea. As I said, it’s a cargo bike. It had a hub motor but that wasn’t strong enough anymore. Why would the tsdz2 not be sufficient to get around every day? Why get it then?

@Chalo is very knowledable and insightful - but he has some sort of hate relation with the TSDZ2.
Best to ignore his comments on TSDZ2 - or at least take them with a big grain of salt.....

Just keep reading here about all the many different problems people have with TSDZ2 that are either rare or non-existent with other systems. Eventually you'll get it.

There are 3 core issues with the TSDZ2, ignoring things solved in the OSF and/or are issues with all chinese motors like cheap bearings:
- Bad axle tolerances
- Bad Circlips groves (aka breaking axles)
- Torque Sensor design/maintenance

All reports of those, are: valid, verified and design related indeed.
But one should just ignore torque sensor related issues because there simply aren't any other mid-drive kits with one.

Though i'm also ignoring bad cooling related issues here, because that's actually what they used to bypass EU regulations (which are measured by power @ highest safe temperature equilibrium). If you know the testing regime there are 2 tricks one can do to pass-off higher-end motors (in terms of performance) as 250w, this is one of them.

Anyway:
The issues shown here, do not seem to coincide with any of these 3 "known big design flaws". Sounds more like it's either electrical/sensor related or a partly burned-out motor.
So I think the "bad motor!!11!!1!" comment is a bit soon

 

[Eggie]

Little update:

Since my last post I added the throttle and tried to change the speed sensor position. So far so good, the engine drives like it should. Als the “low-battery-feeling” is gone. I started monitoring the discharge after a days use (we ride it about 15km a day). Bear in mind we use the tsdz2 on a cargo bike driving 2 kids to school, so the weight is rather big. We use it almost always on pas 4 so it’s always on maximum. We did 3 days (and so 45km) without charging and based on my measurements we’re at about 37% of a full charge. We decided to charge it today so we’re not in trouble tomorrow.
We actually didn’t use the throttle (yet) but I wonder if it could have an impact just to have it plugged in?

On the build quality there’s a lot to say. But (and this is a big but) the serviceability is also extremely high. I don’t think you should open up your Bosch motor and exchange some parts. All in all, the design is simple but efficient (although there are some design flaws). A just added a new TSDZ2b to a new build and will be test driving it next weekend. I didn’t open it up to see what the interior looks like and what the build quality is. I just drove it for 200m just to check if everything is working as it should.

 

[beemac]

Little update:

Since my last post I added the throttle and tried to change the speed sensor position. So far so good, the engine drives like it should. Als the “low-battery-feeling” is gone. I started monitoring the discharge after a days use (we ride it about 15km a day). Bear in mind we use the tsdz2 on a cargo bike driving 2 kids to school, so the weight is rather big. We use it almost always on pas 4 so it’s always on maximum. We did 3 days (and so 45km) without charging and based on my measurements we’re at about 37% of a full charge. We decided to charge it today so we’re not in trouble tomorrow.
We actually didn’t use the throttle (yet) but I wonder if it could have an impact just to have it plugged in?

On the build quality there’s a lot to say. But (and this is a big but) the serviceability is also extremely high. I don’t think you should open up your Bosch motor and exchange some parts. All in all, the design is simple but efficient (although there are some design flaws). A just added a new TSDZ2b to a new build and will be test driving it next weekend. I didn’t open it up to see what the interior looks like and what the build quality is. I just drove it for 200m just to check if everything is working as it should.

Not sure how having the throttle installed would affect your issue - but if you're working the tsdz2 quite hard - and sounds like you might be - i'd strongly recommend installing the temp sensor mod (and osf to support it) and ideally improve the heat dissipation otherwise you run a real risk of burning the motor out and not knowing about it (been there!). You'd lose the throttle function though then since the temp sensor uses the same adc port.

https://github.com/OpenSourceEBike/TSDZ2_wiki/wiki/How-to-install-motor-temperature-sensor
https://github.com/OpenSourceEBike/TSDZ2_wiki/wiki/How-to-improve-motor-heat-dissipation

 

[Eggie]

you run a real risk of burning the motor out and not knowing about it (been there!).

Doesn’t the tsdz2 have its own overheating protection. We never use the throttle actually. I installed it just in case. I don’t know if I’m exhausting it too much. It doesn’t feel like it.

 

[beemac]

you run a real risk of burning the motor out and not knowing about it (been there!).

Doesn’t the tsdz2 have its own overheating protection. We never use the throttle actually. I installed it just in case. I don’t know if I’m exhausting it too much. It doesn’t feel like it.

No not by default - hence the mods that add it. You mentioned you run in max assistance - and on a heavy bike that sounds like it's working it pretty hard. Because the motor is effectively insulated in the case, you'll not know about it until it's too late. Case can feel cool but the motor inside can be hot - and if the motor gets to 100c+ the magnets will demagnetise and you'll need a new motor...

Even with the all the shims and paste in the case to help get heat out of there - i can still get my motor too hot just riding around so it's nice to know the sw can detect and throttle back to protect itself.

 

[ornias]

Little update:

Since my last post I added the throttle and tried to change the speed sensor position. So far so good, the engine drives like it should. Als the “low-battery-feeling” is gone. I started monitoring the discharge after a days use (we ride it about 15km a day). Bear in mind we use the tsdz2 on a cargo bike driving 2 kids to school, so the weight is rather big. We use it almost always on pas 4 so it’s always on maximum. We did 3 days (and so 45km) without charging and based on my measurements we’re at about 37% of a full charge. We decided to charge it today so we’re not in trouble tomorrow.
We actually didn’t use the throttle (yet) but I wonder if it could have an impact just to have it plugged in?

On the build quality there’s a lot to say. But (and this is a big but) the serviceability is also extremely high. I don’t think you should open up your Bosch motor and exchange some parts. All in all, the design is simple but efficient (although there are some design flaws). A just added a new TSDZ2b to a new build and will be test driving it next weekend. I didn’t open it up to see what the interior looks like and what the build quality is. I just drove it for 200m just to check if everything is working as it should.

Instead of PAS 4, you might want to look into emtb mode on emmebrusa firmware with such a load.
Basically it increases assistance percentage with each increase in torque.

I think that might be nice, as basically "taking off with 2 kids" is kinda a comparable load with "taking a small hill with an mtb".
It might save considerable battery, by having a little less assistance when cruising *and* give some thermal headroom as well.

 

[Eggie]

No not by default - hence the mods that add it. You mentioned you run in max assistance - and on a heavy bike that sounds like it's working it pretty hard. Because the motor is effectively insulated in the case, you'll not know about it until it's too late. Case can feel cool but the motor inside can be hot - and if the motor gets to 100c+ the magnets will demagnetise and you'll need a new motor...

Even with the all the shims and paste in the case to help get heat out of there - i can still get my motor too hot just riding around so it's nice to know the sw can detect and throttle back to protect itself.

Hmmm ok. That sounds like a big design flaw. I’ve seen some 3D printed cases with air vents, I wondered why that was and felt like overkill. I really don’t feel like putting on the osf for the moment. I first want it to ride like it should in stock fw. We ride it for 8km a time. Does it heat up that easily? Maybe I can add a temp sensor with read-out and just go with that?

I don’t understand why it would burn with stock fw. It’s the 350w version.

 

[beemac]

No not by default - hence the mods that add it. You mentioned you run in max assistance - and on a heavy bike that sounds like it's working it pretty hard. Because the motor is effectively insulated in the case, you'll not know about it until it's too late. Case can feel cool but the motor inside can be hot - and if the motor gets to 100c+ the magnets will demagnetise and you'll need a new motor...

Even with the all the shims and paste in the case to help get heat out of there - i can still get my motor too hot just riding around so it's nice to know the sw can detect and throttle back to protect itself.

Hmmm ok. That sounds like a big design flaw. I’ve seen some 3D printed cases with air vents, I wondered why that was and felt like overkill. I really don’t feel like putting on the osf for the moment. I first want it to ride like it should in stock fw. We ride it for 8km a time. Does it heat up that easily? Maybe I can add a temp sensor with read-out and just go with that?

I don’t understand why it would burn with stock fw. It’s the 350w version.

Yes it totally is a design flaw - exactly the same product is sold as 250W-750W - and it certainly cant dissipate 750W for any length of time... I haven't run the stock fw for long at all - so can't comment on whether it still overheats but I can't see why it wouldn't since the problem is the case rather than the sw. Don't get me wrong I love the tsdz2 but I wouldn't recommend it to someone who wanted a fit-and-forget solution - only to those who like a bit of a fiddle

you can certainly add a separate temp sensor - lots of discussions on that and other mods here

https://endless-sphere.com/forums/viewtopic.php?f=28&t=102046

 

[Eggie]

The only thing I don’t understand is why it would overheat in my situation. Like you said, it’s the same motor for every version. Since we use the 36v version at 350W it seems like the least straining option? I bought the 250w version but put on the 350w fw. I will be monitoring the temp just in case.

 

[beemac]

The only thing I don’t understand is why it would overheat in my situation. Like you said, it’s the same motor for every version. Since we use the 36v version at 350W it seems like the least straining option? I bought the 250w version but put on the 350w fw. I will be monitoring the temp just in case.

Not saying it will, just saying it might - because it's a common issue. And if it does you're down a bike for a while which is a drag. I've not got any experience with the 36v version but it could be more problematic at the same wattage than a 48v - since it will use greater current to get that 350W - and it's current that causes heat...

 

[Eggie]

Not saying it will, just saying it might - because it's a common issue. And if it does you're down a bike for a while which is a drag. I've not got any experience with the 36v version but it could be more problematic at the same wattage than a 48v - since it will use greater current to get that 350W - and it's current that causes heat...

I understand. So the 48v 250W would be the least straining. I’m reading into the subject. Thanks for the heads-up!

Small question. Would the bafang bbs02 be more plug-in-and-don’t-worry?

 

[beemac]

Not saying it will, just saying it might - because it's a common issue. And if it does you're down a bike for a while which is a drag. I've not got any experience with the 36v version but it could be more problematic at the same wattage than a 48v - since it will use greater current to get that 350W - and it's current that causes heat...

I understand. So the 48v 250W would be the least straining. I’m reading into the subject. Thanks for the heads-up!

Small question. Would the bafang bbs02 be more plug-in-and-don’t-worry?

Yes although some run the 36v with a 48v battery - so it spins faster - better for high cadence riders although the field weakening in the osf helps with that too. So you don't need to bin the 36v if you wanted to try 48v.

I don't have any experience with bafang motors - but they are seen as being much more reliable than the tsdz2 in general - main feature difference is no torque sensing on bafangs - only cadence. Some people think that's more of an issue than others - personally i've never ridden a cadence sensing bike so can't comment...

 

[Eggie]

New update: the issue of stuttering came back. The problem is that I never encountered the issue (my partner drives the bike daily). So I’m a bit stuck right now and don’t really know what to do next. I presume opening the motor and checking the wiring and maybe the torque sensor. She says that when she really pushes hard it starts working a bit.
The throttle does work and gives full power. She also said that from time to time the speed read-out is wrong. I think I tried everything to position it correct but can’t seem to find the sweet spot. I read contradicting distances going from 4mm to 14mm. Fastest thing I can try is to try and fixate the sensor some more and if somebody could give a “correct” distance between the magnet that would be helpful. Can these break or be of inferior quality?
Being able to turn the sensor, does this means it's not tight enough? It seems that if I can easily turn it by hand it can move whilst riding the bike.

 

[beemac]

New update: the issue of stuttering came back. The problem is that I never encountered the issue (my partner drives the bike daily). So I’m a bit stuck right now and don’t really know what to do next. I presume opening the motor and checking the wiring and maybe the torque sensor. She says that when she really pushes hard it starts working a bit.
The throttle does work and gives full power. She also said that from time to time the speed read-out is wrong. I think I tried everything to position it correct but can’t seem to find the sweet spot. I read contradicting distances going from 4mm to 14mm. Fastest thing I can try is to try and fixate the sensor some more and if somebody could give a “correct” distance between the magnet that would be helpful. Can these break or be of inferior quality?
Being able to turn the sensor, does this means it's not tight enough? It seems that if I can easily turn it by hand it can move whilst riding the bike.

I can't test it because all my cables are taped up but I'm pretty sure the tsdz2 will work with the speed sensor removed. Try disconnecting it completely to rule it out...

 

[Eggie]

I can't test it because all my cables are taped up but I'm pretty sure the tsdz2 will work with the speed sensor removed. Try disconnecting it completely to rule it out...

Indeed, I just test drove my new build and didn’t get a speed reading but it continued to give support. So I don’t really need to unplug it, just turn it away. Will try! Thanks

 

[ornias]

No not by default - hence the mods that add it. You mentioned you run in max assistance - and on a heavy bike that sounds like it's working it pretty hard. Because the motor is effectively insulated in the case, you'll not know about it until it's too late. Case can feel cool but the motor inside can be hot - and if the motor gets to 100c+ the magnets will demagnetise and you'll need a new motor...

Even with the all the shims and paste in the case to help get heat out of there - i can still get my motor too hot just riding around so it's nice to know the sw can detect and throttle back to protect itself.

Hmmm ok. That sounds like a big design flaw. I’ve seen some 3D printed cases with air vents, I wondered why that was and felt like overkill. I really don’t feel like putting on the osf for the moment. I first want it to ride like it should in stock fw. We ride it for 8km a time. Does it heat up that easily? Maybe I can add a temp sensor with read-out and just go with that?

I don’t understand why it would burn with stock fw. It’s the 350w version.

Yes it totally is a design flaw - exactly the same product is sold as 250W-750W - and it certainly cant dissipate 750W for any length of time... I haven't run the stock fw for long at all - so can't comment on whether it still overheats but I can't see why it wouldn't since the problem is the case rather than the sw. Don't get me wrong I love the tsdz2 but I wouldn't recommend it to someone who wanted a fit-and-forget solution - only to those who like a bit of a fiddle

you can certainly add a separate temp sensor - lots of discussions on that and other mods here

https://endless-sphere.com/forums/viewtopic.php?f=28&t=102046

It's even more of a design-flaw when you realise the cooling was gimped on-purpose to comply to 250w limits AND they still kept the higher peak power knowing it would overheat that way.

 

[Eggie]

New update:

I’ve pinpointed the problem of the stuttering and lack of support. Since I never experienced the problem myself it took me some time to figure it out but it wasn’t caused by the automatic shutdown but due to starting the bike, starting to talk to somebody and restarting after auto-shutdown, and during all that time keeping one foot on a pedal… so the torque sensor didn’t calibrate correctly at the “reboot”.
The other problem was indeed the speed sensor. It seems really sensitive to positioning since it’s a matter of millimeters and due to some moving whilst biking it seemed to give some errors. I have now positioned it and try to hold in in place by adding some hot glue on the sides (besides the tie wraps). Even still just the turning of the sensor (what kind of happens while biking) seems to give wrong readings.

So I’m happy I found the problems without having to buy anything or having to work too much on the motor. I’m now thinking of there are any better quality speed sensors/magnets out there that aren’t that prone to error?

About the heating: the tsdz2 never really feels hot on the cargo bike (I have yet to install some kind of temperature measuring). But I have a tsdz2b installed on a normal city bike and after a ride it feels kinda warm. That feels strange to me as I’m my opinion the tsdz2 should give more power since the cargo bike is much heavier (and often contains 2 kids and bags). So maybe there’s a better heat management in the tsdz2b.

 

[Elinx]

....during all that time keeping one foot on a pedal… so the torque sensor didn’t calibrate correctly at the “reboot”.
.... was indeed the speed sensor. It seems really sensitive to positioning since it’s a matter of millimeters ......

About the heating: the tsdz2 never really feels hot on the cargo bike (I have yet to install some kind of temperature measuring). But I have a tsdz2b installed on a normal city bike and after a ride it feels kinda warm. That feels strange to me as I’m my opinion the tsdz2 should give more power since the cargo bike is much heavier (and often contains 2 kids and bags). So maybe there’s a better heat management in the tsdz2b.

:thumb: Thanks for the feedback. I'm glad you have solved your problems.
These two are most the cause of this sort of problems, that was why I mentioned these firstly.

About heating
The older Tsdz2 bare motor is isolated from the housing and that is why you don't feel the heat of that motor.
With heat conductive material between bare motor and housing you improve the motorcooling because the aluminium housing is used as a massive heatsink which will be cooled by airstream. But that housing will be hotter.

I don't know if the Tszd2B has some improvements with the bare motor assembly, which gives a better connection with the motorhousing.
If that is the case, the motorhousing will be hotter too which indeed means that the heat management of the tsdz2B is improved.

I know with TsdzB the blue gear is changed to a white gear without clutch and the clutch in the maingear is changed (doubled).
So maybe there is are some small changes too with the motorassembly inside the housing.
At first the space between Tsdz2 motorflanges and housing was about 1.5mm, the latest version was this decreased to almost 1mm.
So maybe with the Tsdz2B that space it is now reduced to zero.
But to be sure we must wait for a confirmation if someone have to open the motor or want to add heat conductive material.