New "TSDZ2 Torque Sensor Central Motor"

[famichiki]

Set your display up to show motor watts, it’s surprising how much the motor will input in without you really being aware.

Thanks for your help, that helped me diagnose the issue.

[ImpulsePurchase]

See here: https://github.com/OpenSource-EBike-fir ... que_sensor

Thanks for the info. Things are clearer now.
What I still do not know is how did the others press in the torque sensor while the inner ring is pressed by the small springs towards it. A lot of people worked on the sensor and I assume there are some ideas.

Some way back in the thread someone suggested the old trick of heating the case lightly and cooling the shaft, went looking for it but can’t seem to find it, or did I imagine this ? Anybody tried this ?

I think the heat and cool was to extract the bearing for the one way clutch. Heating and cooling is normally used for freeing stuck items rather than inserting them together

You can use a home made bearing press/headset press (long threaded bolt with large washers) if you want to be very precise but I suspect most of us just use a mallet with the motor removed from the bike. The roller bearings in the middle of the shaft should be greased before you reassemble and this will make it significantly easier to fit (though not easy enough to do by hand). I have done it 3 times now and as long as you are diligent (take your time to make sure the springs are upright and the key nodules are aligned) its not a big problem. Just be prepared to extract if it looks like things aren't aligned.

[Waynemarlow]

Hi, we are having real problems with downed motors, well not really drowned but we are having probably the worst winter for mud and deep muddles that I can remember and the motors are having to put up with some real adverse conditions. Never the less we are two motors down with the same symptoms.

Symptoms are the motor starts to cut out after being through a deep puddle, the display still works and using the maintenance display, the torque sensor is still doing its stuff. The throttle also no longer works. The battery is not the problem as changing the battery doesn't solve it. You can turn the display off and then back on and it will reactivate the unit for a short period of time but within a few minutes the whole unit fails again. We've tried disconnecting the throttle and that's not the cause.

These units have been reliable in the dry but in the adverse conditions we're having there is a problem. Have any of you had similar problems ?

Speed sensor, will the unit operate if the speed sensor fails ( water and mud covered ) or does it perhaps time out if the sensor doesn't pick up a pulse ?

Any ideas would be appreciated.

[devboy-greg]

Hi, we are having real problems with downed motors, well not really drowned but we are having probably the worst winter for mud and deep muddles that I can remember and the motors are having to put up with some real adverse conditions. Never the less we are two motors down with the same symptoms.

Symptoms are the motor starts to cut out after being through a deep puddle, the display still works and using the maintenance display, the torque sensor is still doing its stuff. The throttle also no longer works. The battery is not the problem as changing the battery doesn't solve it. You can turn the display off and then back on and it will reactivate the unit for a short period of time but within a few minutes the whole unit fails again. We've tried disconnecting the throttle and that's not the cause.

These units have been reliable in the dry but in the adverse conditions we're having there is a problem. Have any of you had similar problems ?

Speed sensor, will the unit operate if the speed sensor fails ( water and mud covered ) or does it perhaps time out if the sensor doesn't pick up a pulse ?

Any ideas would be appreciated.

I have the exact same issue whilst enjoying this wonderful UK winter.

Display working as expected, torque sensor values outputting as expected just no power from the motor, even with walk assist. It tends to happen towards the end of a ride when the battery is below 50% and everything is soaking wet.

I am thinking the problem may lie with the 6pin connector between the KT-LCD3 and the motor, it is very low down near the bottom bracket and gets completely soaked. Upon inspection I can see that water has penetrated the connector.

I am in the process of installing a new controller thinking this was the issue (I tested just about everything else!) but now I am unsure, so as a precaution I have removed the 6 pin connector by running a new 6 core cable from inside the motor up to the handlebars where I have split of the brake sensor wire and added a new Bafang 5 pin connector which I think is superior, but also in readiness for the new SW102 display I have ready and waiting once the firmware issues for SW102 display are fixed.

I plan to have everything up and running again over the coming weekend (some soldering left to do) and will report back.

If I hadn’t had the motor apart I would probably have tried improving the 6 pin connector weatherproofing with silicon and waterproof tape after drying thoroughly.

[Waynemarlow]

Not convinced about the connectors as I have the units out on the Bench with dry connectors and still the same, also battery voltage seems irrelevant as we have spare fresh batteries which we have swapped in without change.

The one thing common is both motors “ slowly “ died as in the time interval between fail and reboot got shorter each time we had to reboot the system, until there was a total failure.

Don’t get me wrong but these really are adverse conditions with mud being anything from 50 mm to 300mm deep with intervening puddles of the same depth, really really not Ebike motor friendly. But it’s really good fun trying to recreate your youth again

[famichiki]

If I hadn’t had the motor apart I would probably have tried improving the 6 pin connector weatherproofing with silicon and waterproof tape after drying thoroughly.

Dielectric grease (silicone-based grease) can be used for this, it's commonly used on automotive connectors like spark plug boots. Much better than using silicone sealant or other permanent solutions as the connectors can be easily unplugged again.

[LePeche]

Diagnosis help please?

Commuting prior to Christmas and the motor started to make a horrible loud noise, and lost power/ assist. Spoke to the seller (aliexpress), read the wiki, and it pointed towards the blue gear. After incident, the noise abated, but no assist...so I ordered new gear. Replaced blue gear today with metal, but still no assist. Checked torque setting - get 90, and when pressure applied the value increases, so assume its working ok. When motor is off, then its quiet ... when I increase power setting, then I get the gear noise, so assume the motor is trying to apply power...although I don't get the impression its high-speed as its not a whine.

Haven't updated to open source firmware as yet, so telemetry limited.

Other -

1. bike seems extremely challenging to push backwards. I really don't remember there being this much resistance.

2. Pedal/ crank is fine i.e. doesn't slip.

3. prior to issue I was having a lot of chain slips at the crank and jamming. Now replaced chainring and added a guide but can't get any assist to ride.

4. Motor when opened looks clean, and can't see any water damage. The motor casing did have some strange debris .. not mud, not grease ... scrapped off and seems ok. Nothing huge.

5. no errors on display and indicates battery is fully charged

6. I do have a slight wobble on the peddles/ spider/ crank .. many people seem to report this so assume its normal.

Other user indicated one way bearing, but reviewing the wiki I don't see the same symptoms. Peddles do not slip. But its my next option to try. Motor, gears, controller all look in good condition in general. Its done around 1300kms.

PAS Sensor or one way clutch, although the clutch doesn't fit the symptoms mentioned in the wiki?

Thanks

[Mike-P]

Got it into maintenance mode but instead of any torque sensor readout it showed beneath the large r e the total mileage covered and the motor did work briefly. Restarted and now in maintenance mode four zeros (still with miles thereafter) and which never changed. Wondering if I have both a faulty torque sensor and display?

[casainho]

New TSDZ2 with pedal axle not aligned

Yes, I just finished my last build with a new TSDZ2 on a Decathlon 540S full suspension model, for my son with 10 years old.
The battery pack is very small and light, a 52V 350 Wh , 14S2P, with 3500 mAh 18650 cells.





Everything went smooth, the installation of 850C display, our OpenSource firmware but finally when I started the torque sensor calibration, I saw that pedals cranks were not aligned!! After testing with a new ones and looking at the axle, I did understand that the axle is not aligned and so the cranks are not vertical aligned... This is so sad to buy a new TSDZ2 and it came like this...

[andrea_104kg]

It's really incredible ... This engine is cheap but in this case the low quality limits have been exceeded !!
Are you sure it's not the hole on the cranks that is wrong?

[Elinx]

If I read sucure the cranks were changed for new ones. If the axle is twisted, I asked myself how this could be possible with mass production. If that is true, than there must be {a lot} more faulty motors and a bad quality check.
Same for the cranks, but imho there is more chance they could be different because left/right are seperated parts. I hope it can be easy solved.

[casainho]

If I read sucure the cranks were changed for new ones. If the axle is twisted, I asked myself how this could be possible with mass production. If that is true, than there must be {a lot} more faulty motors and a bad quality check.
Same for the cranks, but imho there is more chance they could be different because left/right are seperated parts. I hope it can be easy solved.

Yes, I checked that cranks were ok. Then I could see on the square for that the axle is not aligned.

This motor had something new for me, the 8 pin connector to the displays has inverted female/male connector compared the other previous motors I bought.

[dameri]

Did you buy the engine from PSWPower? I think there is pretty good customer service.
At least TSDZ2 motor develops some how. Inverted connectors. Is it good thing or shit happens like axles.

[andrea_104kg]

Where did you buy the engine? probably many other engines will have the same defect, I don't think they could have made a single axis wrong. It would be right to warn users. Perhaps the same engine is produced in different factories, perhaps under license and therefore with slightly different standards and characteristics.

[famichiki]

Did you buy the engine from PSWPower? I think there is pretty good customer service.
At least TSDZ2 motor develops some how. Inverted connectors. Is it good thing or shit happens like axles.

The inverted connectors (motor side female) are good because there are 8-pin to Bafang 5-pin adapter cables available in many configurations.
https://www.eco-ebike.com/search?q=tsdz ... sion+cable

[Mike-P]

Stripping mine down to try to establish what the fault might me and I cannot get anything to budge having taken the circlips off
Any suggestions would be appreciated. The o ring meanwhile has fallen off - presume it needs a new one, what size should I order?

IMG_20200112_115800212.jpg (220.23 KiB) Viewed 1814 times

[devboy-greg]

In the repair manual (available here) they use a gear puller.

From my own personal experience I have always been able to gently ease it out by hand, Jbalat shows it in his video to replace the sprag clutch:

https://www.youtube.com/watch?v=u-SY0rGMEMo

Hope this helps.

[Waynemarlow]

British winter conditions 3 TSDZ2 0, all 3 of our engines have gone down now, we have so much mud and water on the trails that it is really really tough conditions for any electrical motor, never the less we have to find the cause.

Same symptoms as in motor shutting down progressively until the display also shuts down. Interesting to note that the speed display was the first to give weird numbers ( cycling without motor ) until it failed and then soon after the display.

We have both V20 on LCD3 and standard TSDZ2 on VClD5 displays.

Mark P mine starting making the horrible non bearing noise yours had, I suspect it maybe one of the phases not working correctly on the motor.

Would an earthing problem throughout the motor be a cause ? I’ll strip mine down later and check.

[_g_]

Not sure if better here or own thread, but...
My VLCD5 just shows all elements on; ie 88.8, 888.8 all graphs and other elements full.
It does this as soon as I plug it in, without pressing any buttons

Does anyone why and how it can be fixed? Ideally I'd like to use it tomorrow.

I got this motor with a XH18.
I've put the XH18 on the bike the VLCD5 came from and leant it to a friend (my plan was always to have that display on that bike, keeping it compact).
I was hoping to flash the firmware and use the 850C display I got for the purpose. However going to do it today, I realised I hadn't clicked that the USBUART and STLinkV2 were different items - so I've only got the STLinkV2 and will need to order the other.

The cable from the VLCD5 had previously been broken, so I had used the plug end of that to attach to the 850C, which is now soldered up.
I've just wired up the spare plug end from the other cable I had to the VLCD5, keeping the colours consistent.

I could flash the motor firmrware today, but not the 850C display - I'm presuming that won't help me get it to work with either display.

If I disconnect the brown cable, the display it's self works correctly.
However motor still doesn't work. Motor worked fine when I took the XH18 off - unfortunately now leant to a friend, so can't put it back on to try.
The brown cable is listed as "UART TX motor controller", but I'm getting full battery voltage through it - anyone know the likely cause?

When I test with a multimeter, this is what I get for each cable:

Black Ground
Brown 41.7v
green 41.7v
white 180mv
orange 16mv
Purple 4.6v

I've seperately been trying and failing to get the opensource firmware working, though it installed fine (posted in the 850c display thread) - I presume the issues are related.

[Mike-P]

In the repair manual (available here) they use a gear puller.

From my own personal experience I have always been able to gently ease it out by hand, Jbalat shows it in his video to replace the sprag clutch:

Thanks - easing does not seem to be working

[casainho]

In the repair manual (available here) they use a gear puller.

From my own personal experience I have always been able to gently ease it out by hand, Jbalat shows it in his video to replace the sprag clutch:

Thanks - easing does not seem to be working

Nite that there is the torque sensor under the main gear... You should not apply force to torque sensor but only to main gear, that always works for me and I only need two screendrivers to pull gently and apply the force on side only of main gear.

[Mike-P]

Found a youtube video [youtube]https://www.youtube.com/watch?v=-V3G0GSx4zg[/youtube] taking a completely different approach knocking the lot out from the rear which worked albeit probably potentially not too good for the health of the torque sensor. The only thing that seems looks faulty/seems odd is that only two of the three springs appear to exist - have brushed the area where I took it apart and scanned it with a bright light. Wondering if only two did exist and that's why as they weakened it suddenly stopped working / works very occasionally. The loss of power was immediately after the bike bounced, as it does on every commute, over a slightly depressed filled in trench across the road. Alternatively with the display issues as well could it be a fault with the controller? Pondering whether it would be best to get a new torque sensor and a new controller, or simply put it back together with three springs and see of it is okay. EDIT - think I'll get a new torque sensor and see if that cures it,

[Waynemarlow]

Ok, taken my defunct unit apart and its totally ferked with water ingress, the controller has burnt out and there's evidence of water throughout the housings.

So how do we prevent water getting in there. Am I missing something or has my unit got a missing seal. On the main water cover around the chain ring side, the cover has really good seals around the outer parts of the cover, but no seal around the bearing itself, just an interference fit. Am I missing something here or is this a design fault.

Once water is in that section then there is an open hole just beside the torque sensor wire + the opening of the torque sensor wire itself which are all unsealed, which can let water down into the controller and motor cavity.

As it stands if there is no seal around the bearing and just a dust shield, then we cannot guarantee any sort of water protection.

[casainho]

Once water is in that section then there is an open hole just beside the torque sensor wire + the opening of the torque sensor wire itself which are all unsealed, which can let water down into the controller and motor cavity.

I am being happy with my motors on rainy days, riding like 4h under rain and a lot of mud -- but yes, I understand you have much more of this than me

I guess you can block the torque sensor wire hole with silicone. When I remove my motor controller, I always put a lot of silicone on the hole of the cables to block the water and it seems to work well.

[Waynemarlow]

Thanks for the reply, up to this point we have not had a problem but this year the puddles are exceptional.

The torque sensor wire has to be able to float and make a contact as per the tension on the springs beneath the torque sensor, if we silicone them off we run the risk of the wires being pulled from the torque sensor.

We really need to prevent the water getting that far, do you know if there is any seal around the main gear and the cover ?