I've bought and installed two TSDZ2 motors. When I had a problem after the second installation, (an early version of the Model B), I replaced the blue gear with the ModelB white gear which has no needle gear, possible in my case because I had the ModelB main clutch. I didn't find any incompatibilty. It might be worth seeing if this can be done to all Model A's.
New "TSDZ2 Torque Sensor Central Motor"
- Thread starter larsottar
- Start date
I think it is possible, but besides a no-clutch white gear, you have to replace the maingear and plastic coverplate too, or modify that coverplate to fit the new doubleclutch maingear.
There is on the maingear a plastic coverplate. The inner diameter of the old one is 81mm, the inner diameter of the new one 88mm.
imho this is because the clutch is changed to a double clutch that need more space.
Here you can see a comparision of the other parts.
Last edited:
Thanks, sorry, I just deleted my above post before I saw your reply, as I realised I hadn't actually had both gears in my hands at the same time.
No problem,
I can imagine that not everyone knows the differences.
But for Tsdz2 vs Tsdz2B these are only mechanical.
NB
I was forgotten the spider which is different too
Got myself an 8-pin TSDZ2B kit from PSWPOWER with 15 Ah battery (back rack).
I've done the installation and got a few questions (see pictures):
I've done the installation and got a few questions (see pictures):
- The kit came with a small portion of battery cable with blue end-bits, but the battery was preassembled with a cable that had green end-bits. I connected the blue ones coming from the engine to the green ones, and it works. So, what the deal with green /blue, does it matter if it's not matching? and what's the point of the small extra portion of cable?
- I feel concerned because the battery connectors are exposed, better wrap that with some kind of tape but that will make it tricky to unplug should I need to do so. Any suggestions?
- I thought the VLCD5 would mount on a variety of handlebars, but it doesn't on mine. With or without the inner mounting rings (showing without in the picture). Was it supposed to come with other inner mounting ring sizes? I will have a 22 mm handlebar extension tomorrow so it should at least mount to that.
- I have the throttle, but it doesn't connect all the way in the VLCD5 (middle plug) and doesn't work, but if I push the connector a bit harder in, it doesn't stay in, but the engine turns on as if I was activating the throttle (even though I wasn't). What's up with that? Also I know that using the Throttle would not be legal in France, but still it's weird that it doesn't connect properly.
- I noticed that when I hold the minus button down the walk assist mode goes to 10km/h according to the speed sensor, is that normal? I thought it's supposed to be 6 km/h.
Attachments
Last edited:
get watertight connector or solder and shrinkwrap power wires so to be sure they NEVER short
I bought a soldering station and changed those batteryconnectors to amass xt60-w and shrink tube everywhere. Shortened the wires at the same time. Really nice but quite a bit of work
I dont have vlcd5 or throttle so cant help with that. My battery didnt have green plastic but that should hardly matter
I dont have vlcd5 or throttle so cant help with that. My battery didnt have green plastic but that should hardly matter
Got my tsdz2 due to bad knee. I adjust the assist level as my knee requests. The motor has kept me cycling. It's great. I do find that I don't get as much aerobic hard breathing going uphill compared to unassisted.That is the common China-prices. The prices in Europe are higher. Some examples...
elektrofahrrad-einfach.de
conrad.de
ebike-sales.co.uk
For people with health issues and/or special requirements there is only one good advice... Seek local dealers and try different bikes.
Did you see the video posted further up ?
Retrorockit
1 kW
I have a suggestion for a fix to prevent the broken crank shaft at the circlip groove.
The problem is the reduction in diameter at the snap ring groove. Looking at the formula for torsion bar stiffness there is an exponent of 4 related to shaft diameter. I broke the 8mm motor shaft in my BBSHD. The replacement was 9mm, which is 60% stronger. So the snap ring groove is not insignificant. Add in the square corners where a crack can start and failure there is likely. The BBSHD 8mm had a square corner, and a keyway slot all at the the end of the drive pinion. It broke where it was designed to break. The new 9mm shaft fixed all of those issues.
The solution that has occurred to me (I'll give the reason later) is to reduce the section between the bearings to a diameter slightly smaller than the snap ring groove. A radius at each end, and proper surface treatment (polishing, and shot peening) will produce a torque absorbing section of shaft that can absorb and release peak torque loads before they can damage the snap ring groove.
I've seen this used in truck repair for controlling torque on lug nuts. Back in the day when everything was steel lug nits were tightened with a 1" air impact gun which made 800 ft/lb. torque. You would just run the gun until the nut stopped turning. Nothing ever came loose, and there were 10 lug nuts if one ever stripped which was pretty rare, and not too hard to fix (with the 800# air gun).
But then alloy rims, and hubs came along to save weight and fuel, this warped hubs and damaged wheels. The solution was a socket extension with a reduced diameter that would flex at a predetermined torque. Typically 425 ft/lb although other rates were provided.
I'm not actually running my TSDZ2 so someone else will need to try this. Or maybe an actual engineer will look at this idea.
The problem is the reduction in diameter at the snap ring groove. Looking at the formula for torsion bar stiffness there is an exponent of 4 related to shaft diameter. I broke the 8mm motor shaft in my BBSHD. The replacement was 9mm, which is 60% stronger. So the snap ring groove is not insignificant. Add in the square corners where a crack can start and failure there is likely. The BBSHD 8mm had a square corner, and a keyway slot all at the the end of the drive pinion. It broke where it was designed to break. The new 9mm shaft fixed all of those issues.
The solution that has occurred to me (I'll give the reason later) is to reduce the section between the bearings to a diameter slightly smaller than the snap ring groove. A radius at each end, and proper surface treatment (polishing, and shot peening) will produce a torque absorbing section of shaft that can absorb and release peak torque loads before they can damage the snap ring groove.
I've seen this used in truck repair for controlling torque on lug nuts. Back in the day when everything was steel lug nits were tightened with a 1" air impact gun which made 800 ft/lb. torque. You would just run the gun until the nut stopped turning. Nothing ever came loose, and there were 10 lug nuts if one ever stripped which was pretty rare, and not too hard to fix (with the 800# air gun).
But then alloy rims, and hubs came along to save weight and fuel, this warped hubs and damaged wheels. The solution was a socket extension with a reduced diameter that would flex at a predetermined torque. Typically 425 ft/lb although other rates were provided.
I'm not actually running my TSDZ2 so someone else will need to try this. Or maybe an actual engineer will look at this idea.
Last edited:
I have the TSDZ2B motor with 48 volt battery. I have OSF and heat mods installed and can pedal up to about 25 mph without overheating. At 74 years old, and having heart disease, I cannot keep up at the necessary cadence for long. I would like to try a larger chainring, Is there a 50 tooth that will mount directly and has the right offset? Or do I need to buy an adapter? I read that some adapters do not fit the B version. Can someone point me to a sprocket that will fit for sure or a combination of adapter and sprocket that is known to work. Thanks
Narcoleptic
1 mW
- Joined
- Jan 10, 2019
- Messages
- 18
Hi all!
Early 2019 I got help by someone, who is no longer available to assist, to mount a 36V/350W kit from PSW, without foot brake + VLCD-5 350W (handlebar) controller.
This person advised - and helped - me to replace the controller (in the motor) with a 48V(6-pin) version and combine it with a 48V battery to be able to reach higher cadences. I am a reasonably heavy guy and have creaky old knees i want to save for more fun things than commuting. ;p
It's mounted on a bike with Shimano Nexus 8 hub and Shimano Roller brakes.
I did not get brake handles for the motor. Shift for gears is included in right breaking handle.
We did not install any brake sensors.
Last six months or so, speed has more often, then permanently, been displayed as 0 km/h.
Cleaning/adjusting sensor/magnet no longer fixes it.
Distance travelled is shown as (incorrect) 9.999.
Also - at start up, or when putting weight on pedals at lower speeds after resting on them during a ride - more and more often the motor would not contribute power - actually pedaling feels HARDER than with the motor turned off - but make a constant sound, louder than normal.
Taking force off, then on, pedals at first tended to fix this, but worked less often until this has become a permanent state rendering the bike practically useless.
If i do get up to speed (downhill) it seems the motor WILL contribute some power at higher speeds, and noice lessens.
Restarting the system does no longer help either.
I DO know not to touch pedals after startup - this is not the problem.
Now, I guess this points to the Torque sensor no longer working as intended?
If so, is there some suggested fix?
Alternately, could it be possible to rewire the motor to some "dumb" manual (thumb etc) controller, bypassing malfunctioning sensor/s?
I mainly need it to deload my knees on steep/long hills and am on a tight budget.
Thanks in advance!
/Narco
Early 2019 I got help by someone, who is no longer available to assist, to mount a 36V/350W kit from PSW, without foot brake + VLCD-5 350W (handlebar) controller.
This person advised - and helped - me to replace the controller (in the motor) with a 48V(6-pin) version and combine it with a 48V battery to be able to reach higher cadences. I am a reasonably heavy guy and have creaky old knees i want to save for more fun things than commuting. ;p
It's mounted on a bike with Shimano Nexus 8 hub and Shimano Roller brakes.
I did not get brake handles for the motor. Shift for gears is included in right breaking handle.
We did not install any brake sensors.
Last six months or so, speed has more often, then permanently, been displayed as 0 km/h.
Cleaning/adjusting sensor/magnet no longer fixes it.
Distance travelled is shown as (incorrect) 9.999.
Also - at start up, or when putting weight on pedals at lower speeds after resting on them during a ride - more and more often the motor would not contribute power - actually pedaling feels HARDER than with the motor turned off - but make a constant sound, louder than normal.
Taking force off, then on, pedals at first tended to fix this, but worked less often until this has become a permanent state rendering the bike practically useless.
If i do get up to speed (downhill) it seems the motor WILL contribute some power at higher speeds, and noice lessens.
Restarting the system does no longer help either.
I DO know not to touch pedals after startup - this is not the problem.
Now, I guess this points to the Torque sensor no longer working as intended?
If so, is there some suggested fix?
Alternately, could it be possible to rewire the motor to some "dumb" manual (thumb etc) controller, bypassing malfunctioning sensor/s?
I mainly need it to deload my knees on steep/long hills and am on a tight budget.
Thanks in advance!
/Narco
Last edited:
@Narcoleptic
With vlcd5 you can check with hidden settings if the torquesensor still works. Check T0 and T1 values with and without force on the pedal.
What you describe looks more like an overheated and demagnetized motor.
With 48V on a 36V motor you need to hold the current down, which you can't do with stock firmware. You need OSF for that.
But first you have to redone the speedsensor, because it could be a reason too. You must keep the distance of the magnet for about 10mm. Too close or too far give wrong values.
With vlcd5 you can check with hidden settings if the torquesensor still works. Check T0 and T1 values with and without force on the pedal.
What you describe looks more like an overheated and demagnetized motor.
With 48V on a 36V motor you need to hold the current down, which you can't do with stock firmware. You need OSF for that.
But first you have to redone the speedsensor, because it could be a reason too. You must keep the distance of the magnet for about 10mm. Too close or too far give wrong values.
Imho the easiest way is to fill the airgap between casecover and stator with thermoconductive thermopads.
Narcoleptic
1 mW
- Joined
- Jan 10, 2019
- Messages
- 18
Thank you for that robust advice! Will explore!
Narcoleptic
1 mW
- Joined
- Jan 10, 2019
- Messages
- 18
Ok, tried to do as instructed but ran into trouble:
My EN manual has the instruction to hold Power+I for 3 seconds followed by 4 quick presses of I.
My CH/EN manual says hold Power+I for 10 seconds.
No amount of holding Power+I @/after start-up screen/from power off opened hidden settings however.
Instead the start screen would freeze and become unresponsive, apart from Power toggling back light. Had to switch battery off to power down.
I also remembered that the Display lately (months) have tended to freeze seconds after start-up, in this configuration:
So had to adjust assist-level directly after start and stick to that. It used to work during rides.
(I only now noticed it is set to mph rather than kmph, which did NOT use to be the case while speedometer worked.
Some sort of reset happened?
This means I'm doubtful of the possibility to test speed sensor at this time. Even when spinning wheel immediately after startup the speedometer stayed at 0.00.
I did clean magnet and sensor. The placement of the sensor mean that its not possible to increase the distance between magnet and sensor without sensor-cable coming in contact with the wheel:
However this have also worked fine for a few years, only needing a bit of occasional wriggling around if no speed on display.
Any advice on how to proceed?
Cheers!
/Nick
Your magnet isn't in line with the small arrow on the sensor. Also it's a good idea to put a drop of thread-sealant (Loctite) on the threads of the magnet and spoke to secure it in position. Good luck with the hidden menu: my experience has been it's stupidly difficult to get to. My policy to avoid overheating is a low granny gear, thus avoiding overheating, plus only use in the cool of the morning in Andalucian summer. [EDIT: I also fitted thermal pads and made and fitted an Ally shim.
Narcoleptic
1 mW
- Joined
- Jan 10, 2019
- Messages
- 18
Hi! If you are replying to me:
- Magnet has never been farther away.
- Still speedometer have mostly worked good enough for years.
- When speed on display have been "0.00" before this, it has NOT caused the problems described in my initial post.
Conclusion: Speedometer is unlikely to be the culprit here?
Cheers,
/Nick
Narcoleptic
1 mW
- Joined
- Jan 10, 2019
- Messages
- 18
Hi!
Pic was only to illustrate distance. Adjusting does not help. Also see my reply to pxl666 above.
I remember accessing the menu fine before, but was quite awhile since I tried, so something must have happened.
Use have not changed and temps have not been extreme around here, so overheating unlikely cause.
Cheers!
/N
when too close it gives incorrect readings . when cable is damaged - same problem . just to let u know
ebikestuff.eu
100 mW
For anyone interested we have released our TSDZ2/TSDZ2B cooling kits tests. The results are interesting. It's a good read for anyone that is still unsure about switching to OSF or modifying the motor with a better cooling solution.
View attachment 336752
View attachment 336752
is it possible to try print bluegear from this ?3D model TSDZ2 (original) to see without dismantling. 32MB file
Motor_assembly_TSDZ2_edrawing.html
drive.google.com
Similar threads
