Review Tong Sheng TSDZ2 Owner Survey

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
997
428
Havant
Well, I've only done 500 miles on mine so far. It is a bit noisy in some scenarios and I have a spare blue gear to try, but have not had time to do it and won't before I go on holiday.
Enjoy the fettling - it's good to get inside these motors occasionly to check for any water ingress or re-greasing. If for no other reason it is good to explore the motor's layout to get to understand the various parts and their location and to realise that it isn't complicated to swap out most parts.

If you do swap out the Blue Gear, please feed back your findings whether or not the noise goes. I get the feeling that some motor are nosier than others and some say that OSF leads to a quieter ride.
I agree that the crank arms appear to need tightening more than I've experienced in other bikes, I just put a Jimmy bar on the Allen key...although I'm sure my neighbour will have a torque wrench I can use.
The recommended tightening torque is 40nm - in all my years of cycling I've never used a torque wrench on cranks (or much else in fact) and I thought I'd tightened the cranks well enough but after several episodes of hand tightening I got a torque wrench and was really suprised how tight 40nm was - certainly well beyond what I expected.
 

Kevin williams

Pedelecer
Sep 16, 2014
62
12
53
When I changed my
Enjoy the fettling - it's good to get inside these motors occasionly to check for any water ingress or re-greasing. If for no other reason it is good to explore the motor's layout to get to understand the various parts and their location and to realise that it isn't complicated to swap out most parts.

If you do swap out the Blue Gear, please feed back your findings whether or not the noise goes. I get the feeling that some motor are nosier than others and some say that OSF leads to a quieter ride.


The recommended tightening torque is 40nm - in all my years of cycling I've never used a torque wrench on cranks (or much else in fact) and I thought I'd tightened the cranks well enough but after several episodes of hand tightening I got a torque wrench and was really suprised how tight 40nm was - certainly well beyond what I expected.
When I swapped my blue gear I greased with standard lithium grease and it's got quieter by a fair bit
 
  • Like
Reactions: Bikes4two

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
I installed mine in an old 10 speed Raleigh last Fall. I wasn't pedalling hard enough to really get any assist out of it unless I went to turbo, and even then I topped out at 18 mph.

Put in OSF firmware with embrusa configurator. Running his default motor settings. Bike perked up quite a bit. I enjoy riding it in Eco. Can get past 22 mph. Have seen one bug, an unexplained hangup of the display, but has not repeated. Have 700 km on the bike.

I also bought a TSDZB, but still in box. The pedals turn with high effort, much harder than the earlier version. They're supposed to turn easier with a better clutch system? I read somewhere that I must have a gasket binding under the gears?
 
Last edited:
  • Like
Reactions: Bikes4two

Bogmonster666

Esteemed Pedelecer
Jun 6, 2022
343
176
@harrys, your experience is quite different to mine, although I think your usage and expectations are quite different as well. I have the maximum assist speed set to 25kmh so I don't get any assist beyond 25kph and want to remain UK legal in that respect.

I suspect the level of assist between stock and default OSF config is very different - there is less assist in stock firmware. In stock firmware I never use eco (I would consider if the battery was very low as a limp home, but never actually been in that situation), it I'm going for a long'ish ride I use tour and I usually use tour up long steep hills and just go slower using gears. My default for 'proper' rides is tour - mainly because I want moderate exercise and longish range.

If I'm feeling lazy (or at the moment with an ankle injury) I sometimes use speed for shorter rides and on flattish ground it feels like masses of assistance up to the cut-off. This significantly reduced range but is nice for a gentle evening pootle without building up a sweat.

I never use turbo, that feels like ghost pedaling.

I am sure osf has a lot to offer, I'm interested in the higher cadence support (field weakening) and better motor control equating to quieter operation.

I think what the above illustrates is that different people will have different requirements and expectations. One person's 'plenty of assistance ' will be another person's 'is this thing even working?'
 
  • Like
Reactions: Bikes4two

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
I am used to PAS 1 on my ebikes taking the edge of pedaling. There should be a clear difference between pedaling with and without power. I just didn't feel that with the TSDZ2 in ECO. It was very subtle. The embrusa version of OSF is able to do the above.

I could have just ridden in Sport and Turbo, but that didn't seem right, and turbo didn't feel like a turbo should feel anyway.

It could have been that the torque sensor on my motor was reading low, and the OSF just amplified its output. I only ride around at 14-15 mph,
 

Kevin williams

Pedelecer
Sep 16, 2014
62
12
53
Apologies to those who may have read this survey on the Cycling UK forum but as there's no response so far (03May22), I'm posting here as I'm sure there will be greater interest.

Intro

• A bit of a long-ish read, but bear with me if you will – this is all about dispelling urban myth in favour of more factual experiences.
• I am a TSDZ2 user (of around 5 months now and almost 3,000Km) as is my wife and two friends (and I installed all their kits).
• Prior to buying the TSDZ2 I read and watched a lot of comment on the motor and took careful note of the problems that people had had with it.
• Reports of overheating, blue gear failure, sprag clutch problems and axle breakages are numerous but of course forums rarely attract reports of ‘I’ve had a great day/year/experience’ but rather along the lines of ‘something is wrong/broken, how do I fix it’.
• Further, it is often not possible to deduce what type of usage a motor has had e.g. was it a lower powered 36v 250w version taken to the High Street a couple of times a week, or one of the higher powered options with a 52v battery powering along at 750w+ in sub tropical conditions, to say nothing of the rider weight, riding style and terrain.
• Anyway, at a price point of under £300 I didn’t expect precision engineering and in any case, most problems seemed to be fixable by a competent person plus a good level of availability of spares, and forums such as this and Youtube for guidance.

The Survey
My own TSDZ2 experience has been very positive but I am keen to hear of other’s direct (not here-say) experiences with this motor. So to kick off, here’s my input:
• I have had no problems with the motor and neither have the other 3 TSDZ2 users. I’ve done 2,900Km at the time of writing (May 2022) whilst the other users have done, 1,600Km, 1,000Km and 400Km respectively – so no great mileages here and during the winter months of 2021/22!
• I am the only one running Open Source Firmware (OSF) and I’ve only experienced one peculiarity with it in that occasionally the motor ‘drives itself’ i.e. the motor keeps turning without any crank rotation, albeit very slowly and with little power behind it. As I have not fitted ebrakes to my bike (or to any of the others), applying the brakes lightly is enough to stop bike movement and a power on/off resolves the issue. Interestingly, two of the other ‘fits’ occasionally experience the same thing and they are running stock firmware.
• Mechanicals – I like many others, experienced the cranks loosening. I resolved this by buying a torque wrench and properly tightening (as opposed to guessing) to 40Nm. I do re-check for tightness on all motor fixing points and the cranks around every 1000Km – sometimes a bit of tightening is required but not to a degree that you’d say something has coming loose.
• And a comment on chain wear as some owners have commented that mid-drive motors lead to excessive chain wear. After 2,900Km my KMC chain has of course some wear but well within acceptable tolerance and not requiring changing as yet. However, on the 1,600Km bike, I had to change the ‘no name’ chain at around 1,400Km and whilst our riding styles and terrain are similar, I generally use assist level ECO whilst my freind generally uses the next level up - draw your own conclusions?

NB. For chain wear I use a simple 2-wear level go/no-go gauge (0.7% and 1.0%) which from experience is as accurate as measurement with a steel rule and far more convenient.

Phantom Problems
• What are these you might ask? Well, these are problems that I may have initially thought were down to the motor but turned out not to be so, e.g. the loose crank issue as mentioned above.
• Another one was the motor suddenly and randomly powering down. This took a while to resolve and it turned out to be a dodgy BMS in the bottle battery.
• And yet another one was an annoying creak that despite my prolonged efforts, persisted, until that was, I re-greased the headset bearings and peace reigned again.

Over to You
It would be really useful to know of other’s direct experiences with their TSDZ2 (or those of a friend to which they can testify to credible accuracy). So if you’ve had any problems, what were they, how were they resolved and most importantly, what motor spec, type of riding (style and terrain) was the motor used for.
And of course if you’ve had a positive experience, then please share that on here too.

Thanks for sharing and happy e-biking
Update from original post , my original tsdz2 on planet X MTB is now up to 17000 miles with only a blue gear change at 9000 miles,my tsdz2b on an orbea carpe hybrid is now up to 6000 miles and wife's 4500 miles with no issues at all, no maintenance performed or OSF installed
 

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
I installed the TSDZ2B last week on a older 26" Giant RIncon. The unit went in fairly easy. One of the pedals was frozen into the old arm and couldn't be re-used. The chain stays hit one of the screws holding the motor cover. I filed down the boss for the screw to get full clearance. Also some heavy duty hacks on the bracket that bolts to the motor and sits under the chainstay. It was too big, It also forced the use of a rear mounted kickstand as the old one no longer fit,.

IMG_4216.JPEG

Powered with a 48V battery I built two years ago.

Initial thoughts. The TSDZ2B crank shaft is too stiff to be turned by fingers on the axle, and still pretty stiff after the arms/pedals are attached. I cannot tell though when it has to turn the chain and move the bike. I don't sense any more drag.

With stock firmware, I could sense the motor in all four assist levels, with noticeable assist in Speed and Turbo, but ECO is minimal. Throttle takes time, but will pull the bike up to 22 mph if I wait. I put in the OSF/embrusa firmware though.

The big test is to see if my wife will ride it over her current cadence sensor hubmotor bikes. She's not a shifter. She rides in the same gear, and will just push harder in hills. Rarely used the throttle. Might want to put it in Speed mode for her.
 
  • Like
Reactions: Bikes4two and Oldie

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
997
428
Havant
I installed the TSDZ2B last week on a older 26" Giant RIncon. The unit went in fairly easy. ....................... It also forced the use of a rear mounted kickstand as the old one no longer fit,
Nice job there - I love the discrete look you get when you can hide the battery away.

Powered with a 48V battery I built two years ago.
Nice one :)

Initial thoughts. The TSDZ2B crank shaft is too stiff to be turned by fingers on the axle, and still pretty stiff after the arms/pedals are attached.
So is the 'B' stiffer than the original TSDZ2?

With stock firmware, ........I put in the OSF/embrusa firmware though.
Good to hear that the 'B' takes OSF - obviously fitted with the 'old' controller still then? Which version OSF?

The big test is to see if my wife will ride it over her current cadence sensor hubmotor bikes. She's not a shifter. She rides in the same gear, and will just push harder in hills. Rarely used the throttle. Might want to put it in Speed mode for her.
My wife could never get the hang of the triple shifters on her earlier bikes - I even tried twist grip shifters to make things easier but they were too stiff for her wrists - ended up with a Rohloff (eek £££s) which I transferred to her ebike conversion - an Alfine 8-speed would have been adequate for the TSDZ2 fit-out though.

She'll get much more out of the TSDZ2 (feel and performance) if she does learn to shift up/down through the gears as I'm sure you know - good luck with the retraining :rolleyes:
 

Simon Knight

Pedelecer
Aug 19, 2009
49
21
Apologies to those who may have read this survey on the Cycling UK forum but as there's no response so far (03May22), I'm posting here as I'm sure there will be greater interest.

<snip>

Over to You
It would be really useful to know of other’s direct experiences with their TSDZ2 (or those of a friend to which they can testify to credible accuracy). So if you’ve had any problems, what were they, how were they resolved and most importantly, what motor spec, type of riding (style and terrain) was the motor used for.
And of course if you’ve had a positive experience, then please share that on here too.

Thanks for sharing and happy e-biking
Hi,
I installed the 48v version to my ICE recumbent trike. The kit was supplied by Woosh and was straight forward to install although I needed to source a longer cable for the rear wheel sensor. The kit came with a large 720Wh battery and I could achieve high milage per charge riding in flat North Lincolnshire. So far so good. One negative was the indication of battery state did not change until the battery level had dropped quite a long way. The suspicion is that the meter was calibrated for a 36volt battery rather than the 48volt being used. In the end I fitted a small digital voltmeter which was cheap and more useful than the standard five bar display.

Initially I was happy but after a couple of hundred miles I started suffering from pain in the sides of both knees on every ride. This came on at the start of the ride so was not related to milage. I believe the issue is due to the wider Q factor (distance between pedals) and probably more importantly that the pedals/cranks were not centred on the trike.

I reverted to the old non electric cranks and pedals and the knee problem ceased.

Summary : Good system if the pedal location works for you.
 
  • Like
Reactions: Bikes4two

Peter.Bridge

Esteemed Pedelecer
Apr 19, 2023
1,204
540
..One negative was the indication of battery state did not change until the battery level had dropped quite a long way. The suspicion is that the meter was calibrated for a 36volt battery rather than the 48volt being used. In the end I fitted a small digital voltmeter which was cheap and more useful than the standard five bar display...
I have this problem with all battery level displays, 36v, 48v, kit based and on my wife's Wisper, It seems like I get a 1/3 of the way through the useable voltage drop before the battery displays move off 100%
 

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
997
428
Havant
If you look at the discharge curve of a lithium ion cell you'll notice that the terminal voltage difference for a large part of the curve is not a lot making it impossible for those led displays to be very accurate especially given that the voltage dips under load (voltage sag).

The most reliable way to monitor battery usage is to measure the Wh (watt hours) used with a power meter (aka coulomb meter).

@Simon Knight - have you a link to your volt meter?
 

Attachments

Simon Knight

Pedelecer
Aug 19, 2009
49
21
If you look at the discharge curve of a lithium ion cell you'll notice that the terminal voltage difference for a large part of the curve is not a lot making it impossible for those led displays to be very accurate especially given that the voltage dips under load (voltage sag).

The most reliable way to monitor battery usage is to measure the Wh (watt hours) used with a power meter (aka coulomb meter).

@Simon Knight - have you a link to your volt meter?
Here is a link to the volt meter I used : https://www.amazon.co.uk/dp/B09XTMJTHW?ref=ppx_pop_dt_b_product_details&th=1
I re-purposed a nylon tube clamp originally used to fit a bicycle light to mount the meter. I also "gobbed" up any slots in the back plate with silicon to waterproof the meter.

I decided against fitting a power meter as suggested by Bikes4two as the full current has to pass through the meter meaning that they require cooling which presents a challenge making the wiring fully waterproof. Instead I wired a power meter into the charging cable to get any idea of how much energy was being stored in the battery.

I found the volt meter useful as it allowed me to judge when the decreasing voltage was entering the steep part of the discharge curve indicating it was time to recharge. From memory my 48volt battery would do many miles above 50 volts and require recharging at approx' 44 volts. I'm not saying these figures are accurate but I aimed to recharge at 20% capacity in an effort to protect the battery and prolong its life.
 
  • Like
Reactions: Bikes4two

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
997
428
Havant
  • Thanks for the link to the volt meter @Simon Knight
  • I decided against fitting a power meter as suggested by Bikes4two as the full current has to pass through the meter meaning that they require cooling
  • I've tried a couple of different power meters mounted on the bike but they've never got warm enough to consider cooling - maybe we use different power levels. I typicaly draw 3-5amps in ECO mode, occasionally nearer 10A in the next level up.
  • However, I no longer use a handlebar mounted power meter as they are even more clutter on my handlebars. Instead I always recharge the battery through the power meter and note the miles ridden and starting voltage and the WattHours to fully recharge.
  • Obviously this isn't a 100% accurate method of determining power used during the ride but is near enough for me to have gained a very good estimation of cycling range under different cycling conditions, all without having to measure the voltage as I ride.
  • Having said that, I will rig up a little volt meter like you've used, to plug into the battery should I be suffering range anxiety :cool:
 

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
I didn't think my wife would like the bike, but she's a good sport and put a hour's ride on it. She said it was a lot of work. I know she hates it.

I saw there was a hybrid mode available, easily reached with VLDC5. Put it in Turbo level and hit the lights button 4 times. Wow, this was quite nice. It doesn't need to sense much torque to activate the motor. I found it pretty much like riding one of my cadence sensor hubmotor ebikes.

I went out and rode 16 miles in hybrid. Also tried Cadence mode and cruise mode, but they're kind of unpredictable with power delivery at low speed. Hybrid gives our typical 12 mph speed with very little work in Eco, but I can run it past 20 mph with pedaling and still in Eco. It does use a lot more battery, but prior to this, it wasn't using any battery, so it just shows the motor is working more.
 
  • Like
Reactions: Woosh

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
Oh well. My wife thought OSF hybrid mode was too powerful, so I set it in EMTB mode on the road, and she dawdled along at 8-10 mph, and said she was in Sport level assist most of the time.

I spent the rest of the daylight hours, and it was a pleasant afternoon, sitting outside with my PC adjusting the two boxes for ECO in the power and torque assist mode and seeing how the bike responds.. I want to reduce power in Eco for hybrid. Still not there yet. ALmost.

Setting both to 30 really cut the power, so I'm currently at 50 and 40. I've also kicked up the EMTB parameters to make that stronger in Eco/Tour.

By the way, the image is not mine, but one I grabbed off the internet. Hmm, I see their emtb settings are higher than mine which were recommended at 2,4,6,10.

I fear my STM dongle is going to break from the constant use. This is one of the things I dislike about firmware, is the endless tests.


screen_2.jpg
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
20,882
8,517
61
West Sx RH
I haven't dabbled to much with the settings , though have done with the hybrid mode setting of PAS & Torque assist. After a year of using the hybrid mode I then changed to torque assist mode only and find I prefer this setting best out of the two.
Though I have changed the four level settings, it is pretty pointless for me to do so as I find I only ride in Eco mode when towing my trailer and find I never need the higher power assistance.
 

Bikes4two

Esteemed Pedelecer
Feb 21, 2020
997
428
Havant
I haven't dabbled to much with the settings , ......... it is pretty pointless for me to do so as I find I only ride in Eco mode ..........I never need the higher power assistance.
+1 on that one Neal - I set my own (2) motors to power assist mode right from the start some 12+ months ago and haven't touched them since - I couldn't even tell you what those settings are unless I look up my notes and I don't really care as the bikes ride like they do and I've adjusted to their ride characteristics (and only use ECO mode too).

OSF is very flexible and offers many permutations and some folks love to play around with the different settings whilst I just prefer to get out and ride.
 

Nealh

Esteemed Pedelecer
Aug 7, 2014
20,882
8,517
61
West Sx RH
I have unticked the street power mode and it powers beyond 20mph , If I set the speed limit in the display and the OSF it appears street mode unticked simply over rides the speed limit.
 

harrys

Esteemed Pedelecer
Dec 1, 2016
351
93
73
Chicago, USA
I set the OSF/embrusa Eco settings for power and torque mode to 70 and 50, Hybrid mode is default, Everything else is their "proven" settings. This lets my wife pedal her TSDZ2 along at 10-11 mph, about the same speed as her cadence sensor bike. I did the same for my TSDZ2 bike. Both happy now.
 
  • Like
Reactions: Bikes4two

Clueless88

Finding my (electric) wheels
Oct 9, 2022
9
-1
How much would be a fair price for a programming cable? I've made 3 that all work. I'm wondering if I should start churning them out.