New "TSDZ2 Torque Sensor Central Motor"

Starting in a higher gear is not good for the blue gear. You want to start off in a gear that does perhaps a max of 10mph or so at redline speed (90 cadence). Take it a little easy until you get over 50% redline speed. Take the motor to redline in 1st gear, then squeeze the throttle 1/4 second before you squeeze the shifter, then release both all while maintaining cadence. Repeat until you reach cruising speed at or near redline for that gear. I can shift fast and smooth this way but it takes a little getting used to. If you are concerned about trashing your gear start off riding in eco mode and get used to shifting in eco mode to get up to cruising speed. Then move up to the next level and master that. It is much more difficult to destroy the blue gear in eco mode rather than the higher pas levels.

As for hills, make sure you climb at least 75% of your redline speed for the gear you are using. If you drop down near 75% then you need to pedal to maintain at least 75% redline speed. What is really bad for the blue gear is torque at slow rpms and braking with the motor still going.

I think ours is about the same - 15mph flats and 20mph with minimal pedaling. However battery voltage makes a huge difference, and the difference between say 56v and 50v is substantial. You should probably get a little more speed with a full charged battery as Cecil, his carrier, and I weigh in at maybe 100kg.
 
jur said:
I have been commuting with mine (Melbourne), and been using it on Tour setting. Occasionally on Speed. Almost never on Eco. For pulling away, i have been a little careful not to have my pedal at high 2 o'clock position (which is my habit from years and years of commuting and other cycling), or if I have, not to pull away with a lot of torque on the cranks, because the motor controller multiplies that torque.
So the short answer is, just pull away a little gently, in Tour setting, and a bit more gently in Speed. In Eco, never mind, pull away normally.
Click to expand...

Thanks, good advice to not crank too hard. Have to say though, I'm going to find it hard not going straight to the fastest setting - I've been feeling the need for speed as soon as I got onto the bike!

Perhaps the bronze cog will take a bit more bashing in terms of pulling away in a higher gear? I have one on the way and will likely swap it out even before the blue one gives up the ghost.
 
John and Cecil said:
Starting in a higher gear is not good for the blue gear. You want to start off iWhat is really bad for the blue gear is torque at slow rpms and braking with the motor still going.
Click to expand...

This is really going to take some adjusting I think - I'm so used to just ticking over on the highest gear and tickling the brakes as required when going through a pedestrianized area. And then sometimes being so lazy that I don't even change down before cranking hard to get going again.

John and Cecil said:
I think ours is about the same - 15mph flats and 20mph with minimal pedaling. However battery voltage makes a huge difference, and the difference between say 56v and 50v is substantial. You should probably get a little more speed with a full charged battery as Cecil, his carrier, and I weigh in at maybe 100kg.
Click to expand...

Have got the battery on charge now. Any idea how long a full charge usually takes?

P.s. I have to ask... who or what is Cecil?? Judging from your pic he could be a Warner Bros character...
 
I would be careful with the metal gear too as we do not yet know what will be the next weakest link down the line. The nylon gear protects the rest of the motor from damage.

We go pretty fast from a start. Make sure you are in 1st gear and ease into your torque and you should be fine. We have 400 miles so far with no issues and we weigh about 40% more than you.

This is Cecil (he is also on the back of the Acme Roadrunner Chaser in our avatar):

DSCN2829.JPG

coyotes5.jpg
 
John and Cecil said:
I would be careful with the metal gear too as we do not yet know what will be the next weakest link down the line. The nylon gear protects the rest of the motor from damage.
Click to expand...
Very good point.

John and Cecil said:
We go pretty fast from a start. Make sure you are in 1st gear and ease into your torque and you should be fine. We have 400 miles so far with no issues and we weigh about 40% more than you.
Click to expand...

That's good to hear. My next issue is going to be how to control all of my reckless friends who, no doubt, will want to have a blast on it as soon as they see it in action!

John and Cecil said:
This is Cecil (he is also on the back of the Acme Roadrunner Chaser in our avatar):
DSCN2829.JPG
Click to expand...

Well isn't he a handsome fellow! I feel like I should find myself a navigator too...

John and Cecil said:
coyote5.jpg
Click to expand...
:lol:
 
Or a trained wild pig down there in NZ :) I have considered that quite a bit, a trained wild pig with armored back protector, helmet, googles, (and his own tusks!) riding on the back of a supermoto motorcycle and squealing at cars to get out of his way.
 
Can I ask what the reliability of these motors is like compared to the Bafang ones?

I'm really liking the torque assist over the cadence assist that the Bafang ones offer and the motors look very easy to service from the youtube videos I've watched.

I'd be looking at about 60km/day commuting if that makes a difference.

Also is there any real life benefit to going for the 52v 750W model over the 48v 750W model?

Thanks.
 
John and Cecil said:
Or a trained wild pig down there in NZ :) I have considered that quite a bit, a trained wild pig with armored back protector, helmet, googles, (and his own tusks!) riding on the back of a supermoto motorcycle and squealing at cars to get out of his way.
Click to expand...

Now that would be a sight for sore eyes...
 
TSDZ2 original firmware has a hidden assist level and I just added it to KT-LCD3 (thanks to hurzhurz!!) -- now I can choose between 0 up to 5 on assist level.

Here are the assist level average current:
0 - motor disable
1 - 4 amps
2 - 6 amps
3 - 12 amps
4 - 18 amps
5 - 18 amps and torque sensor more sensible

The new assist level 1, I think is good for the ones that has a light ebike, slim tires and use small energy. Good for use with a small and light battery.


1-1.png


The firmware for KT-LCD3 to be used with TSDZ2 has the following unique advanced features over TSDZ2 original LCD:

- 5 assist levels (1 extra level)
- battery voltage (volts)
- motor electric power (watts)
- rider human power measured on pedals (watts)
- motor/battery current (amps)
- trip total in Watts/hour
- rider force on the pedals (Nm)
- pedals cadence (RPM)
- brakes state
 
John and Cecil said:
Very nice Casainho! Are the assist power levels adjustable too?
Click to expand...
That would be great, right?? No they are not. But hurzhurz is getting more and more knowledge and I hope we can get it...
 
Yes, adjustable power levels for the assist modes would be great, as well as a way to ramp up the power levels of the individual assist levels in relation to cadence (to protect the nylon drive gear). The new low gear is nice though, especially for people that really only want the motor for climbing hills. When assist is off the throttle is off too.
 
casainho said:
TSDZ2 original firmware has a hidden assist level and I just added it to KT-LCD3 (thanks to hurzhurz!!) -- now I can choose between 0 up to 5 on assist level.

Here are the assist level average current:
0 - motor disable
1 - 4 amps
2 - 6 amps
3 - 12 amps
4 - 18 amps
5 - 18 amps and torque sensor more sensible

The new assist level 1, I think is good for the ones that has a light ebike, slim tires and use small energy. Good for use with a small and light battery.


1-1.png


The firmware for KT-LCD3 to be used with TSDZ2 has the following unique advanced features over TSDZ2 original LCD:

- 5 assist levels (1 extra level)
- battery voltage (volts)
- motor electric power (watts)
- rider human power measured on pedals (watts)
- motor/battery current (amps)
- trip total in Watts/hour
- rider force on the pedals (Nm)
- pedals cadence (RPM)
- brakes state
Click to expand...
Good work Casainho and also hurzhurz.
 
Does anyone else have a slight gap underneath their motor near the chainring?

It seems to be slightly open (see motor chassis between wires and chainring):

zutxuAE.jpg


Just wondering if I'm putting some kind of undue strain on the motor chassis to cause this?
 
It looks like the plastic cover over the big gear may not be seated properly, you would need to remove the chainring to seat it properly.
 
Marin said:
It looks like the plastic cover over the big gear may not be seated properly, you would need to remove the chainring to seat it properly.
Click to expand...

Thanks, wonder why that happened. Will have a look when I get a chance.
 
Fresh good news!! Seems that I got FOC working on our OpenSource firmware for TSDZ2 :)

Yes, it is to much tecnhical however you can see the comparison videos and hear the sound as also see the battery current used. Also I provide some graphs of real time data like motor speed, phase current, phase voltage and calculated FOC angle.

Why is this important? Because seems the path is cleared to move forward to complete our own firmware. Also I think this is the very first time FOC implementation is done as OpenSource for this cheap ebike motor controllers, that do some "magic" to keep the hardware price at very low cost. This is a great opportunity to learn about the technology for control the brushless motors.

casainho said:
Got an experimental code working near what is expected - I will explain:

On purpose, I put an wrong angle value at startup and I see my motor asking to much current when I increase the speed. When I put a correct value of 0, it simple don't ask much current is about 0.6 amps on full speed and I think that is because it has low current/load.

So I put an angle of 10: #define MOTOR_ROTOR_OFFSET_ANGLE 10

Here the the experimental code, that calcs Iwl and V, than it does asinf and convert to angle values. Yes, I think some part of the code may not doing correct angle conversion.... can you please review?

The code prints the variables and also puts the negative value to ui8_angle_correction, that is used to correct the angle.

I did 2 tests, by keeping the ui8_angle_correction always at 0 (angle offset error of 10) or by apply the value calculated. I show t
he graphs and logged as also videos, where you can see the power supply current for each case and listen the motor noise.

Code: 
  ui32_l_x1000000 = 135;

  ui32_temp = ui32_w_angular_velocity_x10 * ui32_l_x1000000;
  ui32_temp *= ui32_i_phase_current_x2;
  ui16_iwl = (ui32_temp / 2000000);
  
  f_temp = asinf((float) ui16_iwl / (float) ui16_e_phase_voltage) * 5732.0;
  ui16_temp = f_temp;

  printf ("%d,%d,%d,%d\n",
          ui16_e_phase_voltage,
          ui16_i_phase_current,
          ui16_motor_speed_erps,
          ui16_temp/100);

  ui8_angle_correction = - (ui16_temp/100);

FOC calculation disabled:
- speed ERPS = 450
- phase current = 3.5A

FOC calculation enabled:
- speed ERPS = 350
- phase current = 0.5A

FOC calculation disabled


FOC calculation enabled



FOC calculation disabled
[youtube]-cUIJlzcs2c[/youtube]

FOC calculation enabled
[youtube]8uOsVOv5oS4[/youtube]
Click to expand...

casainho said:
I tested again just to make sure it is working. This time I setup the initial angle as zero and the adjust FOC algorithm will put a negative angle value. I also decided to low pass filter the FOC calculated angle.

Also I did another test, I increased the the lab power supply voltage from ~36V up to ~48V and the phase voltage increased from 22V up to 34V (duty_cycle was kept constant but lower than 100%).

Here the results:
- 22V, I phase 1A, 350ERPS, FOC angle 7.5 degrees --> 16ERPS / volts
- 34V, I phase 1A, 525ERPS (4200RPMs), FOC angle 10.5 degrees --> 16ERPS / volts

So, the speed scaled as expected and FOC had to adjust the angle by +3 degrees. Also the speed is near the motor rated 4000RPMs (motor has 8 pole pairs).

I tried to decrease the speed of control loop to something like 50ms and I saw motor kind in oscillations, as I expected. Lowered to the 5ms and works as seen on the previous videos. I am not sure it is at 5ms because control loop code may take more time than that...
Next time I would like to improve this code, starting by measuring the time it takes to process and maybe use a look up table for asin() instead of using that float version...

Click to expand...
 
hi casainho, I'm a very very dummy, but in near future i will do some experimentation.
i wait the eprom programmer, first i will try to increase th voltage of controller from 36v to 48v, flashing a new standard rom 48v on a 36v controller.
I see from the video the new foc. But on the road are there real advantage? low ampère isn't low watt? and low assistence of the motor?
excuse for my bad english :(
 
Hi all,

I have bought one of these motors (48V 750W) a couple of weeks ago. I tried my first commute with it last Monday and all I have to say is that I'm not that happy with this thing.
First, I never got 750W, well round 700 when the battery is full. Never give me more than 14 amps whatever the setting in the hidden menu. I have connected a CA in order to measure the power, because it seemed to be too low. It's forcing me to pedal quite slowly otherwise the power is decreasing rapidly.
One trip of my commute is 16 km long, with a big downhill at the end. No particular problem to go (excepted that I lost the battery on a small jump, I forgot to lock it...). But on the way back, after 3 km uphill, the controller limited the current at ~8amps. It was a bit warm outside, near 30°C, and the motor was really hot. I made the rest of my trip like this, now the bike is waiting for another solution and I use my other ebike with (unrestricted) Bosch Performance, which is at least as powerful and allows me to pedal faster. The only problem is that the 400 Wh Bosch battery is not made for giving so much power that long, and ages quite fast. I will rebuild it with other cells that will be able to give more amps. I'll keep the TSDZ2 for another bike that will just make some short trips, I'm sure I will fry it rapidly if I use it like I use my Bosh. And I'll keep searching for another high power "torqued" solution for this bike...
 
On long hills you probably need to keep the rpms up pretty high, did you do that? We have only climbed hills about 1km long so far but I maintain at least 80-85% redline speed up the entire hill for the gear you are currently in. Think of it this way, if you maintain 100% redline speed by pedaling along then even though you are climbing a hill the motor is only as stressed as it is on the flats. If you climb long hills and the motor is running at low rpms then the motor will get hot.

Example: My bike will do 10mph in 1st gear on a flat surface with the motor redlined. When I climb a hill I make sure to never drop below 7.5mph. If I climb a long hill I try to not drop below 8.5mph. If I were climbing a really long hill I would try to keep my speed above 9mph or higher.

Perhaps you gearing is too high for hill climbing, they make wide ratio rear cassettes that will reduce your 1st gear redline speed and thus take a lot of stress off your motor when climbing hills (at slower speeds).
 
hi
i have the tsdz2 500w 48v, I have the little display
Xh18 lcd
and in the last days I have a problem with the indicator of battery level.
it is fluctuates.
switch from full level to red in few seconds.
I tried changing batteries.
Can you help me?
 
The torque sensor is really simple. I have nothing to compare it with, so perhaps they all are more or less the same in all torque sensing mid drives. Too me it seems to sense the difference between light pedaling and very light pedaling, sort of. But no difference between hard and very hard. Maybe with some clever programming it could be made to perform differently, but tongsheng seem very reluctant to let anyone but resellers use that programming tool.

Anyway, I recently took out the torque/cadence sensor mechanism when replacing my old worn out one-way bearing. The sensor is fixed with one small screw and sits on tube that surrounds the crank arm axle. It looks to me like it measures the very slight mechanical flex in that tube when you are pushing on the pedals. I thought perhaps it can be tweaked by loosening or tightening that screw and maybe moving it a little? I didn't try, but it would be interesting to know how or if tweaking the sensors positition would change the way it responds.
Click to expand...

Had the same impression, seems like the TSDZ2 is made for casual, laid back riders only.
Probably it's not easy to optimize to motor power to preserve the nylon gear in accordance to pedaling since legs have said to be able to provide fairly high torque at low cadence.
 
Yes, if possible it would be great to reprogram it so that power output is limited at lower cadence levels. Something like 1-10rpm - 2a, 11-20 - 4a, 21-30 - 6a, 31-40 - 8a, 41-50 - 10a, 51-60 - 12 a, 61-70 - 14a, 71-max - 16a. It would make starting out on hills more difficult but it would probably put less strain on the nylon gear.
 
John and Cecil said:
Yes, if possible it would be great to reprogram it so that power output is limited at lower cadence levels. Something like 1-10rpm - 2a, 11-20 - 4a, 21-30 - 6a, 31-40 - 8a, 41-50 - 10a, 51-60 - 12 a, 61-70 - 14a, 71-max - 16a. It would make starting out on hills more difficult but it would probably put less strain on the nylon gear.
Click to expand...

Yes, that would be very nice! Thinking about it though... perhaps set the current limit according to speed. Although I hate to incorporate the speed sensor, this would carry over smoothly to a wider platform of gearing ratios (which have varying cadence to speed ratios) while concentrating on the problem area - the get up and go... once your rolling, stronger current is fine. BUT thinking about it more, what about when you hit the low gear for a hill, you will be going slow, but high cadence. SO, in my dream world of TSDZ2 programming, the current limit would be set by MPH/KPH, BUT OVERRODE by CADENCE. That way, No matter what the gear ratio you run, you would be fine, and if you have to hit a low gear at slow speed, you could still crank the amps safely. I am definitely over complicating that, although that is the type of calculations some of these higher end mid drives make, correct?
 
eyebyesickle said:
I am definitely over complicating that, although that is the type of calculations some of these higher end mid drives make, correct?
Click to expand...
So, just out of curiosity, what higher end motors could take a bit of thrashing?

I think the TSDZ2 is very good value for money but it's probably not quite what I was looking for in the end.

I really wanted to be able to put the same amount of effort in as I do on my manual bike but I wanted the motor to just make me go even faster. In reality what seems to happen is I can go roughly the same speed as I do on a manual but without having to make any effort.
 
You must log in or register to reply here.

Similar threads

R
    • Like
Really good article comparing Bafang vs. Tongsheng vs. CYC
Replies
12
Views
10,379
999zip999
999zip999
E-HP
    • Like
DIY Newb FAQ?
Replies
20
Views
2,345
neptronix
neptronix
B
Intro guide to TSDZ2 & BBS02 (updated 2021)
2
Replies
35
Views
17,158
docw009
docw009
Back
Top