TSDZ2 OSF for all displays, VLCD5-VLCD6-XH18, LCD3, 860C-850C-SW102.

Be aware that theoretical W/h's never tally with actual on the bike totals. My batteries the BMS cuts the output ( voltage sag in the last part of the charge ) about 20% ahead of where theoretically it should. I have a test rig where I can put a constand 250W drain and that also confirms that you never get as good as advertised.

I checked this morning the pwm values during the commute. Urban environment, lots of red lights and stops. With the boost active with 250 factor and 25 step the pwm is 254 from light to light. Is this the reason why the battery consumption is higher than with 1.0.0 version? I changed the step value to 50 and the pwn is still 254 for 10 seconds.

Waynemarlow said:

Be aware that theoretical W/h's never tally with actual on the bike totals. My batteries the BMS cuts the output ( voltage sag in the last part of the charge ) about 20% ahead of where theoretically it should. I have a test rig where I can put a constand 250W drain and that also confirms that you never get as good as advertised.

Click to expand...

Indeed, and there's also the influence of the amount of current drained : if you draw 15 amps continuous you'll get less total Wh than if you draw 5 amps because of heat losses in the battery. Battery makers indicate the capacity for a low discharge rate.. That's why Casainho's implementation is really good as he takes into account the internal resistance of the battery for the calculation of SOC.

JohnAnanda said:

Waynemarlow said:

Be aware that theoretical W/h's never tally with actual on the bike totals. My batteries the BMS cuts the output ( voltage sag in the last part of the charge ) about 20% ahead of where theoretically it should. I have a test rig where I can put a constand 250W drain and that also confirms that you never get as good as advertised.

Click to expand...

Indeed, and there's also the influence of the amount of current drained : if you draw 15 amps continuous you'll get less total Wh than if you draw 5 amps because of heat losses in the battery. Battery makers indicate the capacity for a low discharge rate.. That's why Casainho's implementation is really good as he takes into account the internal resistance of the battery for the calculation of SOC.

Click to expand...

Yes, the implementation I did of battery SOC on V1.1.1 is the best until now because it includes the battery resistance on the calculation. What is missing is the effect of ambient temperature and on our TSDZ2 wireless project, we are using a cheap NRF52840 board but there is another NRF52840 board that costs a bit more but includes a temperature sensor - the Arduino Nano 33 -- using this board we could improve even further the battery SOC. Also this board includes an accelerometer, meaning an algorithm could be developed to automatic increase the assist level on hills and reduce on downhills (this NRF52840 supports machine learning which I think would be great to do this task).

mbrusa said:

Mr.Flibble said:

That's very kind of you, but I have ordered a new controller as it has locked up, and gone off all by it's self.

If it was just a change in readings with temp, surely restarting would reset "zero" and fix it?

Click to expand...

Yes, with the restart the zero is reset, but only with "Torque Calibration" disabled.

Click to expand...

I tried turning off torque calibration on my last long ride when it started ghost pedaling, it made zero difference

Reading all these other posts, I suspect it's the heat issue adversely affecting a dry joint.

Surely FOC should be linked to motor RPM, not PWM?

Mr.Flibble said:

I tried turning off torque calibration on my last long ride when it started ghost pedaling, it made zero difference

Reading all these other posts, I suspect it's the heat issue adversely affecting a dry joint.

Surely FOC should be linked to motor RPM, not PWM?

Click to expand...

To understand what it depends on, I need complete information.
So, you had the problem, you disabled "Torque calibration", you turned the display off and on again, but the problem remained.
Did you check after turning on again if the "Pedal torque delta" value was 0?

I'm on a test drive. Much better than 1.1.0, you can say a masterpiece, the hybrid mode is a revelation. But, it's terribly power-hungry. After 12 km on flat terrain, I heated up to 72 °, the consumption was quite high in the 5th mode out of 9 practically around 500W pulls. 3V battery drop after 12km. Where on 0.8 I did 2x as much on a similar support factor (50%), but I never reached 60 ° on the same route. And the power fluctuated in the range of 100-250W with the same sensations of support.

This weakening of the field does not suit me terribly, I like a low cadence, but when I crank the crank around 110, I do not feel any resistance at all and the 350W power consumption jumps in. It is a pity that it can not be turned off because it may be the reason of power-hungry in the lower term, and the effect of 0 ...

hetm4n said:

I'm on a test drive. Much better than 1.1.0, you can say a masterpiece, the hybrid mode is a revelation. But, it's terribly power-hungry. After 12 km on flat terrain, I heated up to 72 °, the consumption was quite high in the 5th mode out of 9 practically around 500W pulls. 3V battery drop after 12km. Where on 0.8 I did 2x as much on a similar support factor (50%), but I never reached 60 ° on the same route. And the power fluctuated in the range of 100-250W with the same sensations of support.

This weakening of the field does not suit me terribly, I like a low cadence, but when I crank the crank around 110, I do not feel any resistance at all and the 350W power consumption jumps in. It is a pity that it can not be turned off because it may be the reason of power-hungry in the lower term, and the effect of 0 ...

Click to expand...

I hope soon to be merged the improvements of all this new firmwares back to my firmware, and for sure I will keep the option to disable the Field Weakening. When we are doing long travels, we want to have best efficiency possible, like doing 100 kms with the same battery I would do only 40 kms regularly. That is why there is the graph of battery used Wh/km, so we can track our energy usage and try reduce it if needed.

hetm4n said:

I'm on a test drive. Much better than 1.1.0, you can say a masterpiece, the hybrid mode is a revelation. But, it's terribly power-hungry. After 12 km on flat terrain, I heated up to 72 °, the consumption was quite high in the 5th mode out of 9 practically around 500W pulls. 3V battery drop after 12km. Where on 0.8 I did 2x as much on a similar support factor (50%), but I never reached 60 ° on the same route. And the power fluctuated in the range of 100-250W with the same sensations of support.

This weakening of the field does not suit me terribly, I like a low cadence, but when I crank the crank around 110, I do not feel any resistance at all and the 350W power consumption jumps in. It is a pity that it can not be turned off because it may be the reason of power-hungry in the lower term, and the effect of 0 ...

Click to expand...

Without torque sensor calibration I ride on level 3 of power asist mode (70%) with PWM of 254 90% of the time. The calibration gave me a narrow ADC range from 198 to 280, and still got to 254 PWM easily. So I increased the maximum ADC to 380 and now some effort is required to get to the maximum PWM, and therefore, the less power consumption.

Nfer said:

Without torque sensor calibration I ride on level 3 of power asist mode (70%) with PWM of 254 90% of the time. The calibration gave me a narrow ADC range from 198 to 280, and still got to 254 PWM easily. So I increased the maximum ADC to 380 and now some effort is required to get to the maximum PWM, and therefore, the less power consumption.

Click to expand...

Good to know regarding manual torque sensor calibration. Easy fix if you are chewing through your battery!!
I’m happy with all the extra power despite some minor overheating up hills so will leave it how it is for the moment

hetm4n said:

I'm on a test drive. Much better than 1.1.0, you can say a masterpiece, the hybrid mode is a revelation. But, it's terribly power-hungry. After 12 km on flat terrain, I heated up to 72 °, the consumption was quite high in the 5th mode out of 9 practically around 500W pulls. 3V battery drop after 12km. Where on 0.8 I did 2x as much on a similar support factor (50%), but I never reached 60 ° on the same route. And the power fluctuated in the range of 100-250W with the same sensations of support.

This weakening of the field does not suit me terribly, I like a low cadence, but when I crank the crank around 110, I do not feel any resistance at all and the 350W power consumption jumps in. It is a pity that it can not be turned off because it may be the reason of power-hungry in the lower term, and the effect of 0 ...

Click to expand...

The first time you try this version it is normal to experience higher consumption, I have experienced it myself. Especially if you keep the same driving habits.
When the motor responds well it is easy to get carried away.
It is about finding the right balance between performance and consumption, there are the parameters to do it.
Now for me, of all the versions I have tried it is the one that consumes the least.

As for Field Weakening, it is activated only when the PWM reaches the maximum, but the loss of efficiency occurs only at high cadences, I do not understand the need to reach 110.
I will however add the ability to disable it, maximum customization is a priority for OSF.

mbrusa said:

I will however add the ability to disable it, maximum customization is a priority for OSF.

Click to expand...

Będzie super

mbrusa said:

I do not understand the need to reach 110.
I will however add the ability to disable it, maximum customization is a priority for OSF.

Click to expand...

Guys look at your engine watt output on your display. If you are using 350W’s on level 3 and not 250W’s as previously on level 3, then be logical and ask yourself where that extra 100W’s is being used. It’s obviously not simply disappearing into thin air, it can only be extra power at the wheel and heat. Now if all the 100W’s was now only producing heat and no power, then you would fry you motor in a few minutes. So where is the only other place it can go, power at the wheel.

Just because you cannot feel the extra power, doesn’t mean that it’s not happening. Why not all power output and no extra heat, that’s down to the engine design specifically for low cadence and low power outputs to power millions of commuter bikes. The designers never ever considered a few 1000 converted mtbs pulling 650W.

So how do you now get back to the 250W’s on level 3 ? Change the settings in the set up menu, it’s not difficult and if one sits down for 5 minutes and reads the setup notes, it’s actually not hard to do.

I set mine up by firstly setting the highest level 5 in my case to produce the max Motor Watts, in my case 650W with my max pedal power at around 50 -60 pedal rpm in a middle gear. Slow down and then pull away peddling as hard as you can and watch the motor Watts on the display. You can also lightly touch your brakes if you are on flat road to add a bit of drag but you will eventually heat the engine up. Please don’t just stand on the pedals in a low gear from standstill, the blue plastic gears just do not cope. Now you may have to do this 3 or 4 times adjusting the numbers to just reach max Watts.

Once you have found the top level, I simply divide that number into 5 levels and enter the numbers into levels 1-4. MBrusa has given basic settings to get you going, the exact settings to suit your bike and you, can only be finely tuned and inputted by you.

Why the higher cadences, most experienced riders who also ride analogue bikes regularly will be in that 75 - 100 rpm cadence which is the most efficient for the human body. I swap rides constantly between analogue and EBike and it’s hard to drop back to lower cadences when on the EBike, so I”ll take the slight extra use of the battery and pedal at my normal cadence of 90 to simulate my analogue bike. Both bikes are setup suspension and geometry wise the same, just one is heavier than the other.

HI everyone. I have been on that forum for a long time ( 2015) as a reader. Only few post Im sorry... But it is gonna change soon
I have been using TSDZ2 since first model become available on the market. Now I have 4 bikes converted. I always used OEM firmware mostly because limited display options. I like things to be made and look like OEM. Now we have finally a choice. 860 C and other displays looks very good and promising so for the first time i have decided to go with OSF. For me most important is efficiency and distance and reliability of the motor . ( besides its know mechanical issues)Acceleration and top speed is on the end of the list if i have to choose. My question is what OSF firmware would U choose in such a case? 1.01 or 2.0?
My riding style.
On the OEM firmware Im using XH18. Is a good display with the knob . I use the knob to switch assistance mode on the fly. So i start with eco. Then i use most of the time "sport" mode. I pedal hard like on the normal bike. Then when I need some extra power i go for "Turbo" for some short time and if the road is good I use standard mode. This style of riding gives me around 90 km ( around 57 miles) on my power pack ( 48 V , 17.5 A) with "750" TSDZ2 motor. and average speed around 30 km/h.
So what would U suggest . What firmware and set up ? I think I do not need 20 steps of assistance. Or may be I do?
Im waiting for programing cables and bootloader from Eco cycles and im going to reprogram on Monday. Any suggestion would be very helpful as Im confused with all firmware's and options available.

martin.shane said:

HI everyone. I have been on that forum for a long time ( 2015) as a reader. Only few post Im sorry... But it is gonna change soon

Click to expand...

I've been down the road you are going. Started with XH18 and OEM firmware. Then 860C, v1.0.0, v1.1.0 and now v20.1C.

From my experience, I recommend 860C display and v20.1C. Also suggest get a temp sensor. (Eco cycle sold me a controller with temp sensor soldered in. Requires you to attach sensor to motor). OSF can provide lots of additional power at the expense of extra heat, especially with initial settings. Think of the temp sensor as insurance while you adjust the settings to your specific riding style.

Cheers!

martin.shane said:

HI everyone. I have been on that forum for a long time ( 2015) as a reader. Only few post Im sorry... But it is gonna change soon
I have been using TSDZ2 since first model become available on the market. Now I have 4 bikes converted. I always used OEM firmware mostly because limited display options. I like things to be made and look like OEM. Now we have finally a choice. 860 C and other displays looks very good and promising so for the first time i have decided to go with OSF.

Click to expand...

I am not sure if you are aware, but if you like to keep the bicycle look like OEM, you can make a minimalist build with TSDZ2 and using wireless only and no display -- see here our TSDZ2 EBike wireless controller: https://opensourceebike.github.io/

Thanks Casainho and seattlesockey
Casainho , WOW!. As always awesome work! I will try it for sure.
I will need to check what controllers i have on my motors. As most of them I have bought in 2020 and 2021 im afraid that it is a new controller .
For now I have some OEM look with SW102 from Eco Cycles with original firmware. It looks great but now way i can use it with OSF without opening it.
I will try wireless. It would be awesome.

martin.shane said:

For now I have some OEM look with SW102 from Eco Cycles with original firmware. It looks great but now way i can use it with OSF without opening it.

Click to expand...

The TSDZ2 EBike wireless controller is just like a display and costs only 25€ in materials, meaning at anytime you can swap it with other display, if you prefer.

Really beautiful bicycle!! I guess using less wires will keep like that. For instance, I have short cables from the brake sensors to the fully wireless remote. And I do not install the speed sensor, If I am on a long travel, I just attach to the handle bar the GPS display for navigation that will also show me the EBike speed using the GPS, the EBike battery state, assist level, etc.

Thank U Casainho
Wireless with Garmin edge would be perfect. Im frame builder and i have strong mechanical skills however my electronic skills are limited. I mean electricity YES but electronics not so much. So I must wait for some " ready to go" wireless module . Maybe Eco Cycle will do it in a near future... I hope so.
For sure for me going wireless i a best option.
Im building 29r now on own frame and solving some chainline issues . It will be a bikepacking bike. The idea is to make it look analog as much as I can do , including internal routing for cables and so on.
For me in OSF most important is that I can see power consumption and how hard I need to pedal to make a long distance ( human power indicator). I will try to find the way to optimize it for a long distance and as much support I can get.
I have another bike , full MTB and this one will be optimized for Torque. I think it would be grat option if we could use profiles .
It would be helpful also when somebody like me is using 2 different battery packs ( small one is 14 Ah , and bigger one 17,5 Ah - 2 different cells are use so different parameters in set up)
Again , thank u for U hard work.

martin.shane said:

Thank U Casainho
Wireless with Garmin edge would be perfect. Im frame bulider and i have strong mechanical skills however my electronic skills are limited. I mean electricity YES but electronics not so much. So I must way for some ready to go wireless module . May be Eco Cycle will do it in a near future.
For sure for me going wireless i a best option.
Im building 29r now on own frame and solving some chainline issues . It will be a bikepacking bike. The idea is to make it look analog as much as I can do , including internal routing for cables and so on.
For me in OSF most important is that I can see power consumption and how hard I need to pedal to make a long distance ( human power indicator). I will try to find the way to optimize it for a distance and as much support I can get.
I have another bike , full MTB and this one will be optimized for Torque. I think it would be grat option if we could use profiles .
It would be helpful also when somebody like me using 2 different battery packs ( one is 14 Ah , 2nd 17,5 Ah - 2 different cells are use)
Again , thank u for U hard work.

Click to expand...

Oh, frame builder... I am in contact with one that will build a micro cargo bike for me and where I will install a TSDZ2

I already did some long travels but never really bikepacking, but I wish to do it some day, I really like the adventure side

So, for optimizing to low power for bikepacking travels, our firmware is the best you can get and I would:
- reduce current ramp
- disable field weakening
- config motor max power like if my battery have 300Wh, I would configure motor max power to 75W, so 300 / 75 = 4 hours continuous riding. Assuming I have the same amount of ascendings as descendings, etc, I assume I will use the motor only 60% of the time, then I hope to have battery for 6.5 hours of riding
- reduce battery low voltage value to be the minimum as possible, to fully discharge them and so get the max energy possible
- configure the battery resistance to have the best accuracy on the battery SOC

And yes, our fully wireless remote controller, other than controlling TSDZ2 with the 3 buttons, the 4th button is used to change the page of Garmin Edge!! The remote connects simultaneously to the TSDZ2 Ebike wireless controller and to the Garmin Edge!!

Profiles? could be implemented on the mobile app. Mobile app reads and sends the configurations to the TSDZ2 EBike wireless controller, so various profiles would be on the mobile app and each one would send each configuration to the controller.

If you have strong mechanical skills, did you thought already if there is a way to reduce the noise of TSDZ2 on the mechanical gears?

So, for optimizing to low power for bikepacking travels, our firmware is the best you can get and I would:
- reduce current ramp
- disable field weakening
- config motor max power like if my battery have 300Wh, I would configure motor max power to 75W, so 300 / 75 = 4 hours continuous riding. Assuming I have the same amount of ascendings as descendings, etc, I assume I will use the motor only 60% of the time, then I hope to have battery for 6.5 hours of riding
- reduce battery low voltage value to be the minimum as possible, to fully discharge them and so get the max energy possible
- configure the battery resistance to have the best accuracy on the battery SOC

Click to expand...

THANKS a lot.

I

f you have strong mechanical skills, did you thought already if there is a way to reduce the noise of TSDZ2 on the mechanical gears?

Click to expand...

[/quote]
I was quite lucky as for my four motors only one make a very strange noise. It have pushed me to do some kind of reverse engineering . TSDZ2 is a very cheap motor but it has some problems. I think every single thing makes it noisy.
Precision of part making, loose bearings, wrong design of axel ( crank shaft) and so on.
Im working with my friend on that. Starting with redesign internal parts , find better bearings and we will see how it will work.
I have 3d printer and it makes design easier but making later prototypes on CNC machines is expensive . But I have invested some money on that . I should have some results very soon.
In todays technology , when everything is changing very fast in eclectic motors , batteries and mobility I strongly believe in idea of interchangeable motor. Frame will not change so much over next few years. Motors and batteries YES. All modern electric bicycles form today with Bosh, Yamaha and Shimano motors will be worth nothing in 2-3 years. New versions will come, same expensive as today and old version will become obsolete. ) like in a sports cars) This is why i follow that idea and try to make it stronger , modern geometry frame with possibility to change motors and batteries in a futures for whatever will be available.
In a mean time i think TSDZ2 is a great , small and cheap motor and with a lit bit of work from people like U ( software side) and like me from engineering side can change this little thing to something very competitive. And Im gonna proof it. Very soon. Riding from Poland to Katmandu with trailer and solar panels. Big project I believe it can be done...

I think Idea of wireless controller in super cool.
What are dimensions of the box?
Is like a separate controller and it will work with the new version of TSDZ2 controller?
And the most important question. Can I buy from U such a box. Off course in a retail price.
Let me know
Anyway i will donate U work. Tons of work has been done and a lot of cost. I know how it is...

martin.shane said:

I think Idea of wireless controller in super cool.
What are dimensions of the box?
Is like a separate controller and it will work with the new version of TSDZ2 controller?
And the most important question. Can I buy from U such a box. Off course in a retail price.
Let me know
Anyway i will donate U work. Tons of work has been done and a lot of cost. I know how it is...

Click to expand...

You can find the build instructions here: https://opensourceebike.github.io/ebike_wireless_controller.html

I am not selling anything, the idea is that user should buy the components and DIY. We made a circuit with the smallest components possible and widely available, as also cheap. Maybe in future, if the interest for the project grows, that other shops that now sells the TSDZ2 and displays with our OpenSource firmware, may sell the TSDZ2 EBike wireless controller.

That box was 3D printed and is very simple but I do not provide yet the 3D files I used because I used a DC-DC converter that is not the one we recommend, when I receive my new DC-DC, I will shared the new files for 3D print. But you can do as is easier for you, any simple plastic box is ok.

You are asking for the new version of TSDZ2 controller - if you can, then donate to me (or other interested developers) that controllers so I can start the development work.

Thanks
I will donate.
So I need to wait for some company make it commercial. Its fine. For now I have 860C and new controllers so i need to wait for some firmware for that.

Hello guys, thank you all for awesome work!
I've been reading a lot here but it is my frist post.

I made many tests of different versions with tsdz2 48V, 52V battery, lcd-3 and 860c also, and I found something important.
I have a test route by my house, 4km steady uphill tarmac ride. I made tests in one day, same ambient temperature, same speed, same effort by me, same cadence, same motor wattage (on display I tried to keep 500w - power assist).
In new firmwares like 1.1.0 or 20.1C - temperatures are much higher on the top of the 4km hill ride - 85-90c compared to 50-55c on firmwares 0.8.0 and lower. So after few tests I settled with old 0.20 on lcd-3 because I like eMTB, and I found it most stable.(0.8 with newer display had weird motor stopping behaviour, and no emtb)

I really don't know what is inside different firmwares, but from what I can see, after 0.8.0 - high cadence mode was introduced - maybe thats the cause? Every version above 0.8 is heating my motor much more.

If you guys could make version like mbrusa 20.1c with all the awesome features like eMTB and Hybrid mode, all the graphs, but normal temperature rise like old firmwares, I would be sooooo happy.
For me temperature is the most important thing, more important than all the new features. I am riding MTB bike. Maybe in city it is ok, but in mountains heating is the biggest problem. Cadence in old 0.20 firmware is enough for me. I am riding it for a month now.

Thanks again!