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

[dgc]

Doe's the B02N (EKD01) display male cable connector have all 6 pins showing as I like to use brake sensors, so I normally shave back the outer sheath and pick out the purple and black wires and graft in a pair of cables with connectors for plugging in brake sensors.

 

[Piper J3]

Just a quick data point….

I have TSDZ2B on my Trek Dual Sport for three years and 3000 miles. I’m running early version of OSF that has bug with trip distance resetting once in a while and motor temp graph showing twice value. I have added motor temp sensor and I modified motor cover to allow air flow for cooling. I ride only in dry weather. I make many rides at +30 miles range. I love this motor, OSF software, and 860C color display. Very natural and smooth power addition…

 

[Elinx]

.... I modified motor cover to allow air flow for cooling. I ride only in dry weather. ....

For the sake of completeness, you can find the implementation from @Piper J3 here

 

[mbrusa]

Just a quick data point….

I have TSDZ2B on my Trek Dual Sport for three years and 3000 miles. I’m running early version of OSF that has bug with trip distance resetting once in a while and motor temp graph showing twice value. I have added motor temp sensor and I modified motor cover to allow air flow for cooling. I ride only in dry weather. I make many rides at +30 miles range. I love this motor, OSF software, and 860C color display. Very natural and smooth power addition…

Great.
The temperature graph bug is already fixed in the current version, while the random trip reset will be fixed in the next version.
It's ready, but I need to update the manual before releasing it.

 

[dgc]

I am running the TSDZ2 OSF on the bike in the photo with a muck shield in front of the drive and have not had any problems with overheating, perhaps it may be because I am 75% of the time on the flat and 25% hill climbing, this is a 36v combo but it also applies to my 48v combo on my other bike. I do like the alternative idea for cooling which makes one think is there another way, I am into to high power computer so water cooling come to mind.

 

[mbrusa]

Perhaps someone can help me with some troubleshooting ideas.

I have a TSDZ2 running OSF Version: 860C 0.20.1c-3 (originally purchased from Electrify Bike in Utah back in February 2022). It's been great over the last 2000 miles... until recently.

Some background.. the motor has the cooling mod from early on, and I also limit it to 450W max. No temp sensor (I have throttle instead, but rarely use that). Since I enjoy the exercise, I usually run it at 100-200W, with my legs providing the other 200-300W at 70-90 RPM cadence. When I installed the system, I even followed the steps to calibrate the torque sensor curve to actual pedal force using a spring scale. In other words, the motor has had a pretty easy life.

At random times on my rides, the pedal assist completely cuts out. Display stays on. Battery voltage is normal (even happens with a full charge), indicated speed is normal, even the "Human Power" readout is normal. The assist will resume on its own after some random number of seconds as if nothing happened. During the fault event, the throttle may or may not work. If the throttle works, triggering it sometimes resolves the fault (could be coincidence). The fault generally happens right when I stop pedaling, like at the crest of a hill, or after a quick sprint to cross traffic, but not always. Maybe 10% of those times.

Technical Screen shows the following (typical) during the fault:

ADC battery current	0
ADC throttle sensor	0
Throttle sensor	0
ADC torque sensor	253
ADC torque delta	83
ADC torque boost	83
ADC torque step calc	44
Pedal cadence	62
PM duty-cycle	0
Motor speed	345
Motor FOC	0
Hall sensors	(counting normally)

Once power resumes, all values are similar except now (typically):

ADC battery current	22
PWM duty-cycle	73
Motor FOC	2
(everything else)	(same as above)

My initial thought was the brake cutouts, but those are working normally, and indeed, when triggered they bring the motor speed to 0, even if I keep pedaling, whereas the fault shows the motor still spinning. Triggering the brakes during one of these faults has no effect.

Thinking maybe it was the torque sensor, bad connection, or perhaps water ingress, I opened up the motor and connectors and it was quite clean inside, just like I remembered when I did the cooling mod (pads, paste, and plate). Just some road grime on the oil gasket sprocket side, which cleaned up easily enough.

The torque sensor looked new. No cracks, no chafing wires. The inductive power/data loops looked good. They did have a gummy layer of gray grease on their interfaces, which I cleaned up and replaced with a light high temp silicone lube. At least now the bike pedals much easier without motor power! Those inductive loops were causing quite a bit of drag with whatever gummy grease was on them.

Upon reassembling the motor, not only did it still work (yay!), but the occasional PAS glitch still happens (boo!).

One last thought. This has the 860C display (which I quite like, otherwise). The glass is starting to delaminate around the edges likely due to the Arizona sunshine, and the USB out port broke apart. So I'm probably going to be in for a new display. Are the 860C's currently available for purchase capable of being re-flashed with the OSF?

Wondering if I should try re-flashing both the motor and the display? I don't currently have the USB programmers to do that, but have flashed plenty of other microcontrollers in my day job.

Any thoughts, anyone? Thanks in advance! This forum has been an invaluable resource.

The problem of losing pedal assist has already been reported by other users. It is not frequent, it only happens under certain conditions.
It happened to me twice last year and once this summer.
I think I have identified the cause and solved the problem.
It will be in the next update.

 

[Zipzap]

The problem of losing pedal assist has already been reported by other users. It is not frequent, it only happens under certain conditions.
It happened to me twice last year and once this summer.
I think I have identified the cause and solved the problem.
It will be in the next update.

Great! Thank you. I’ll make sure I’m ready to flash when you have the update ready. Let me know if you need a beta tester.

Now wondering why I never experienced this in 2,000 miles until recently, like the last couple months. Maybe I changed a setting that makes it more likely to run into… ah, middle-aged memory.

In the mean time, is there a particular exacerbating corner case that I should try to avoid?

 

[jgs1]

Hi Mbrusa
Some really cool software here , thankyou .
I have several tsdz2/b motors attached to Marin mountain bike frames
the configurability of OSF is very good but I wondered if you could add an " overrun" feature .
Already using minimum deceleration 2% ( sort of works a bit )

 

[mbrusa]

Great! Thank you. I’ll make sure I’m ready to flash when you have the update ready. Let me know if you need a beta tester.

Now wondering why I never experienced this in 2,000 miles until recently, like the last couple months. Maybe I changed a setting that makes it more likely to run into… ah, middle-aged memory.

In the mean time, is there a particular exacerbating corner case that I should try to avoid?

If my guess is correct, there is a way to avoid this problem.
When you stop pedaling, wait until the motor is at zero rpm before you start pedaling again.
How many times has this happened to you in the last two months?

 

[mbrusa]

Hi Mbrusa
Some really cool software here , thankyou .
I have several tsdz2/b motors attached to Marin mountain bike frames
the configurability of OSF is very good but I wondered if you could add an " overrun" feature .
Already using minimum deceleration 2% ( sort of works a bit )

I'm glad to hear that there is someone, besides me, who appreciates a long deceleration. I have it set to 0%.
I don't know if anyone has noticed, but I have already increased the deceleration time.
Do you think it should be even longer?

 

[Zambam]

I'm glad to hear that there is someone, besides me, who appreciates a long deceleration. I have it set to 0%.
I don't know if anyone has noticed, but I have already increased the deceleration time.
Do you think it should be even longer?

I have KT LCD3 TSDZ2 V20.1C.4-2 OSF . Just searched my manual for "deceleration" but it was not found. Is it called something else or is it a feature not available in my version?

 

[dameri]

If overrun mean that motor keep running after stop pedaling. It happens to me on every day when I'm bicycling. It usually do not start immediately. Maybe TSDZ2B get warm and then it will begin. It also includes a banging sound. It seems to me that the reason is the big gear. I think TSDZ2 had better big gear than TSDZ2B. TSDZ2B big gear is too complex and fault-prone.

 

[jgs1]

I'm glad to hear that there is someone, besides me, who appreciates a long deceleration. I have it set to 0%.
I don't know if anyone has noticed, but I have already increased the deceleration time.
Do you think it should be even longer?

is it easy to make the deceleration ( overrun ) a bit longer ? . Really useful when using the tsdz2 on mountain trails

 

[mbrusa]

is it easy to make the deceleration ( overrun ) a bit longer ? . Really useful when using the tsdz2 on mountain trails

It is true that a long deceleration time (overrun) improves the use on mountain trails, but it also has some contraindications and above all it requires greater attention when changing gear.
To increase overrun, in addition to setting "Motor deceleration" to zero, the value of PWM_DUTY_CYCLE_RAMP_DOWN_INVERSE_STEP_DEFAULT must be increased. I will try to do so.
I will also have to try to decrease the minimum overrun, there are also requests in the opposite direction.
A clarification, the deceleration time, with the same setting of "Motor deceleration", is not constant, but depends on the pwm duty cycle, the cadence and the motor rpm.

 

[mbrusa]

I have KT LCD3 TSDZ2 V20.1C.4-2 OSF . Just searched my manual for "deceleration" but it was not found. Is it called something else or is it a feature not available in my version?

I checked, the manual has not been updated.
The item 10.9 is not "Min current ADC step" but:
10.9 - "Motor deceleration adjustment".
If set to zero, the maximum deceleration ramp is active, if set to 100%, the minimum deceleration ramp is active (quicker stop).

 

[Zambam]

I checked, the manual has not been updated.
The item 10.9 is not "Min current ADC step" but:
10.9 - "Motor deceleration adjustment".
If set to zero, the maximum deceleration ramp is active, if set to 100%, the minimum deceleration ramp is active (quicker stop).

Thanks for checking. Now I will have to go play with 10.9- "motor deceleration adjustment" to see what it does in actual riding.

 

[arkanum]

Yes the common and controller folders have been deleted, all the files are in the src folder.
The wiki needs to be updated.
With the latest version it should compile and flash automatically just by clicking Compile&Flash.
However, it is necessary to install xterm, it works with Linux.
Maybe also with MacOS, but I don't know it and I don't have the possibility to try.

so, just to get i right, flashing to the motor can be done on linux using the java configurator just like on windows?

i just made my way through the manual and wiki and got quite confused by the improvements IArchi made: Compile and flash automatically from Linux and Mac OS by IArchi · Pull Request #44 · emmebrusa/TSDZ2-Smart-EBike-1

 

[mbrusa]

so, just to get i right, flashing to the motor can be done on linux using the java configurator just like on windows?

i just made my way through the manual and wiki and got quite confused by the improvements IArchi made: Compile and flash automatically from Linux and Mac OS by IArchi · Pull Request #44 · emmebrusa/TSDZ2-Smart-EBike-1

Yes exactly, in Linux the Java configurator can be used as in Windows.
I tried with Ubuntu, it is necessary to install xterm (sudo apt install xterm).

I adopted the IArchie proposal even if it did not work automatically for me.
There was also another proposal for MacOS systems
Configurator UI and error handling improvements by ec-bikes · Pull Request #92 · emmebrusa/TSDZ2-Smart-EBike-1
I did not adopt it because it did not work in Linux and it was not possible to investigate further.
However, I understood that to work automatically in Linux a terminal emulator was needed to show the compilation output, hence the choice to use xterm.