WheelLog for Inmotion - Development

[palachzzz]

11 hours ago, WARPed1701D said:

Had trouble getting the app to work for my return journey today. Connects to the wheel and logging starts but nothing updates on the screen, or it only updates once every so often. It could be my phone though. Still need to check the csv to see if any data made it that far. Tried clearing cached app data but didn't help.

Yes, I found same problem day ago, and also I found related bug in code, the problem is - incorrect calculated checksum when getting basic info about wheel. A temporary solution for the problem - change the volume of the wheel speaker to another one with the standard application (checksum will change), and maybe it will work again 

I will put here fixed version some time later. Thank you for feedback)

[WARPed1701D]

13 minutes ago, palachzzz said:

Yes, I found same problem day ago, and also I found related bug in code, the problem is - incorrect calculated checksum when getting basic info about wheel. A temporary solution for the problem - change the volume of the wheel speaker to another one with the standard application (checksum will change), and maybe it will work again 

I will put here fixed version some time later. Thank you for feedback)

Thanks for the update and you continued work on this. I'll try the work around.

Even though the app appears unresponsive it was logging data to the CSV file this morning. Lat/Lon and Location Distance (all gathered from phone sensors) was changing across rows but the wheel data remained static. It is as though it managed to get some inital wheel stats but then couldn't update them. So from the interface it looks like the app has hung (even though you can change screens and the live log is scrolling) there just either isn't any data (everything 0) or it managed to grab an inital value for everything but then failed to grab further updates.

I for one will be very happy to make a donation to you if you can get this running smoothly for the V8 and will also donate to the original fund setup by JumpMaster for his maternity unit.

[palachzzz]

6 hours ago, Chriull said:

The power values shown are not valid since the wheel just sends battery voltage and motor current, but not the motor voltage! So the difference between shown and real power gets bigger the lower the speed (load) is!

Yes, but motor voltage will not be significantly different, in the system there is simply nothing to dispel even 100 watts. Inmotion V8 delivers up to 2.8 kW of electric power, with a real power of 2.2 kW, which is almost 80% efficiency on a standard battery. On the custom battery (Samsung q30 cells) it gives up to 2.7 kW of real power.

[palachzzz]

New version, 

FIX: 
- Incorrect calculation of the CRC, leading to the inability to read data from the wheel.
Add:
- Wheel alert logging.

Wheel sent some kind of alerts, that has different meaning.I tried to analyze and understand it, maybe I was mistaken with the interpretation of some of them. It looks like:

Speed cut off at speed 38.82 and current 24.60 [2102F600705FFF]

If you see there "Unknown", please send me hexvalue from square brackets. Also, I would be grateful if you could send me a piece of the log (without coordinates), before and after the event

 

[Chriull]

45 minutes ago, palachzzz said:

Yes, but motor voltage will not be significantly different,

Sorry no - it will can and is (1) significanly different (depending on speed). Read the link for more details.

Quote

in the system there is simply nothing to dispel even 100 watts. Inmotion V8 delivers up to 2.8 kW of electric power, with a real power of 2.2 kW, which is almost 80% efficiency on a standard battery. On the custom battery (Samsung q30 cells) it gives up to 2.7 kW of real power.

Thats quite the only case were the motor voltage is more or less the battery voltage - at higher speed with max load at almost 100% duty cycle...

edit: forgot (1) and has to be

[Mono]

1 hour ago, Chriull said:

i am taking output power into account - speed just happens to be a main factor of the output power.

But we got quite offtopic here - in the other topic are the details to discuss it without speculations.

Sure, just trying to help. 

[Mono]

58 minutes ago, palachzzz said:

Yes, but motor voltage will not be significantly different, in the system there is simply nothing to dispel even 100 watts.

Motor voltage is significantly modified through back EMF,^1 which is essentially the reason why the current goes to zero when speed approaches max speed. 

Given that the efficiency of the system is not always even close to 100%, in particular at low speed and max load,^2 it seems there is something in the system that can dispel some energy at least for some short period of time, for example by heating up the motor. 

^1 https://en.wikipedia.org/wiki/Counter-electromotive_force

^2 

 

[WARPed1701D]

50 minutes ago, palachzzz said:

New version, 

FIX: 
- Incorrect calculation of the CRC, leading to the inability to read data from the wheel.
Add:
- Wheel alert logging.

Wheel sent some kind of alerts, that has different meaning.I tried to analyze and understand it, maybe I was mistaken with the interpretation of some of them. It looks like:

Speed cut off at speed 38.82 and current 24.60 [2102F600705FFF]

If you see there "Unknown", please send me hexvalue from square brackets. Also, I would be grateful if you could send me a piece of the log (without coordinates), before and after the event

WheelLog-inmotion_09082017.apk

I grabbed the new version. The problem I am seeing at the moment is the data refresh rate which stops for several seconds and then starts/stops again. I've attached the log file (it got converted to excel when I removed my positional data, let me know if you can't open it) which reflects the updates seen in the GUI. The clock in the top left keeps ticking but ride time (as one example but this affects all metrics) sits frozen for many seconds and then will update. Sometimes it will update several times and then stop again for a random amount of time. To be clear my wheel was on its stand when I recorded this so no chang ein speed or power is expected but you can see that roll change a little between polls showing there was a flow of data.

I didn't see this sporadic updating when I used the app flawlessly yesterday morning so I doubt it is my phone even though it is older.

2017_08_09_12_40_19.xlsx

[palachzzz]

On 8/9/2017 at 7:53 PM, WARPed1701D said:

I grabbed the new version. The problem I am seeing at the moment is the data refresh rate which stops for several seconds and then starts/stops again. I've attached the log file (it got converted to excel when I removed my positional data, let me know if you can't open it) which reflects the updates seen in the GUI. The clock in the top left keeps ticking but ride time (as one example but this affects all metrics) sits frozen for many seconds and then will update. Sometimes it will update several times and then stop again for a random amount of time.

I didn't see this when I used the app flawlessly yesterday morning so I doubt it is my phone even though it is older.

2017_08_09_12_40_19.xlsx

Fix: 
- Main screen freezing

 

[WARPed1701D]

4 minutes ago, palachzzz said:

Fix: 
- Main screen freezing

WheelLog-inmotion_09082017_2.apk

Downloading now. This IS service I tell you!

[Chriull]

57 minutes ago, Mono said:

Motor voltage is significantly modified through back EMF,^1 which is essentially the reason why the current goes to zero when speed approaches max speed. 

Given that the efficiency of the system is not always even close to 100%, in particular at low speed and max load,^2 it seems there is something in the system that can dispel some energy at least for some short period of time, for example by heating up the motor. 

^1 https://en.wikipedia.org/wiki/Counter-electromotive_force

^2 

 

Heating up the motor by voltage drop along the coils (their ohmic resistance) is one way that has to happen to get the back emf nearer to battery voltage, the other is the mosfet half bridges together with the pwming and the coil inductances working as dc/dc (1) converter.

edit (1) its a dc/ac converter - but the main thing is it is a step down converter to reduce the voltage as needed.

[WARPed1701D]

Getting there. In 4 attempts I got a constant flow of data 50% of the time. The remainder was the same problem of sporadic GUI and log updates but this is more stable than it was.

2017_08_09_13_26_11.xlsx

2017_08_09_13_27_27.xlsx

2017_08_09_13_28_24.xlsx

2017_08_09_13_28_56.xlsx

[palachzzz]

Tested new feature: unlock max speed to 35 km/h 

 

[Mono]

31 minutes ago, palachzzz said:

Tested new feature: unlock max speed to 35 km/h 

Aren't you asking for people to get seriously hurt or die by implementing this feature? Just think twice.

FWIW, if this feature remains implemented, I will consider the app to be unsafe and will not be able to recommend it to anyone to use.

See also 

 

 

[palachzzz]

19 minutes ago, Mono said:

Aren't you asking for people to get seriously hurt or die by implementing this feature? Just think twice.

FWIW, if this feature remains implemented, I will consider the app to be unsafe and will not be able to recommend it to anyone to use.

See also 

As you can see, this function is not yet available in the versions I upload. First I tested by myself. If it will be released, I'm not forcing to use this function, nor the application as a whole. And if someone unlocks the maximum speed - it's his choice, not mine, and not yours.
But anyway, V8 not stupid, it has adaptive system that decrease maximum speed depending on battery condition. At battery charge at 96% the max speed was decreased to 33.1 km/h.

Also, you can see this topic:

 

[palachzzz]

26 minutes ago, Mono said:

See also 

 

"but the battery of the device was around 50%. I pressed the device really hard to accelerate uphill."

For me it's enough to understand that the rider does not understand anything about how the wheel works, and what it can do and what not.

[Mono]

5 hours ago, palachzzz said:

And if someone unlocks the maximum speed - it's his choice, not mine, and not yours.

To be clear, we set the maximum speed of a self-balancing device to a larger value than specified by the manufacturer.

I would certainly feel some responsibility if I would recommend an app that can do this, even if I didn't make the final choice to actually do it.

 

5 hours ago, palachzzz said:

"but the battery of the device was around 50%. I pressed the device really hard to accelerate uphill."

For me it's enough to understand that the rider does not understand anything about how the wheel works, and what it can do and what not.

Let me agree for the sake of argument. Now I see, his ignorance is a perfectly fine reason to let him kill himself.

[Mono]

On 10/08/2017 at 0:59 PM, palachzzz said:

Tested new feature: unlock max speed to 35 km/h 

 

FTR, the speed at which forward leaning becomes impossible (i.e. leads to outleaning the wheel) because the V8 cannot accelerate anymore is probably less than 40km/h and drops on a 5% incline for an 85kg rider to about 30km/h, as documented here:

or for a KS16C which also has an 800W motor here:

 

[palachzzz]

On 11.08.2017 at 5:23 PM, Mono said:

FTR, the speed at which forward leaning becomes impossible (i.e. leads to outleaning the wheel) because the V8 cannot accelerate anymore is probably less than 40km/h and drops on a 5% incline for an 85kg rider to about 30km/h, as documented here:

 

Nothing but theory, based on wrong characteristics. Read my post there. 

[WARPed1701D]

I've had several successful logs since I last posted. Occasionally data seems to stop briefly but for the most part my ride produces a good stream of data. Still testing. Picked up a Pebble watch today and going to test the integration tomorrow.

[palachzzz]

New version, 

FIX: 
- minor and major bugs
- decrease polling rate to 8 times per second (125 ms) to improve logging stability

Added:
- inmotion control (Light, Led, Button, Speed, Volume)

USE THE APPLICATION AT YOUR OWN RISK.
Neither the application developers nor the Inmotion company is responsible for the damage caused by the unlocking speed above 30 km/h, it is not a declared feature.

 

[WARPed1701D]

26 minutes ago, palachzzz said:

New version, 

FIX: 
- minor and major bugs
- decrease polling rate to 8 times per second (125 ms) to improve logging stability

Added:
- inmotion control (Light, Led, Button, Speed, Volume)

USE THE APPLICATION AT YOUR OWN RISK.
Neither the application developers nor the Inmotion company is responsible for the damage caused by the unlocking speed above 30 km/h, it is not a declared feature.

WheelLog-inmotion_14082017.apk

Downloading now.

It will be cool if this fixes the data stream stability. Thanks for adding configuration options too. If I can get away from the Inmotion app I will be very happy. I'll test the additional functions, including the speed control, but won't be taking it over the manufacturer wheel rated speed spec. I like to ensure plenty of headroom for the wheel to compensate for challenging conditions.

Thanks again for your continued efforts.

[WARPed1701D]

I haven't looked at the logs properly yet but I'm seeing trip distance is actually showing total distance on the GUI and in the first few log lines. I think this was fine before.

Full disclosure. This morning was the first time I used a Pebble watch with this version of WheelLog. Not sure if that could have caused this data error. More later.

[WARPed1701D]

1 hour ago, WARPed1701D said:

I haven't looked at the logs properly yet but I'm seeing trip distance is actually showing total distance on the GUI and in the first few log lines. I think this was fine before.

Full disclosure. This morning was the first time I used a Pebble watch with this version of WheelLog. Not sure if that could have caused this data error. More later.

OK this may be related to Pebble Integration as my ride home last night with the new version did not show the wrong distance, just my ride to work this morning. The difference being I used the Pebble this morning. I will test again tonight.

[palachzzz]

Strange, I will check, but it will be difficult without Pebble)