V11 cut-out report

[avevo]

A few weeks ago my V11 cut-out at 37.9 kph for no apparent reason.
No overpower events, no battery issues, no obstacles (flat road, no dirt, etc), no high temperatures, nothing.
Just a shut-down while riding.

Lucky me, only a few scratches and bruises for me and some muscle pain (collateral damages: 260 bucks sunglasses totaled, 300 bucks jacket destroyed).
V11 got a a few scratches and the 3D-printed frontlight bumper savaged.

After the fall, V11 was found turned off and won't turn on again. (This seemed really bad).

Last meaningful records from euc.world log file reported reasonably safe values:
- voltage: from 78.7 to 80.5V (reported as "89%" battery)
- current: from 6 to 19A (way below the 65A limit)
- power output: from 200 to 1500 watts (way less than 2200W rated motor output)
- speed: from 36 to 39.34 kph
- safety_margin: 48 to 52% (way above than 20% minimum recommended)
- ambient temperatures around 20°C

The last three meaningful records in the euc.world log file (0.16 seconds apart) also show reasonable values and report respectively:
- 79.8V 9.6A 772W, +0.1° tilt, 49% safety margin
- 80.5V 3.4A 275W, +0.1 tilt, 51% safety margin
- 78.9V 17.8A 1412W, +0.1° tilt, 47% safety margin.

Then something happens (between rows 12531 and 12532 of the euc.world CSV log file), and in 0.16 seconds the wheel just shut down. I remember a slight "clank!" noise, I'm not sure if it happened before or after the shutdown (neither I'm sure it has anything to do with the shutdown); upon inspection of the V11 there were no signs of anything wrong.

Log file literally shows nothing interesting except the "now is on, now it's off" thing:

Negative forward tilt values mean "accelerating" (expect -0.1 to -0.5° usual values), positive equals "decelerating". The +0.1° backwards tilt looks like a slight miscalibration (not enough to have the wheel "going backward" if left free).

Safety margin never went below 37% for the entire 12.9 km trip.

Note that 1412 watts output is within safe limits at 78.9 volts even if only one of the two batterypacks was available.

The next few records show:
- invalid voltage values (meaning the mainboard got actually shut down; just like when you turn off the wheel with euc.world still logging a few seconds of data and the voltage readings go down)
- quickly increasing tilt values (up to +70°: frontlight literally looking at the sky)
- euc.world safety_margin alarm (256, going off after the shutdown, thus based on invalid data).
Note that those log records are about 0.15 seconds apart.
No way at 11:56:15.968 it's 78.99V and drops to 44.53V at 11:56:16.129. The V11 just shut abruptly off, for some reason. My main proof is the CSV log file with those timestamped rows.

Thus I call it a cut-out (is "sudden shutdown while riding with enough headroom" better?) because a "normal" fall would have logged more records before shutting down (possibly recording a freespin as well). I also checked the place where I fell for small bumps and cracks but there was nothing big enough to have the slightest worry. Sunny day, clear road, tire/shocks pressures OK, I wasn't distracted.

At home, connecting the charger for a very few seconds was enough to turn on the V11 again (this is spectacularly bad because if it happened 24 hours before, when I was 30 km up the mountains...).

The Inmotion app diagnostics didn't report any issue before and after that fall.

In the previous days I had enjoyed a number of trips of 60-75 kms each without issues, always riding in Fancy mode (55 kph limit, but rarely went above 50, never exaggerating up to tilt-back or alarms), enduring some stress (up to 55A output, etc) always without problems.

I also rode many times in the following weeks, without problems, in Normal/offroad mode (instead of Fancy mode), including doing 40+ kph in the very same spot I fell.

So I've still got no clue as why the heck it cut out while riding. I just found myself falling towards for no apparent reason.
The wheel was activated on May 21, 2022; only had ~2700 km (~700 mine, ~2000 previous owner); latest firmware update; shocks and tire in reasonable state.
I can only think of:
- some sensor fault (hall sensor? on/off key? can't believe it could believe it's enough to shut down in less than 0.16 seconds)
- some main board firmware bug.

I switched from Fancy mode to Normal mode just to tell myself I took some countermeasure. Then I did a 32 km trip, going safely above 40+ kph and 2000+ watts output multiple times, as if the fall never happened.

I had 10,000+ kilometers on a V10F and even in the worst conditions (overheat, or 0% battery, etc) it never cut out, neither required the charger to be turned on again after a fall. I enjoyed lots of solo trips on the mountains, only limited by its limits (battery and overheating). Using the V10, I only fell either because I was tired/distracted/overconfident, or because of really unexpected obstacles (like a large stone in the middle of the road): I just found myself falling without any other hint than a small "clank!" noise (still unsure if it is really related to the cut-out).

While I'm still open to some reasonable explanation (is there anything I didn't consider till now?), remember that I expected that the V11 to show no less quality than the V10F, even in "Fancy mode". Instead I got:
- "V11 can cutout without any warning sign"
- "V11 can cutout way below critical levels"
- "V11 can shutdown and require charger connected to turn on again (what if on a mountain trip?)"

[chanman]

That's a new one for me, it really does look like it just decided to turn off mid ride, which it should be programmed to not do under any circumstances. That's the voltage drop i see for a second when I power the V11 off. I thought maybe some hardware failure but if you got it to turn on afterward after some prodding I guess not.

Best guess I have is some intermittent fault with the battery packs where they both decided to disconnect at the same time? Maybe Inmotion support can look and give a more detailed explanation.

[Rawnei]

Hi,

Sorry to hear, glad you are ok even with those material damages.

Two things are interesting in that log, first is alarm 257, from this thread:

we can determine that safety margin + speed alarm together.

And we can see voltage drastically dropping at the same time.

From this I would guess there was some kind of battery fault, safety margin and top speed are both tied to level of charge.

I think best course of action here is to talk to Inmotion, perhaps they can see more information from their black box logging.

[avevo]

2 hours ago, Rawnei said:

we can determine that safety margin + speed alarm together.

And we can see voltage drastically dropping at the same time.

I already explained above that isn't the case, so let me clarify:

• when the wheel shuts down, the bluetooth module has enough energy to emit a few status packets
• those packets, on the V11, report steadily decreasing voltages (I guess its voltmeter depends on timing some capacitor discharge; it also happens when regularly shutting down the V11)
• between row 12531 of the log file (motor still running, voltage and other parameters are OK) and row 12532 (inconsistent voltage and other parameters, telling us the wheel shut down; also note the tilt value abruptly increased to +1.4°) there are only ~160 milliseconds
• the "speed alarm" (1) generated by euc.world was there because the speed limit at the time was set to 25 kph; in the previous weeks I rode about 700 km with that setting (up to 54.3 kph and 3900+ watts output, up to 72+ kilometers on a single trip) without problems
• the "safety margin alarm" (256) in that log file only happens when the wheel is already shut down, and only happens because inconsistent record values; in the whole 12.9 km trip the minimum safety margin was 37%, way safer than the default 20% threshold for the alarm
• in the log file the speed_limit enforced by the wheel was never below 55 kph (Fancy mode and battery never below 89%); during that 12.9 km trip, max speed was 46.2 kph, with 28.1 kph average speed riding
• in the last record before shutting down, reported power was 1412 watts, that is, there was plenty of headroom even if one of the two battery modules failed
• I had the latest firmware (1.4.22 EUC, 1.3.3 BLE, 4.2.21 inverter)

I have disabled the Fancy mode (and set a speed limit of 48 kph) only to tell myself "I did domething".

I still don't have a clue about "why" and I guess Inmotion guys won't be able to tell me more (diagnostics, both battery and wheel, reported no issues soon after I turned it on again).

I hope someone could tell me why the heck after that cut-out it had to be connected to the charger for a couple seconds to switch it on again. And whether that means some sort of damage/issue.

I'm riding it again (including off-road and doing 40+ kph on the very same spot where the V11 cut out), hoping it was an event "once in a lifetime".

[Rawnei]

26 minutes ago, avevo said:

I already explained above that isn't the case, so let me clarify:

• when the wheel shuts down, the bluetooth module has enough energy to emit a few status packets
• those packets, on the V11, report steadily decreasing voltages (I guess its voltmeter depends on timing some capacitor discharge; it also happens when regularly shutting down the V11)
• between row 12531 of the log file (motor still running, voltage and other parameters are OK) and row 12532 (inconsistent voltage and other parameters, telling us the wheel shut down; also note the tilt value abruptly increased to +1.4°) there are only ~160 milliseconds
• the "speed alarm" (1) generated by euc.world was there because the speed limit at the time was set to 25 kph; in the previous weeks I rode about 700 km with that setting (up to 54.3 kph and 3900+ watts output, up to 72+ kilometers on a single trip) without problems
• the "safety margin alarm" (256) in that log file only happens when the wheel is already shut down, and only happens because inconsistent record values; in the whole 12.9 km trip the minimum safety margin was 37%, way safer than the default 20% threshold for the alarm
• in the log file the speed_limit enforced by the wheel was never below 55 kph (Fancy mode and battery never below 89%); during that 12.9 km trip, max speed was 46.2 kph, with 28.1 kph average speed riding
• in the last record before shutting down, reported power was 1412 watts, that is, there was plenty of headroom even if one of the two battery modules failed
• I had the latest firmware (1.4.22 EUC, 1.3.3 BLE, 4.2.21 inverter)

I have disabled the Fancy mode (and set a speed limit of 48 kph) only to tell myself "I did domething".

I still don't have a clue about "why" and I guess Inmotion guys won't be able to tell me more (diagnostics, both battery and wheel, reported no issues soon after I turned it on again).

I hope someone could tell me why the heck after that cut-out it had to be connected to the charger for a couple seconds to switch it on again. And whether that means some sort of damage/issue.

I'm riding it again (including off-road and doing 40+ kph on the very same spot where the V11 cut out), hoping it was an event "once in a lifetime".

Ok I understand what you're saying, regarding why plugging it in again made it work again might have been some kind of software soft-lock which was reset by plugging it in.

[chanman]

If there was a battery issue, false alarm or otherwise, that cleared up later when connected to charger, that might be logged and be able to be viewed by the support techs. Even though you can't see it now, they might be able to dig up something if you contact them.

Edited April 17, 2023 by chanman

[mrelwood]

What a sad, one in a million (or 10’000) event. The analysis you have performed seems all sound. It was indeed a cut-out, out of the blue, with no apparent reason.

On 4/16/2023 at 8:33 PM, avevo said:

- power output: from 200 to 1500 watts (way less than 2200W rated motor output)

The peak power output is something like 3500-4000W, so you are definitely at safe levels through and through.

On 4/16/2023 at 8:33 PM, avevo said:

I remember a slight "clank!" noise

This could support your theory of a hall sensor failure. The EUC motors commonly jerk with a “clank” noise if the hall sensor data is at a fault.

Please, upload the logs to Inmotion from the Inmotion app, and reach out to them at service@imscv.com . They really need to get their eyes on this so they can fix whatever caused the cut-out for future firmwares and wheel models.

 Before continuing to ride the wheel, I’d definitely triple-check the hall sensor cable connection on the mainboard.

On 4/16/2023 at 8:33 PM, avevo said:

At home, connecting the charger for a very few seconds was enough to turn on the V11 again

I’m not sure if the V11 goes into such a lock out mode after a hall sensor failure, but I can see that making sense as well.

On 4/16/2023 at 8:33 PM, avevo said:

on/off key?

The power on/off and lift-kill button functions are disabled if the wheel is traveling faster than about 4km/h, or if there’s enough torque at the motor to suggest that there’s a rider on the wheel (staying stationary on an incline for example).

On 4/16/2023 at 8:33 PM, avevo said:

- some main board firmware bug.

Plausible.

On 4/16/2023 at 8:33 PM, avevo said:

I expected that the V11 to show no less quality than the V10F

Here’s how statistics work though: One can own the most failure prone of all wheels for 10 years and have no issues, and then purchase the most trustworthy wheel only to have it fail immediately. One person’s experiences are only a tiny blob of a data point. All electronics can have issues, it’s up to chance whether you get them or not.

On 4/16/2023 at 8:33 PM, avevo said:

Instead I got:
- "V11 can cutout without any warning sign"
- "V11 can cutout way below critical levels"

These apply, have always applied, and most probably will always remain to apply to all wheels. That’s one big reason to ride geared up no matter where and how you ride.

 

 Please, reach out to Inmotion. By email, Telegram, etc. That might save others from experiencing the same fate.

[avevo]

On 4/19/2023 at 12:29 AM, mrelwood said:

This could support your theory of a hall sensor failure. The EUC motors commonly jerk with a “clank” noise if the hall sensor data is at a fault.

If this was the case, would a mainboard replacement reduce the probability of happening it again?

(I also remember how much the V13 prerelease videos insisted on featuring redundant hall sensors...)

I rode 300+ kilometers after that event and everything seems fine. (The soft-lock nonsense makes it scarier to do solo trips)

Also note that most Inmotion app 8.0 features require internet connectivity. (Also, its calibration thing still don't work, while FAQ's are still empty).

[mrelwood]

2 hours ago, avevo said:

If this was the case, would a mainboard replacement reduce the probability of happening it again?

Probably not. Hall sensor issues are usually in the cable connector, the cable itself, or the motor.

2 hours ago, avevo said:

(Also, its calibration thing still don't work

Strange, I just tested the V11 calibration last Sunday on the v8 app, and it worked perfectly.

 

[alcatraz]

Is the V11 seemingly fine after the crash?

Did one pack blow a fuse or something?

[alcatraz]

If the packs are fine then that volt reading can't be the input voltage. Why did it drop?

[avevo]

On 4/22/2023 at 1:33 AM, alcatraz said:

If the packs are fine then that volt reading can't be the input voltage. Why did it drop?

When you regularly shut off the V11, its bluetooth section takes a while before stopping sending data, so you have a very few seconds of bizarre voltage values. In fact, the best proof my V11 cut off was that ordinary 78.99V reading, followed by a weird ~44V happening 160 milliseconds later (and note that 150-200 milliseconds is the usual delay between bluetooth data packets from the V11). I guess the voltage is measured by precisely timing some capacitor discharge.

The wheel didn't turn on again until connecting the power supply. Nothing else came up; the self-diagnosis just reported everything fine. I did some 350 km riding in the following weeks with no problems whatsoever.

The most annoying "feature" was that I had to connect the charger to switch it on again. If the accident happened on a mountain trip, it would have been quite a hassle to get back home.

I can't be sure about what kind of critical component is the culprit. Maybe the hall sensor, who knows.

I wish future wheels had some sort of "black box". Maybe there's a reason why V13 had been singing "redundant hall sensors".

[mrelwood]

1 hour ago, avevo said:

I wish future wheels had some sort of "black box".

Your wheel already has! Which is exactly why you were advised to upload the logs to Inmotion so they can check out the data from the black box.

[alcatraz]

What does the board look like? Dirty?

What about software updates or calibrations? Did you do any recently?

[avevo]

On 4/27/2023 at 10:04 AM, alcatraz said:

What does the board look like? Dirty?

What about software updates or calibrations? Did you do any recently?

I did not yet inspect the main board; since then, I rode 400+ km without issues.

The log data of the last ride is apparently saved only if you use the Inmotion app and have Internet connection working (also, it is apparently limited to ride statistics). I always used euc.world instead because of its automatic log file saving (Inmotion app doesn't have that feature), thus the log of that day is not present.

[avevo]

Possible update: I had a near-overpower event, 4200+ watts output at 63.5 ampere (14% safety margin left). First time in more than 2700 km riding I went above ~50A output.

At first it looks like the V11 hit some road obstacle (speed humps), yet it sounds weird that switch from 1000-2000 watts output to an instant 4200+ and then back to 1000-2000, within a very few hundreds milliseconds, while keeping 47-48 kph speed. (I have set "2nd prio3rity speed alarm" at 48 kph, "2" implies that, "256" is the "low safety margin" alarm). I can't rule out yet the possibility of some motherboard problem. (If only the granularity of the logging was higher than "a log row every 160-240 msec"...).

[Chriull]

On 4/26/2023 at 4:11 PM, avevo said:

The most annoying "feature" was that I had to connect the charger to switch it on again. If the accident happened on a mountain trip, it would have been quite a hassle to get back home.

In the ancient times this also happened with my Ninebot One E+. Once the bms shut off due to overcurrent (?and cell undervoltage?) one had to unlock the bms by connecting it to the charger. The second way to unlock the bms was to dis- and reconnect the battery from the mainboard. That's definitely something to try once stranded somewhere in the mountains...

Single cell bms cut off was just with these "ancient" wheels, overcurrent cutoffs should only trigger with "huge" values for some time. Definitely not at ~17A for some 100ms.

But not much is known about inmotion bms and they seem to go very own ways from time to time...

With ~89% charge one cell would have to be very deteriorated to drop below some sane threshold voltage wise.

Some loose contact/micro crack in a copper track could be the culprit?

53 minutes ago, avevo said:

yet it sounds weird that switch from 1000-2000 watts output to an instant 4200+ and then back to 1000-2000, within a very few hundreds milliseconds, while keeping 47-48 kph speed.

The control loop likely regulates the motor current some hundred times a second, current change time is just limited by the systems impedance.

Reported values could be averaged or just some 'random" values from within the time intervall - so it's about impossible to deduct much from such a peak.

Inmotions internal data log, which can be send to inmotion via the app seems to have much more precise data.

[dycus]

9 hours ago, avevo said:

At first it looks like the V11 hit some road obstacle (speed humps), yet it sounds weird that switch from 1000-2000 watts output to an instant 4200+ and then back to 1000-2000, within a very few hundreds milliseconds, while keeping 47-48 kph speed.

That's pretty much exactly what you would expect from hitting even a gentle speed bump at speed. If you're cruising on flat ground, then suddenly the EUC encounters a slope (the base of a bump), the power needed is equivalent to if you were going up a slope of that angle at that speed (just only for a moment).

For example, if you hit a speed bump that had a 20 degree slope (pretty average I think) at 50kph, that's like trying to ride 50kph up a 20 degree/36% grade road! You pass the bump in less than 100ms, but the power requirement during that time is the same. 100ms is a very long time in electronics.