EUC World Alarms (or other App Alarms) for Sherman

[Rawnei]

I'm curious as to what alarms you guys have set for your Shermans in EUC World (or other apps?), both speed and temperature alarms would be very interesting to know!

Is there some specific battery level and speed that you are mindful of, say specific speed under X% battery? I see only discussions about top speed before the wheel beeps at high battery but nothing about low battery behavior.

What's a reasonable temperature to be aware of? I saw on Martys second hill climb it hit 76C, maybe 75 is a good threshold for temperature alarm?

Let me know what you use and what you think. 😎

[Rawnei]

Still no replies, that makes me sad 😥😅

The main thing I would like to avoid having the motherboard burn on steep and long climbs if it's possible to alert on some parameters before that happens, @Seba you have any input on good alerts for a Sherman? And what is the difference between "peak current" and "current" from an alarm perspective?

It's a shame we don't have log data from someone who managed to burn their Sherman to see if something stands out. 😔

Looking at previous tours it's difficult to make out good current number to watch for to not get false positives, I commonly see peaks in the range of 110-130A (even as high as 170A!) during normal operations, analyzing logs it seems to happen in usual scenarios, when going downhill on a normal incline, starting from a standstill or climbing a not so steep hill at low to medium speeds (10-40kmh) so using peak current for alarm seems not very good, what does "current" alarm represent, some average over x amount of time?

Also @Seba I noticed a strange thing happening multiple times now, when you park your wheel for a while (5-10 minutes) EUC World will sometimes produce some bogus data and trigger an alert out of nowhere, happened to me 3 or 4 times now startling myself and people around me since I'm using a speaker (which I now started to turn off whenever I have a pause), when it happens it seems it loses GPS data for half a second and the log looks like this:

speed   speed_max   speed_limit   voltage   current    power       battery    temp    alarm    gps_datetime    
0            65.1               280               98.21       0.5          49             90            46        0           2021-05-13T15:46:30.999+0200
0            65.1               280               98.21       0.5          49             90            46        0           2021-05-13T15:46:30.999+0200
0            65.1               280               98.09       0.3         29             89            46        0
0            65.1               280               98.09       0.3         29             89            46        0
0            65.1               280               98.18        0.4         39             89            46        0
0            65.1               280               98.18        0.4         39             89            46        0
0            65.1               280               98.15        0.5         49             89            46        0
0            65.1               280               98.12        0.4         39             89            46        0
0            65.1               280               98.09       0.4         39             89            46        0
0            65.1               280               98.09        0.3        29             89            46        0
0            65.1               280               98.12         0.5        49             89            46        0
0            65.1               280               98.12         0.5        49             89            46        0
2358.4  2358.4          0                  -91.25        2153.9  -196543    0              0          2
0            2358.4          280               98.12         0.7        69             89            46        0           2021-05-13T15:46:36.999+0200

Never had the same thing happen on my MSP only with the Sherman.

Another thing I noticed is that even if I have overvoltage alarm disabled I get overvoltage alarm very early in the tour at full battery, looking at the log indeed I can see it briefly peak at 102V but since overvoltage alarm is disabled it shouldn't trigger, seems like a bug?

Edited May 18, 2021 by Rawnei

[Seba]

1 hour ago, Rawnei said:

@Seba you have any input on good alerts for a Sherman? And what is the difference between "peak current" and "current" from an alarm perspective?

No, I don't have such data yet  The problem for me is to get Sherman for extensive testing. As for now, I can't afford buying Sherman for EUC World development. Anyway, I hope to get Sherman from Eunicycles.eu or from EUCService.com soon. Peak current alarm will react to momentary current values, while current alarm uses some filtering so it will activate when sustained current will exceed the alarm threshold.

1 hour ago, Rawnei said:

Looking at previous tours it's difficult to make out good current number to watch for to not get false positives, I commonly see peaks in the range of 110-130A (even as high as 170A!) during normal operations

This is because of Veteran legacy, Sherman shares the same mainboard idea with Gotway/Begide. Especially, it doesn't report battery current but motor phase current. But starting from upcoming 2.6 version, EUC World will estimate battery current in Veteran/Begode/Gotway wheels so it will be more "real".

1 hour ago, Rawnei said:

Also @Seba I noticed a strange thing happening multiple times now, when you park your wheel for a while (5-10 minutes) EUC World will sometimes produce some bogus data and trigger an alert out of nowhere, happened to me 3 or 4 times now startling myself and people around me since I'm using a speaker (which I now started to turn off whenever I have a pause), when it happens it seems it loses GPS data for half a second and the log looks like this:

This is a glitch in the data transmission caused probably by the bug in wheel firmware. It sometimes happens with certain Veteran, Begode and Gotway wheels. Starting from 2.6 version of EUC World there will be a special filter that will discard invalid data.

[Rawnei]

22 minutes ago, Seba said:

This is a glitch in the data transmission caused probably by the bug in wheel firmware. It sometimes happens with certain Veteran, Begode and Gotway wheels. Starting from 2.6 version of EUC World there will be a special filter that will discard invalid data.

Thanks for the replies, what about the overvoltage alarm being triggered altough being disabled? Also fixed in 2.6? 😉

[Chriull]

2 hours ago, Rawnei said:

It's a shame we don't have log data from someone who managed to burn their Sherman to see if something stands out. 😔

Looking at previous tours it's difficult to make out good current number to watch for to not get false positives, I commonly see peaks in the range of 110-130A (even as high as 170A!) during normal operations, analyzing logs it seems to happen in usual scenarios, when going downhill on a normal incline, starting from a standstill or climbing a not so steep hill at low to medium speeds (10-40kmh) so using peak current for alarm seems not very good, what does "current" alarm represent, some average over x amount of time?

As @Seba stated Sherman reports motor current - which is great for your alarm purpose.

The burdening of the wires and mosfets is proportional to the square of this motor current. Afair @Marty Backe started to use an alarm on this motor current  with his GW wheels to protect the motor wires and mosfets from frying. Afair it was mostly in the range of some 80-100A?

As Marty rides a Sherman too maybe you find some sane values in his youtube videos?

2 hours ago, Rawnei said:

(even as high as 170A!)

Here one has to filter - some peaks are no problem.

The effective heating goes with the power dissipation over time - for this the root mean square of some recent time intervall of the current would be of interest!

Until now riders afaik accomplish this by accepting (ignoring) motor current alarm beeps coming "seldomly" during short time periods (show off accelerations, inclines) and reduce the burden immedeately once this beeps come regularly over some seconds?

1 hour ago, Seba said:

This is because of Veteran legacy, Sherman shares the same mainboard idea with Gotway/Begide. Especially, it doesn't report battery current but motor phase current. But starting from upcoming 2.6 version, EUC World will estimate battery current in Veteran/Begode/Gotway wheels so it will be more "real".

As written above - for mosfet/wire burdening the motor current is the value of interest! Battery current is "just" for the power values - which are not really corresponding to "motherboard" burdening.

The root mean square (1) of the last x seconds (configurable?) would be the absolute perfect value! Mabye once you want to play with this

Edited May 18, 2021 by Chriull

[Rawnei]

2 hours ago, Seba said:

Peak current alarm will react to momentary current values, while current alarm uses some filtering so it will activate when sustained current will exceed the alarm threshold.

The values we see in the tour and log is "peak current" right? Is it possible to record in log and show "current" calculation as well as a graph? That way we can distinguish some sort of baseline for current alarms.

[Rawnei]

Since nobody is sharing here are my EUC World settings from experimenting around:

Speed and distance correction at 6% to get more accurate readouts

Current Alarm: 100A, This alarm is some sort of short sustained average as stated by @Seba above. I asked @Marty Backe who used 90A alarm when doing the second hill-test and there it beeped now and then, I did a test at 95A because I don't want false positives but I got it to trigger on hard accelerations on flats so I raised it by 5 more too 100A which for now seems OK for me, still experimenting and testing this setting.

Peak Current: Off, not interesting.

Two speed alarms:

Speed Alarm Highest Prio: 54kmh at or below 30% battery, I'm not sure what the low battery speed limits of a Sherman is so just set something for testing and will be mindful of Sherman beeps at low battery, 54kmh is equal to about 57.5kmh without speed correction.

Speed Alarm Medium Prio: 68kmh at or below 100% battery, this is equal to about 72kmh without speed correction

Temperature Alarm: 75C

Edited May 20, 2021 by Rawnei

[Dominic Winsor]

On 5/20/2021 at 7:48 AM, Rawnei said:

Speed and distance correction at 6% to get more accurate readouts

Hi @Rawnei this is an interesting thread. What was the error you are applying a correction to? Was it reporting too fast or too slow, and against what benchmark?

I haven't set any current alarms in my Sherman yet, but I will give this a go now as I think it quite sensible.

[Rawnei]

4 hours ago, Dominic Winsor said:

Hi @Rawnei this is an interesting thread. What was the error you are applying a correction to? Was it reporting too fast or too slow, and against what benchmark?

I haven't set any current alarms in my Sherman yet, but I will give this a go now as I think it quite sensible.

It over-reports speed i.e. the wheel is showing more speed than you are really going, you can compare it to GPS speed which is also recorded in EUC World, analyzing several tours it seemed to me that 6% gives a more accurate representation of actual speed. @Seba can probably look at statistics over a lot of tours to get an accurate number like he did for the V11.

For now 100A current alarm feels good for me, I trigger it now and then when I push the wheel but one or two triggers is no panic, if it triggers multiple times it's a good indicator that the wheel is under heavy stress. Anything below 100A would trigger too often for me.

[Rawnei]

An update for anyone basing alarms on this thread, since updates to EUC World was done that affects the current reported a current alarm at 55-60A is more reasonable (not phase current).

Safety Margin I found 14% to be most accurate and close to wheel beeps but sometimes even that a bit conservative.