Metr Pro Updates

Murloc992 · September 26, 2019, 11:22am

VESCs are made by my friend, VESC6 variant, 1200+KM distance. No other UART devices connected.

This is the first cutoff a day before (I will get that log too soon):

Voltage dip is me disconnecting and reconnecting the loop key. Board was completely ignorant/dead. App was connected but no stats updated.

Then it cut out here, straight lines at the end. Metr disconnected.

The last one, same thing with straight lines at the end. Metr disconnected.

I have rechecked VESC boards, all connectors, battery, nothing seemed to be out of ordinary.

When BMS kicks it looks way different, I get undervoltage errors and DRV errors. I had this only once when I forgot that I am low on battery.

23rd day log: https://www.dropbox.com/s/j74o5ehfu95bay6/log_2019-09-23.txt?dl=0
24th day log: https://www.dropbox.com/s/7359zkjdecjbxos/log_2019-09-24.txt?dl=0

Friends IOS 26th day log: https://www.dropbox.com/s/l5bgkgv85iomg60/log_2019-09-26.txt?dl=0
He will try to access earlier logs. How can he do that?

Murloc992 · September 26, 2019, 3:59pm

My board just did the same cutoff while doing nothing, bisected on my work area. Now metr blinks every second, data updates are like 2-3 seconds between each…

rpasichnyk · September 26, 2019, 6:10pm

I still have no idea why your VESC cuts off. Curious where the problem is. Have you been running without problems on earlier VESC FW? Did the issues start after you upgraded to FW 3.61? Try lower baud rate, Metr Pro additionally supports 57600, 38400, 19200, 9600.

Murloc992 · September 26, 2019, 6:13pm

Previously I used 3.58 or 3.59(can’t recall) firmware, the one before 3.60 and 3.61.

I think I found the issue but I still need to verify it is that.

(max audio, low volume sounds here)

I’ve found a piece of metal in my receiver. It’s conductive, it doesn’t stick to solder, even with flux and leaded one… I have no friggin idea how it got into the deepest point of the receiver as the whole enclosure is spotless and dust free…

I will verify and let you know. This would be the weirdest “bug” I’ve ever seen… I have no idea how VESC reacts to shorts in the receiver.

xsynatic · September 30, 2019, 7:35am

Does that mean sonething in particular?
Because i had a little cutoff at 13:22 while going “uphill” (it was a 2m tiny and not steep at all hill.) No faults registered.

bevilacqua · September 30, 2019, 1:35pm

share the record URL, might find something in the CSV Data

xsynatic · September 30, 2019, 1:49pm

Wasn’t sure because i didn’t notice anything weird.
It was between 13:22:23 and 13:22:28 (more like 28)

bevilacqua · September 30, 2019, 2:40pm

plotted(motorC, batC, duty, voltage)
data looks clean, whatever it was, your vesc did not capture it

Ace · September 30, 2019, 5:46pm

This exact thing happened to me except my Vesc wouldn’t connect via usb to my computer as well. My metr. Saved my ass. I was able to connect via blue tooth and the Vesc tool app. I uploaded new FW to my Vesc through my cell phone. Afterwards I was able to connect to the usb and start from scratch. Fixed everything with the new FW

Soflo · September 30, 2019, 5:51pm

Im wondering if this may be causing my fsesc to be problematic. I honestly have 1 side drop out all the time. Im moving over to a wand now, might try to connect 2 dongles and run split nrf lol.

ducktaperules · September 30, 2019, 9:53pm

finaly got round to testing this.

Firstly verify issues still exist setting battery max from metr. i have a profile set up for testing these things but basically i have set battery amps to 5A max.

quick ride down the street and you can see that realtime data is showing 10A from battery.

Next to verify this behaviour in the official unity app. same battery settings in the battery panel.

another quick ride and you can see that this is also allowing 10A from the battery. Seems like this is also more than the app expects so the graphics are broken.

I guess this means the currently Unity battery amps settings are PER MOTOR not for the whole system.

I guess this also means that 95% of unity users (probably including raptor users) have there battery amps and battery regen set twice as high as it should be.

dont know what the best solution is here. your app settings are consistent with official unity settings. but unity settings are broken, resulting in many people running unsafe battery limits.

Meeep · October 1, 2019, 3:16am

I just want to say that I’ve had the same experience. Contrary to what I’ve been told as well.

rpasichnyk · October 1, 2019, 6:21am

OMG!

Can be 2 things:

UNITY allows 2x amps specified in the configuration (this is bad)
UNITY incorrectly reports battery amps in realtime data (2x)

Are you sure this is #1 and not #2?

ducktaperules · October 1, 2019, 8:04am

im 90% sure its the top one because otherwise i would expect Wh used and Wh/mile measurements to be way out, but i also have no real way to check it. I have my BMS in bypass so dont know how i can verify actual current from the battery.

anyone have a build with Unity, Metr and DieBieMS (or other bms with current monitoring) ???

EDIT: also if it was over reporting real-time data then this would be an unusually high motor current for only 5 battery amps.

ducktaperules · October 1, 2019, 8:10am

should be easy to figure this out by looking at the source code right?

rpasichnyk · October 1, 2019, 8:16am

I have such a build Will verify that!

Pedrodemio · October 1, 2019, 2:12pm

I remember Jeffrey stating several times that it was battery current for both motors, if it’s not the case this is a major slip that could, and still can end up really bad

pookybear · October 2, 2019, 4:53am

Still getting weird numbers from my rides.

Logs sent.

rpasichnyk · October 2, 2019, 7:45am

Yes, something weird happened and your VESC sent this

I will think about it, I don’t want just to filter out all bad samples, because you obviously want to see if something goes wrong. But it doesn’t make sense to include bad samples in stats and calculating average values

mackann · October 3, 2019, 4:14am

Fixed today in firmware update