UART issue on high-voltage VESCs

The boards I noticed it on were 12s but running D75s

1 Like

Ah, great. So we don’t even have a simple definition. :frowning_face:

Thanks for the info! This is helpful.

Troubleshoot:

Start at 12S,
Send continuous packages over UART,
Probe comm lines during throttling/load and see how that affects the SNR on the UART.

Repeat for higher cell count untill problem appears.

Oh now that I read this, I experienced and have the same exact issue on little focer v3 20s (2x connected via canbus), i noticed the red thing happening on the top also but was solved by changing the vesc ID of one controller. I got two identical vesc ID when I got the issue.

1 Like

I’ve never tried the DAVEGA, but the D75 with Bluetooth module works well in 16s.

I did some testing today with a selfmade shielded cable (which was not touching the Davega Display housing) The shielded cable is the original Davega cable with a braided shield wire which is not connected to GND.

When I was riding I just used a bit of current, not super high values.
The issue video shows how much EMI is affecting the UART signal. When I let roll out the kart without applying current, the „V“ was green as it should be. As video quality is not so good and there is a reflection on the display, the red „V“ appears often and the speed is slowly changing on the display.

I was able to replicate it with a different unshielded cable. I will use a different cable later, which will look better.

I have no idea why this problem did not occur on Vesc 100/250 with an unshielded cable.


2 Likes

This is great news. Thanks a lot for the info @Oliver. So the solution should be as simple as using a copper sleeve on the DAVEGA cable.

It’s still a mystery why the interference issue is only reproducible on 75/300 and not on 100/200. @Trampa, any ideas?

Could the VESC setup be a factor? A different FOC switching frequency maybe?

2 Likes

I was unable to get a Metr Pro unit to work with my Little FOCer 3.1 without getting shut off when current ran through the controller. DIY Onewheel, 19s battery. I am able to see the LEDs on the Metr through the controller enclosure. Just balancing in place caused the bluetooth to disconnect, even the slightest acceleration would cause the Metr to shut off entirely until the controller was rebooted. I replaced it with a Metr Pro CAN and it’s working fine now.

2 Likes

@janpom So i now have also a weird problem with DavegaX and Ennoid Vesc. As soon the DavegaX get contact with the basplate DavegaX has connection problems when I use the throttle. When I keep it in my hand or isolated it from the base plate it work normal. I checked also different cable and other DavegaX and also with 12s and 16s. So it’s for sure the combination with ennoid Vesc but it is still weird.

1 Like

I had a similar issue but I’m not running anything wild - 12s8p on a stormcore 60D. Tried lots of different things but it kept coming back to the baseplate. I ended up printing my own 2-piece PLA baseplate (I’ll make one from PETG when I’m not feeling lazy) and since then I haven’t had any problems. Jan mentioned the carbon deck could possibly cause some electromagnetic interference.



I bet it has something to do with y’all’s carbon fiber decks

5 Likes

Ah ok ! Yes this could be the issue. First time I use a carbon deck. Thanks!

1 Like

Yeah that was kind of the general conclusion after troubleshooting with @janpom

@philo not sure if you’re in US but if you want to try one of my weird baseplates let me know

2 Likes

Hi, for now I solved by using the ruber gasket and plastic screws. This seems to work. I use an angled basplate. So iI will just print me one but thanks for you offer and help :smiley: This issue really drove me crazy…

1 Like

I have a hypothesis about this. On a CF deck, there may be an electric connection between the motor cans and the DAVEGA baseplate. The current may travel: motor can → motor mounts → trucks → deck → DAVEGA baseplate.

Also, raw aluminum parts may be more susceptible to this than anodized parts since anodizing makes aluminum nonconductive.

4 Likes

Ok this make sense. The board with the issues is a kaly xl deck and all parts are raw Aluminium from trucks to motor gears and DavegaX. I have some isolating spray so when I have time I will coat the baseplate with it and test if this will work.

1 Like

I haven’t dived into all the details in this thread but if any conductive material is not grounded when around EMI/RFI or intentional radio signals (BLE, etc.) it will become an antenna, sucking it all up, and then re-radiating that stuff into places you might not want it.

This is commonly seen with ungrounded heat sinks being used for MOSFETs in motor controllers and power supplies.

IMHO, if there is any CF then connect it to the battery pack’s NEG terminal, shunting any absorbed EMI/RFI/ration frequency energy to ground. These emissions can come from badly designed ESC’s and be spread via interconnecting cables or radiated directly.

Grounding other metal parts might help too. This would involve maintaining electrical contact for all the pieces though so one connection grounds what is now the “frame”. This is similar to older cars/trucks which used the chassis as the common ground for all the electronics.

Anodizing, paint, lubricants, dirt, etc., between touching surfaces can interfere with this though.

Shielding cables can work but only if the interference is radiated and not conducted down the wires. Grounding the shielding needs to be done well too otherwise it will just turn into another antenna, directly injecting interference into the wires it surrounds.

2 Likes

I dont have issues with my davega on a kaly deck :person_shrugging:

This is extremely dangerous. It will massively increase your risk of a battery fire.

Any interference is worth dealing with in other ways.

1 Like

Do you also have raw Aluminium trucks and davegaX?