USB comms

FreeEMS topics that aren't specific to hardware development or firmware development.
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: USB comms

Post by Fred »

I wouldn't mind hacking the same behaviour into my problem vehicle and seeing if it happens, or not.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
tmbryhn
DIP8 - Involved
Posts: 15
Joined: Mon Jan 06, 2014 8:38 am
Location: Norway

Re: USB comms

Post by tmbryhn »

Got some more updates on the whole USB-coms issues subject.

I recently got a new engine built for my good old thrustworthy Beetle. Check it out! Norwegian forum, but the pics from my machining sessions are quite nice :)
http://www.vwnorge.no/index.php/topic,69232.0.html

Firing up the above mentioned engine and also another I finished recently running passive coil-on-plugs in the engine test stand, both gave huge USB issues! I could not stay connected for more than a couple of seconds before it said good bye while the engine were actually running. This is with the stand running on its own battery power, and the laptop running with its charger plugged into mains. Although coms is stable while engine is not running.
Same problem in the car along the road; running with a relatively expensive 230AC-12VDC inverter screws up USB immediately after connection of the power adapter to the laptop - even when the engine is not running, while running USB with the laptop disconnected from the inverter using its battery as the energy source works like a charm.
Even though I have also updated the previous design found on page 1 in this thread to a new one utilizing isolation between the MCU and the FT232RL using a proven ADUM chip, it does not seem to make a difference. In my mind, that is not surprising; it's the FT232RL that gets upset and shuts down/locks up resulting in lost USB coms, while the engine keeps on running with the main MCU (microcontroller) crunching through its firmware without interruption. Maybe I should have the 5V fed to the FT232RL internally in the ECU instead of through the USB cable?
I have good filtering with both inductors and tantalum/ceramic caps on the USB-VCC line, but might still be to noisy?

I was forced to go all out bluetooth comms while running the engines in the engine test stand, and that worked extremely stable with no problems at all.

So, this seems to happen when there is a charger connected to the laptop via either mains or a noisy inverter.
Should I connect the GND of the 230V output of the inverter to chassis GND in the vehicle/battery?

Tomorrow, I'm gonna remove my ECU from the Beetle, and dig out the test board with stim, coils and injectors once again, and try with different power sources; battery, inverter doing USB with my tuning laptop and also my desktop computer in the electronics workshop and see what I discover from it.

I'll keep you guys posted with my findings :)
User avatar
Fred
Moderator
Posts: 15431
Joined: Tue Jan 15, 2008 2:31 pm
Location: Home sweet home!
Contact:

Re: USB comms

Post by Fred »

tmbryhn wrote:I'll keep you guys posted with my findings :)
Keep em coming! :-)

Love air cooled krauts! Thanks for the link. I might run it through a translator later.
DIYEFI.org - where Open Source means Open Source, and Free means Freedom
FreeEMS.org - the open source engine management system
FreeEMS dev diary and its comments thread and my turbo truck!
n00bs, do NOT PM or email tech questions! Use the forum!
The ever growing list of FreeEMS success stories!
tmbryhn
DIP8 - Involved
Posts: 15
Joined: Mon Jan 06, 2014 8:38 am
Location: Norway

Re: USB comms

Post by tmbryhn »

Alright, so I got some testing done...
Could not even get my inverter running on my old weak POS 12V car battery that I'm using from time to time in the electronics workshop. Scoped the +/- terminals while loading the inverter with my laptop charger, and it went down from 12 to like 7 volts and took bout 200ms to recover, so the inverter would just go in and out like crazy...

So, I guess the problem with the inverter noise while being used in the car, had to do with lack of capacitance on the 12V input terminals inside the inverter (a 1000uF 25V electrolyte was installed originally).
The inverter is connected to the cigarette lighter underneath the dash in my VW Beetle, and the battery is located underneath the backseat, so pretty long positive wire and multiple connections points in between. Right, so because of large inductanse on those long wires, why not just take one of my huge ass 6800uF 100V Nichicons and solder directly across the input terminals inside the inverter case? :geek:

While I was inside, I bit the bullet and connected a wire between the negative input (chassis GND) and the GND on the 230 "mains" AC output.
You know, just for the heck of it :lol:

Hooked the shit up in the Beetle and went for a ride. Wow! All the noise problems with regards to the inverter are completely gone!
I had like two falling out on the USB link while driving for 10 minutes because of noise from the ignition system, so that's still an issue of course, but then again, a completely different subject.
That being said; the ECU is mounted inside the engine compartment like 20-30cm away from the old fashioned distributor having the USB cable hangin' in the breeze from the ECU, alongside the car and in through the door :lol2:

My experience is that when the ECU gets mounted inside the car behind the torpedo wall, interference from the ignition system is no longer an issue on the USB link. That would be the next step in other words.

Still gonna find out why the USB link drops out when the engine is being run in the test stand like seen in the video attached. Coms is via Bluetooth because USB keeps disconnecting like every ten seconds or so when running the engine like this:
https://www.youtube.com/watch?v=BBI3_qtuoyU

I wonder if the reason that I get so much USb noise disconnect problems is because of the the lack of EMI shielding in terms of having a metal chassis built around the engine that absorbs a lot of the shit, and the fact that the laptop is isolated from the battery GND in the test stand. Because when I have done tests in the electronics workshop with coils/injectors firing like crazy, I don't get NEARLY as much noise problems on the USB when powering my "engine simulator" from my bench PSU. And when powering the "engine simulator" through my PSU, the laptop charger is eventually connected to the same mains GND as the PSU afaik.

Anyway, I'm glad i got to improve my crappy inverter
Kenmogg60
TO220 - Visibile
Posts: 4
Joined: Tue Nov 16, 2021 7:19 am

Re: USB comms

Post by Kenmogg60 »

Following...
EFIRick
TO220 - Visibile
Posts: 9
Joined: Sat Apr 11, 2015 10:26 pm

Re: USB comms

Post by EFIRick »

There have been instances of ground loops and other badness if the laptop is powered (not running on battery).
Post Reply