I’m having a lot of intermittent issue with USB logging. I’ve tried a few different wiring configurations flowing the recommendations in the manual with no real improvements. My current setup is a ~5in stub of USB 3 extension cable with the male end cut off and terminated straight into the ecu connector. Shield of plug is verified to be connected to chassis ground. I’ve tried a few different flash drives I have around of various sizes, some are better than others but all eventually fail and enter the “Accessing USB” → “MemoryNotReady” loop.
I’m kind of disappointed this setup isn’t more robust but could be a hardware issue with my unit after all. I will still need to do more bench testing to verify.
FW 261.2
Here is a log of successful usb storage of 4 minutes, no changes in any state. The car is just sitting key-on with the engine off on the battery tender.
Not much to help with, but when you say “chassis ground,” do you mean that you actually tied the shield to chassis ground instead of a ground output of the ECU? And did you use the 4 adjacent pins of the ECU connector for the USB drive? Since the wires tend to be pretty small, did you fold them back over themselves before crimping to give the pin more to hold on to?
I’ve successfully written hour or more files using 2-3’ of wiring and a Samsung 64GB USB-A drive (not that those details should matter) using firmware 261.0 or earlier. Haven’t tried with the latest one yet, will this weekend.
EDIT: I have written 17+ minute log files using FW 261.2.
yes, sorry, ground at the ECU but that just passes directly though the ECU to the “chassis ground” as far as I can tell via continuity tests, it drains at the ECU though.
I’ve tried a few different configurations now and running out of ideas.
ya, I originally had it as you describe with the data wires going through a twisted pair shielded cable and the power going through a separate cable and that worked about the same as the short USB wire stub.
Oddly, the most successful attempt yet was running into a Raspberry Pi in gadget mode (future project in the works ???), that seemed to be logging perfectly but then I unplugged it to try a flash drive again and it’s back to the same old behavior only swapping drives back and fourth. Another strange thing is on the Pi, there aren’t any signs of disconnecting or dropped connections with the host, no automatic reduction of the connection speed or anything.
I had to move onto a different project for the day but I’ll be able to pick back up on it in the next few days and see what else I can come up with.
edit: spoke too soon, logs seemingly forever when car not running but the second I start the car, same testing memory loop. It continues even when the car is turned back off and until the ECU powers down completely and then gets a reset.
Today we tried to reproduce the issue you reported.
To simulate severe electrical interference, Marek @mkuklis_ECUMASTER disconnected the D+ line while an ignition coil was discharging close to the USB flash drive wiring. Using this setup, we were able to reproduce the infinite loop, although under conditions different from those in your case.
Regardless of the exact root cause, we have implemented a fix for the infinite loop. USB communication is now reset if the “Testing Memory” state fails.
This should address the symptom you observed: if USB communication drops, the logging system will now be able to recover and start recording a new log file instead of remaining stuck in the loop.
The underlying issue is the USB communication drop itself. To reduce the likelihood of such drops, we have also made additional improvements to the reliability of the logging process.
We now have a new beta build and will continue validating it tomorrow. We plan to share it by the end of the day tomorrow.
I appreciate you guys looking into it a little deeper! Seems to be reliable for most others so not sure what exactly is going on in my case.
This car does have passive coils so that could cause interference back to the ECU but the wiring is what it is. Besides shielding all of the coil wiring from the ECU to the coils I’m not sure what else can be done there.
Probably, I tried so many arrangements at this point. Currently running just a 2" stub of wire off the ECU connector to a female USB A terminal, no shielding on that because there isn’t a way to shield such a short stub of wire.
Key off =
everything off, except the constant power inputs to the ECU and when I’m powering the flash drive from an external source
Key on , engine off =
everything that is normally on is on
power to coils on (separate circuits)
power to injectors on (separate circuits)
no DBW on the car
IAC 4 wire stepper hums so maybe just holding a park position
Lambdas are the default settings, powered from a auxiliary circuit, ground switched by ECU
Key on, engine running =
same as with engine off except ECU enables things like the Lambdas and starts controlling the IAC stepper and all other engine functions
