Gen 1 Cruze: Importance of battery voltage during flash

[Dutch]

This is in the L4 tuning forum as it relates to tuning the first gen Cruze and Sonic.

Complaint is that a tuned 2014 Cruze eco suddenly "looses all power during a downshift or under heavy partial throttle, almost like severe turbo lag, then suddenly gets power but sounds weird" after flashing a very minor change to the main calibration. Reverting the change does not solve the issue, nor does reverting previous changes. The only thing that seems to resolve it is reverting to the factory calibration. Customer also mentioned writes occasionally fail, and sometimes the 'release parking brake' warning would blare for no reason until the battery is disconnected/reconnected.

Test drive with logs confirm the complaint, error seems to be camshafts are, for whatever reason, commanded full advanced before abruptly returning to their normal positions. The VVT tables were never touched. The engine is mechanically healthy and commanded phaser angle matches actual even when manually controlled. Reverting to the factory calibration restores normal operation.

The issue ended up being a marginal battery and failure to use a battery tender to keep battery voltage up during the flash process.
After a fresh battery was installed, I was no longer able to replicate the issue.

Voltage was apparently high enough for the write to succeed but clearly something went wrong during the write.
Possibly contributing to the issue is the Eco's tendency to reduce alternator output, preventing the weak battery from staying fully charged even though the vehicle was driven for a decent amount of time.

Although using a battery tender is always best practice, we tend to get complacent with these ECMs and assume if it says "Write completed" with no check engine lights, everything must be good. This example highlights the importance of
- maintaining battery voltage during writes
- having a drive cycle at operating temperature to confirm vehicle behaves as expected
- adopting a white glove approach whenever anything aftermarket is connected to the diagnostic port of a car

I am creating this thread for tuners to share their experiences dealing with weird issues that ended up being due to voltage dropping during a write, and potentially ask those with experience reverse engineering these modules as to what could be the underlying mechanism of the fault (tables being partially written?)

Additionally, could HPTuners implement something in software to detect these potential issues? Perhaps verifying what was written, or giving a warning if voltage to the device drops below a threshold?

Any comments are appreciated.

[blindsquirrel]

If the dealer tools don't stop you from doing something stupid and/or not following the directions why should an aftermarket tool? The service manuals you're supposed to be familiar with before flashing anything with any type of tool are explicit and unambiguous about the importance of system voltage during programming.

[Dutch]

1. This was escalated to me.

2. Given the number of blatantly dangerous calibrations and misinformation I seen in the repositories (before they locked it down), various posts, and customers I'm sure you've seen over the years, most don't follow best practices.

3. This information is for others who run into similar problems, making them aware before they start blindly replacing parts, and perhaps act as a warning for those who aren't so careful.

If you have nothing to add, its best you don't say anything at all.