L1 vs L2 charging SOC algorithms seem different

[sukhoi_584th]

Is this well known? This summer I've L1 charged up using the Ford charger about 50-60% SOC for four times at a family member's house. Normally I L2 charge at home, with some occasional DCFC tossed in. I've noticed after L1 charging to ~90% SOC over a couple days that once I start driving the SOC will quickly drop ~4% over just a mile or two then proceed like normal. L2/DCFC charging doesn't do this. It seems like L1 algorithm is different from the L2/DCFC algorithms, and has a disconnect with the discharge SOC algorithm.

Usually I start driving immediately or shortly after finishing L1 charging, so I'm not sure how a longer post-charge dwell would affect this. It's odd too because you'd expect L1 to have the least SOC relaxation due to the lowest power input. Thoughts on if there's a mechanical/electrical explanation for this, or is it just poor quality software?

Sponsored

 

[Mach-Lee]

I think it’s mostly to do with temperature differences and the time since charging finished. But it very well could have something to do with L1/L2 algorithm differences. I never use Level 1 so I haven't studied the differences. Charge termination seems based on voltage targets, after that is reached then it shuts down and the pack voltage and SoC does drift down slightly as the cell voltages relax. You're correct that in theory, the voltage/SoC sag with L1 should be less than L2 due to the lower amperage input.

If you pay attention to the actual SoC vs. the displayed SoC, there is some hysteresis built into the displayed SoC after charging. For example, if you have it set to stop at 80% and the actual SoC is 77% when it finishes, when you come back in 5 hours to it you might see the actual SoC drop to 74% but the display will still show 80% until you start driving, then it will gradually drop to "catch up" to the correct value. I'm sure they did this so the user doesn't wonder why their car didn't charge fully when they turn it on in the morning. The propensity to do this is also tied to the HVBJB and powertrain software version the car has. I think in later software they upped the charge termination targets a little more so there is less SoC sag after charge termination.

Having a calibrated BMS and the pack in consistent warm temp conditions also helps reduce the drop. I don't notice the drop as much in the summer, but in the winter it can get worse when the pack cools off a lot after charging.

In your case, I suppose it's possible they are applying the hysteresis for Level 2 but not Level 1 charging. Or Level 1 uses lower voltage targets for charge termination. It might be a case where they forgot to update the Level 1 strategy to match the Level 2 strategy in terms of termination point and hysteresis. You could try a powertrain software update and see if it changes.

 

[sukhoi_584th]

I'd have to pay much more attention to be sure, but temperature never jumped out at me as a difference. The L1 sessions have all been 50-75 °F, but the L2 have been all over the place from nearly freezing to scorching hot.

Good call on the hysteresis...I suspect differences with handling it are what I'm seeing. I did calibrate my BMS in February via your suggested method, but then had the HVBJB recall done a couple months later. So, I'm not sure if that managed to kill the calibration or not.

I don't think it's termination voltage. My L1 sessions have all ended 85-95% SOC, while I've L2ed to 100% SOC probably a dozen times and it stays at 99%-100% for even longer than I'd expect it to.

Anyway, not a big deal just something interesting I've noticed. I'm not exactly sure what software versions I have on which modules, but I believe I'm current on OTAs.

Sponsored