[RobbertPatrison]

I installed a Sense power monitor in the electrical panel of my house. It's a device that measures the current and voltage real time, uploading it to the cloud via wifi and showing it on a neat app (see www.sense.com). After a while Sense learns devices and is able to identify what is running in the house. The machine learning part doesn't work well in practice, but I can easily identify my Volt charging from empty to full in 3:30 hours last night:

A few interesting things:

Every hour there is a short 1-minute period where the charger uses only half of the rated power. Why is that? Is it to re-calibrate the voltages in the battery management systems without a charge current? It's not clear to me. This power dip is always the same and exactly on the 1-hour mark, so it's not to control thermal issues.

At the end of the charge cycle, the power draw reduces gradually over a 10 minute period. That surprises me. Why did GM engineers go to the trouble to program this gradual decline? Is this perhaps done to approach the maximum cell voltage across all 97 groups slowly? That could be, but the normal charge/discharge currents while driving are 10x higher, so it would not make a difference.

Both the above things puzzle me. Anyone has an idea?

From the graph I calculate a full charge at 11.92Kw, which is less than I always had assumed. On average I see 10.2Kw usable in my 4.5-year old Volt, so the battery efficiency is ~86%. The rest is lost as heat in cables and in the charger.



Sent from my iPhone using Tapatalk

 

[ClarksonCote]

I'm not sure about the reverse pulse every 1 hour, but the gradual decline at the end is likely because the battery charging goes from constant current mode to constant voltage mode as the battery nears its full state of charge. This would keep ramping down even further if GM didn't have a buffer at the top and bottom of the battery to keep the chemistry from its extremes (which extends the life of the battery).

This switch from Constant Current to Constant Voltage is very textbook for any Lithium Ion battery charging process once it reaches a certain voltage level during charging, and is a requirement of the battery chemistry.

 

[hvacman]

Robbert - Interesting! The net efficiency doesn't surprise me. At 240 VAC input, I believe the Gen1 OBC is about 90-91% efficient (Gen 2 is 95% efficient!). I also believe that the battery self-resistance losses during charging is about 5%, so that all works out to net 86% wire-to-charged-battery efficiency. The hourly dip and end-of-charge taper - hmmmm. good question.

It would be interesting to see what the charging watt profile looks like with the portable L1 EVSE and 115 VAC plugged into it, set at 12 amp charge rate - if it has the same hourly dip and the end-of-charge taper.

 

[2VoltFamily]

Volt charging from empty to full in 3:30 hours last night:

This is probably the only thing that surprises me. Usually from completely empty to a full charge takes over 4 hours.

Every hour there is a short 1-minute period where the charger uses only half of the rated power. Why is that? Is it to re-calibrate the voltages in the battery management systems without a charge current? It's not clear to me. This power dip is always the same and exactly on the 1-hour mark, so it's not to control thermal issues.

I have measured mine with an ODB2 and android software and never see the dip. It may be possible it is somehow related to the tool you use to measure. I will have to go back to my data and see if that occurred and I just missed it.

At the end of the charge cycle, the power draw reduces gradually over a 10 minute period. That surprises me. Why did GM engineers go to the trouble to program this gradual decline? Is this perhaps done to approach the maximum cell voltage across all 97 groups slowly? That could be, but the normal charge/discharge currents while driving are 10x higher, so it would not make a difference.

The dip at this end also occurs on my measurements. This is typical of any smart battery charge system. What does not show on your graph is after the dip there is some slight bump in charging or draw which seems to be using the pumps on-board for cooling the battery and on-board charger.

 

[edk-austin]

I would expect the Volt to monitor the cell leveling status during the "pause" in full charging. Could be a point where it decides to continue or throw a CEL. The end ramp is where the charger goes into constant (maximum) voltage charge. Another interesting point that I've measured is the ramp at the beginning. Volt doesn't go full on to 3.3KW (or even 1KW L1), but will slowly ramp up to full power over about 10-12 seconds. You can even see this on a Kill-a-Watt meter for L1.

 

[RobbertPatrison]

I would expect the Volt to monitor the cell leveling status during the "pause" in full charging. Could be a point where it decides to continue or throw a CEL. The end ramp is where the charger goes into constant (maximum) voltage charge. Another interesting point that I've measured is the ramp at the beginning. Volt doesn't go full on to 3.3KW (or even 1KW L1), but will slowly ramp up to full power over about 10-12 seconds. You can even see this on a Kill-a-Watt meter for L1.

You are right: The ramp-up is indeed about 12 seconds.

Sent from my iPhone using Tapatalk

 

[siai47]

I have never watched this on most of the other EV's I have owned. However, I did observe this on a Mitsubishi I-MiEV that I owned for a couple of years. About 10 minutes into a charge cycle, the charger power was reduced to almost zero for about 30 seconds and then ramp back up again for the remainder of the charge. It would taper off at the end and if cell balancing was required, it would stay on for additional time at a very low level until balancing was complete. This "pause" must be something the BMS does to check cell voltages at rest during the charging cycle to confirm the status of each cell before resuming the charge. Most likely found in many EV's if you are looking for it.