Battery pack balance charging times

[aisbell]

The goal of a balance charge is to raise the charge level of all cells in the battery pack to "full" so that all cells have the same charge level (i.e., the battery pack is balanced). Because many of us have Insights to reduce our energy usage, I would add to this goal that the charge time should be the minimum required to reach the fully-balanced state (i.e., to continue charging the battery pack after all cells are full is a waste of energy). Various balance charging times have been recommended. What balance charging time is perfect?

To answer this question, I will try to estimate some charging times. The numbers that I will use are approximations that some of you could improve. I might also make some incorrect assumptions that you could correct.

The minimum charging time would be when the battery pack charge level is at its IMA maximum charge level and the battery pack is fully-balanced (i.e., all cells are at the IMA maximum charge level). The charger would then raise the charge level of every cell from the IMA maximum charge level to 100%. If any cell is at a lower charge level than the IMA maximum (i.e., some imbalance exists), more charging time would be required. If the charging time were less than this minimum, the battery pack would not be fully-charged and some imbalance might remain.

The maximum charging time would be when the battery pack charge level is at its IMA minimum charge level and the battery pack is fully-balanced (i.e., all cells are at the IMA minimum charge level). The charger would then raise the charge level of every cell from the IMA minimum charge level to 100%. If any cell is at a higher charge level than the IMA minimum (i.e., some imbalance exists), less charging time would be required. If the charging time were more than this maximum, the battery pack would be fully-charged and fully-balanced, but energy would be wasted as heat which could raise the temperature of the battery pack to a possibly damaging level.

Assumptions:
- the charger's output current is 100% efficient in charging the battery pack;
- the charger's output current is constant at 0.35 amps;
- the battery pack's capacity is 6.5 amp-hours (Ah);
- the battery pack's capacity at the IMA full charge level is 80% of the battery pack's capacity, or 5.2 Ah;
- the battery pack's capacity at the IMA empty charge level is 20% of the battery pack's capacity, or 1.3 Ah.

Minimum balance charging time = (6.5 Ah - 5.2 Ah) / 0.35 A = 3.7 hours

Maximum balance charging time = (6.5 Ah - 1.3 Ah) / 0.35 A = 14.9 hours

Do these calculations make sense? I realize that one of Mike's chargers starts charging at 1 A and eventually shifts down to 0.35 A which complicates this calculation.

So if these calculations are accurate, charging an IMA battery for more than 14.9 hours accomplishes nothing, wastes energy, and raises the battery pack temperature unnecessarily. Charging less than 3.7 hours would not fix an imbalance. But our battery packs are almost always somewhere in between these minimum and maximum states. With the instrumentation that's now available (e.g., the OBDIIC&C gauge), we can know much more about the current state of our battery packs. Do we know enough to be able to adjust our charging times based on our current battery state while being certain to balance our packs and to not charge longer than necessary?

[Cobb]

Good question. I do mines til it stops wondering and the voltage is dead on for a 30 minute window. WIth the i2 its typically 121 volts + or - a few decimal places depending on outside temperatures. Since its a no thrills charger with no safety's than a fuse, I watch it like a hawk.


Sent from my Autoguide iPod touch app

[Mike Dabrowski 2000]

The grid charger will stop charging when the plateau in voltage is detected, so it is not wasting energy. The soak on the initial charge will top all cells and will turn it all into heat at the end, but my experience is that some cells can lag the others by a substantial amount, and an extended low current charge will get those lagging cells to full, and from there, a few good cycles, and the pack should get better.

The time to reach the plateau will vary based on where in SOC you start the charge. If the data logging is used, you can see clearly what the plateau looks like, and as the pack capacity gets better, the slope of voltage vs time will get extended as the higher capacity cells with the same charge current will take longer to reach the plateau.
Log the first charge, and the subsequent ones, and compare the graphs.
The capacity settling into a maximum will indicate when further frequent charging is not required.

[aisbell]

I currently use a simple IMA battery charger like those that HBR is selling with their BetterBattery. I use it only to keep my battery pack well-balanced, not in "PHEV" mode where I would discharge my battery pack using MIMA as I drive home and then fully charge it with my charger prior to driving. Apparently, there are quite a few of these simple chargers with some people unsure how long to charge their battery packs.

When I begin using Mike's charger, I would be most concerned about how long a soak time to use since that is what brings the cells into balance. I envision using a soak time with every charge since my goal would be balance rather than just adding power to my battery pack. I would think that the battery voltage continues to increase while "laggard" cells reach full charge. Can the voltage increase that occurs as cells are being balanced be used to determine when balance has been attained (i.e., maybe no further voltage increase indicates balance)? Maybe an increase in the rate of battery pack temperature increase could be used to determine when cells begin to reach their full charge state. I'm trying to figure out whether there's a way to determine how much soak time to use with Mike's charger and how much total charging time is needed for a simple charger. Maybe we don't have enough information available about the state of our battery packs to be able to figure this out.

[Cobb]

As for the i2 it takes 3-4 hours before the pack is charged to 117 volts. The balance act depends on how out of whack the pack is. Ive seen 36 hours tops 12-17 hours shortest. Id rather error on the side of caution with forced cooling in place and waste a few watts leaving it on longer than needed vs stop before its all done as it does effect performance, not necessarly mpg.


Sent from my Autoguide iPod touch app

[Mike Dabrowski 2000]

Art, I suggest that you datalog right from the first charge.
The graph clearly shows the plateau, and will give you some baseline data, and a good insight as to how much soaking will be required. The discharging side of the cycle is just as important, as that lets the sticks stretch their capacity which was lost due to short cycling.

The grid charger code soon will have a mode 6 auto cycle charge mode .
The prototype discharger accessory would use 2 100W tungsten bulbs in series, and have a 250VDC SSR that would control them.
The dual floodlight socket for the bulbs would be mounted to a regular lighting electrical box with an inlet and outlet harness that would put it in series with the charger to car harness.
The control cable would plug into the AUX port on the charger (under the big black hole plug). The AUX connector was not populated on any of the chargers I shipped, so you would have to solder in the right angle connector or send the charger back to me to have that done.
The charger would watch the discharge voltage, and discharge to an adjustable voltage threshold, but would look at the voltage rate of drop to determine if a subpack starts to drop a cell, at which point it would stop discharging and begin another charge, which would charge to the plateau, while measuring AH from a known discharge reference point, so any improvements could be measured between cycles.

The car without MIMA makes it difficult to do a clean deep cycle on the pack, the dis charger will do it while monitoring the pack voltage.

Another feature that may be possible is to use the the fast switching speed of the SSR to measure the IR of the pack by pulsing the discharge on and off with a square wave drive while measuring the voltage difference to determine IR.

[Hybrid-Battery-Repair]

Originally Posted by aisbell

Assumptions:
- the battery pack's capacity is 6.5 amp-hours (Ah);

This is not an assumption you can make. As the batteries age, their capacity drops. In a 2003-2005 Civic, a P1433 error is "battery is below 10% of original 6.5AH capacity". This means that the car is designed to use the battery until it reaches 0.65AH capacity.

[Cobb]

.65?!?!?!? Amps?!?!?!


Sent from my Autoguide iPod touch app

[Eli]

Yes, 0.65Ah. Same as a P1447/P1449 in the Insight.

This is what is stated in the technical documents, but there seems to be a fudge factor.

In the Insight, the IMA light is thrown on the charge cycle, not the discharge cycle, and it can be thrown with significantly more than 0.65Ah inputted.

It seems to be ultimately more voltage dependent than amp hour input dependent, because voltage determines when the positive recalibration will happen. For example, with the weak stock pack I've been testing, I can charge the pack at ~4A to 60% SoC as per the BCM and no IMA light will be thrown. However, if I charge at over 6A, it will do a positive recalibration at less than 40% SoC and the IMA light will illuminate with a P1447/P1449.

Still lots to learn about our batteries and their control systems....

[Hybrid-Battery-Repair]

Originally Posted by Eli

Yes, 0.65Ah. Same as a P1447/P1449 in the Insight.

It seems to only be equivalent to a P1449-78 aka P0A7F in 2005+ cars. P1447 is an out-of-balance pack, and P1433 is always balanced, otherwise it would throw a P1447/P1449 pair instead of the P1433.

I looked at a pack that came in yesterday with a P1433 and every single stick was 8.04V. We haven't capacity tested yet, but it will probably give up 0.4AH - 0.6AH @ 5C discharge.