Estimating HV battery capacity in-car?

[RobertSmalls]

I'm dropping my cycled & balanced junkyard HV battery in this weekend, and I wanted to record the usable capacity of my old battery versus the new one, in the car.

With my current battery, I was able to get 91s of assist, at two bars less than full assist (it's cold here) before the assist died down to four bars, then zero bars. This brought the SoC down from 100% to ~65%. A few minutes later, I got a recal and a forced regen. Is there any way to guesstimate how many mAh this represents?

[Mike Dabrowski 2000]

Robert,
Since the assist guage is not giving you true assist current, the test will not be very useful or repeatable.
You can use an external Hall effect current probe to see the true current, or tap into the white battery current probe voltages. The output of the current sensor is about .02V/Amp referenced to chassis ground.

A suggestion made several years ago was to let the car sit in ACC mode, with the lights on as the load, when the 12V system needs a charge, it gets it from the HV battery, so this will drain the HV battery at a reasonably uniform and repeatable rate. I don't know what happens when the HV battery runs out of juice, as I never tried the technique.
Good luck

[JimIsbell]

I use an estimation method that seems to work fairly well. The SOC represents 100% (4 ah) so each bar represents about 5% You had 35% (seven bars) available before RECAL so you had somewhere in the neighborhood of 1.4ah capacity.

This is not perfect, but it comes pretty close to what I get with the CBA after I get it out.

I am currently running a battery with an 8 bar window, one bar more than yours, and I can keep it down to one RECAL a week easily. If I didnt have to drive at night on Friday night I could keep from having any RECALs. But when I drive at night it gets charged all the way to the top and then RECALs at 12 bars, before the background charge ramps up. During the week I try to keep it below 15 bars, RECAL would then be at 7 bars, after the background charge ramps up, which prevents it from getting down to 7 bars.

[E27006]

Isn't it an Either/Or situation during a Recal?

The Charging starts and the SOC increments steadily, a fixed quantity of charge (4Ah or so) is delivered to the pack,Either, when that measured quantity is reached ,OR, the pack voltage reaches a maximum of about 170V then the Recal is complete and the SOC adjusts to read 19 bars
You could have an aged pack and only a couple of AmpHours of charging would fill the battery, the SOC would do a positive recal in this situation.

It must be very hard to work out the capacity of the pack from observing a recal

[JimIsbell]

My car always charges from 2 bars after a recal.. then it discharges until the computer thinks it is empty.

This could mean charging till full 20 bars or maybe only to 10 bars if I begin discharging it back down before it gets to 20 bars. But the constant is that, from the "top" where ever that was, I can discharge 8 bars and then will get a recal. This means that the computer thinks my battery has a capacity of 8 bars.

If its night time or if its a long level road trip, it will charge all the way to 20 bars. I have NEVER seen a positive recal. It just charges till the SOC shows 20 bars. Then, if I begin discharging it, at 12 bars I will get a recal down to 2 bars from where it will begin the recharge again.

If I keep it from going above 15 bars by using the IMA aggressively (cant be aggressive enough with the headlights on, so night time means I will get up to 20 bars within 20 miles) then whatever the top is when I begin a long discharge, it will drop 8 bars from that top and generate a recal back down to 2 bars.

Most of the time I can keep it between 15 bars and 7 bars and never generate a recal. The problem comes if it charges to 20 bars, because at 12 bars it will generate a recal. BUT...the aggressive background charging from the computer doesnt start till you are down to 10 bars so its easy to discharge it below 12 bars and generate the recal. But by keeping the "top" below 15 bars, it is easy to prevent discharge below 8 bars because the aggressive background charge will provide a substantial floor to the discharge.

BUT...and this is important, a recal is only the computer scratching its head and wondering what the battery contains. The actual charge on the battery is not changed by the recal...until it begins aggressively charging because it thinks the battery is discharged. When the bars go down in a recal the battery is NOT going down. So the recal only affects the drive ability of the car for a few miles until the computer decides there is enough charge to start using the IMA again.

I consider a recal as a normal and un eventful occurrence.

What you need to know, and what is the most useful thing you can get from the SOC, is what is the "window" of your particular battery. Mine is 8 bars. Then with that information, manage the charge so that you stay within the window. In time mine may deteriorate to 7 bars and then 6 bars, etc. It will become very hard to manage at that point. To find that window, charge up to 20 bars and then begin driving aggressively until it recals. Watch the SOC carefully for that break point. When you hit it, you know that your window is 20 minus the break point.

Now to determine the capacity...APPROXIMATELY...multiply the window times 0.2 ah. For me that is 1.4 to 1.6 ah depending on the day. When I pulled the battery and did charge discharge cycles on individual sticks using the CBA I found two sticks with a capacity of approximately 1.5 ah. When these two drop below a certain voltage they trigger a recal. Thus the capacity of my battery is confirmed at 1.5ah just as predicted by the SOC.

[3-Wheeler]

Jim,

If you had the Calpod IMA disable, you would then be able to control the amount of regen that is done while driving at night.

By using this switch, there are at least several more mpg to be had too.

Jim.

[RobertSmalls]

Originally Posted by JimIsbell

Now to determine the capacity...APPROXIMATELY...multiply the window times 0.2 ah. For me that is 1.4 to 1.6 ah depending on the day. When I pulled the battery and did charge discharge cycles on individual sticks using the CBA I found two sticks with a capacity of approximately 1.5 ah. When these two drop below a certain voltage they trigger a recal. Thus the capacity of my battery is confirmed at 1.5ah just as predicted by the SOC.

Shouldn't you get more Ah on the bench than in the car, because the Honda BMS avoids fully charging and discharging the battery?

[Mike Dabrowski 2000]

I think that the bottom line here is the there is so many variables and uncertainties in the accuracy of the SOC display, that it is not a useable indication of capacity.

[JimIsbell]

Originally Posted by Mike Dabrowski 2000

I think that the bottom line here is the there is so many variables and uncertainties in the accuracy of the SOC display, that it is not a useable indication of capacity.

I agree, but this method is the closest thing that I could find for a rough estimate without pulling the battery. And, come to think of it, do you really want to know the capacity on the bench anyway. What you want to know is the capacity IN THE CAR, thats all you are going to be able to use,

[Cobb]

The main idea is to make the test repeatable. Like take a route on a long hill and set the curse control and see what happens, then try the other battery.