Days Fighting Software

I spend my days working around things written in software, but I guess nothing unique about that. It may be a road you've been down, or it may be useful to someone else, so let me tell you a story. 

The center of this story is a computer interface for battery testing from West Mountain Radio, the exact product I'm using apparently is obsolete, but they have a page full of what they call computerized battery analyzers here.  The one I have, the CBA-IV has been replaced by the CBA-V on that page. I originally got this at the Orlando Hamcation from West Mountain on a Hamfest Special price, and judging by my saved pictures of plots, that must have been in February of 2017.  

This is not a charger; all it does is a controlled discharge of the battery, and handful of other tests.  In the world of battery makers and their specifications, the rates at which you can discharge, as well as recharge, are important limits.  For batteries that might run a power tool, they tend to be rated to the capacity in Amps*hours (Amp Hours or AH) while for something like a car starter, or the various jump starter batteries you can buy now, those are rated in "Cold Cranking Amps".  In the first category, the Ryobi tools have batteries rated in Amp Hours, in different output voltages. 

Judging by the number of Ryobi tools I see around, I'm guessing a fair percentage of you have some.  I have three 4AH 18V One+ batteries, along with a 2 and a 1-1/2 AH.  With these batteries, it's pretty easy to decide on a discharge test. Consider the 4AH battery.  That means, if you were to draw a constant 4A, just the capacity C, it would discharge to "dead" in a bit under one hour.  If you discharge it 2A, that's half the capacity or C/2, and the time to discharge goes to a bit under two hours.  To get closest to the 4AH rating, the convention has been that the manufacturers pretty much rate them at the C/10 discharge or 0.4Amps. For the 2AH battery, C/10 is 0.2A, and so on. As rule of thumb, the lower the discharge current the longer the battery charge's life. That is, if you discharge a battery at C/20 and C/10, the C/20 discharge lasts longer than exactly twice what C/10 lasts and so on.

I use the CBA-IV a couple of times a year, and doing the cycles on my various rechargeable batteries pretty much can take up a month. I started out in the middle last week and had some strange things happen.  Let me start with an example of the output from the CBA.

The software popped up that screen and I just screen-captured the whole software window with the discharge curve and the measured results. Note it measured 3.78 AH, and the tab conveniently says 240625_Discharge.  The numbers are the date of the test, 6-25-2024, so last June 25th. 3.78 AH out of a 4AH hour claim is pretty good - 94.5% - especially for a battery that's several years old.  

When I went to test this same battery this week, I got a result closer to 1.2 or 1.3 AH, which is awful.  So what's going on? 

It's in plain sight on this plot.  See that horizontal line two minor divisions above the bottom of the plot?  There's a green arrowhead on the left end of the line. That line is the voltage it was discharging the battery to in order to decide the test was complete there.  That's 15.5 Volts.  While setting up the test to run this week, it said the test should end at 18.6V.  If you eyeball the plot to see where the red curve looks to hit 18.6, it looks to me like about 1.5 hours.  On the other hand, if you look at the right side of the curve, you can see that it goes into a dive downward toward fully discharged at just about 17V. Setting the discharge to 17 would get all but that last roughly 0.2AH of discharge time.

So today, after charging this battery fully, I set the discharge voltage to 17.0.  The software immediately popped up a warning that it wasn't their recommendation, "are you sure?"  I said  yes. The results came out at 2.96AH which is quite a bit degraded from this test which looks to be 3.6 at 17V.  At the moment, I'm testing the other old 4AH battery to see how it compares to this one.  Then I do the new battery.

What's up with the 18.6V vs 15.5 or the compromise 17V?  The gotcha is that I updated the software in the CBA at the start of the testing - call it Tuesday the 3rd. They apparently changed the recommended voltage in the SW update.

When things go wrong, as they sometimes will, to butcher the poem, look at the software first.

EDIT at 0930 Monday Dec.9:  It became clear this morning that I had remembered the results of that all important last test very incorrectly after I turned the computer on and compared reality to my memory.  The last three paragraphs (before the last line) have been revised to reflect the real data from the shop computer.