Just to clarify, I am testing whatever cells I have 'harder' than manufacturer tests them:
| EHW5 manufacturer's test | mudder's test |
| Charge current (amps) | 40 | 52.5 |
| Discharge current (amps) | 40 | ~90 |
| Ambient Temp (degC) | 23 | 25-35 |
| Forced air cooling? | Yes | No |
| Linearized SoC capacity reduction per 10000 cycles (%) | 4 | ???** |
**My first cell test data is not valid, as the cell under test was accidentally pulled all the way down to 0.000 volts (due to a data corruption issue caused by rain/dirt shorting out the datalogging PCB).
Regardless, whatever cell I'm testing is certainly 'good enough' to use in the G1 insight. What I mean by that is that age - not cycle count - is going to ultimately kill these cells... by a country mile. It's not even close. The QTY950 cycles I ran on the first cell is equivalent to driving the car hard every single day for three years... and only resulted in "5%" SoC reduction over ~QTY650 cycles (again, flawed data).
...
Let's look at cell#2, which has been under 'proper' test for QTY43 hours. It's certainly too soon to draw meaningful results, but thus far we've logged QTY349 complete cycles (averaging one complete cycle every 8.1 minutes). Right now cell#2 is losing on average 284 uA/cycle, which is ~-5.7% capacity per 1000 cycles. So actually it's only slightly better than the last test... right now. But as you may know, lithium cells tend to lose quite a bit of capacity over the first several hundred cycles, and then level off... so we'll need to wait maybe a week before we draw any "long term" conclusions.
Fortunately, as I mentioned above, 5% capacity reduction per three years of driving is pretty minimal, and certainly cell aging will be what kills these cells... not cycle count. That's ultimately all I care about. Whether or not an 'actual' EHW5 module performs better isn't terribly important; cell aging will kill them both, and that's all I care about.
batteryhookup.com
