dnmun said:
ping gave .015 ohm for the 4aAh cells and from my analysis of RLT's last graph i would estimate the internal resistance of these 10 Ah cells to be about .030 ohm. so twice the resistance for twice the size or is this just the peukert effect?
there are now 2 independent measurements of 6 mOhm for this cell. is the remaining 24mOhm due to the peukert effect? it should drop to zero as the load current drops to zero then. how to separate the 2?
If we look at RTL's discharge graph of cell #25 above, we effectively see somewhere around 30-40mohm internal resistance:
0.1 Volt difference / 3 Amp difference = 30mOhm :|
As far as I know, peukert's effect has nothing to do with internal resistance of a cell. We use Amps to check the voltage drop produced by internal resistance, not Amp-Hours (Ah).
LiFePO4 supposedly has little peukert effect compared to other chemistries (from what I've heard - may partially be hype), but we DO seem to be observing a fair amount of capacity loss with higher discharge rates, which would mean it does suffer from peukert's effect after all. Just not as heavily as lead, that's for sure.
i used a rough estimate from an architects scale on the screen. averaging by eye the difference between the 8 A test and the 5 A test, ignoring the freshness or charging effect on the battery. i measured the voltage differnence at each Ah from 1 to 7Ah.
1,2,3Ah all about .09V, 4,5Ah up to .1V, 6,7Ah up to .11V, and by 8Ah up to .14V
maybe you can measure more accurately off your CBA.
the diffence in current is 3A, so the effective internal resistance under load as it is discharges begins at about 30milliohm and climbs to about 35 after 6Ah of discharge, and then breaks dramatically to 40milliohm at 8Ah as the cells get close to cutoff.
This is what I observe too, but this data comes from three
different charge cycles on this cell, which may be a problem. Cells #14 and cells #15 give about the same results as this #25 cell (see RLT's post above).
RTL, could you double-check to make sure your cell analyser is not measuring any voltage drop that may be happening outside the cell (in its own wiring I mean). Use a multi-meter to check the voltage directly on the cell's tabs while it is under test to see if you get the same reading as the analyser is giving you.
We either are getting bad readings, or these recent cells are far from what they really should be!
I'm going to check back in old posts to this thread, because initial results from these cells gave around 8-10mohms resistance... FAR from these last results!
i think there is an education in the cell's lifespan in these graphs, it would also be nice if we thought that headway had their engineers analyzing this too.
Hummm.. I'm not so sure if this indicates problems with the cell's lifespan, but it does show that there is some problems with consistency of the manufacturing process (which may also result in reduced life cycles with some cells, but we have yet to prove this).
we actually need to have headway's technical staff come on here and tell us what they think. i will recommend immediately that headway begin marking and recording a tracking number for each individual cell and tracking them during manufacturing in order to establish feedback for the kinda quality control they will need to maintain stability in their manufacturing process long term. in a previous life i helped bring an IC manufacturing process to life. we wasted literally 10's of millions of dollars making dead IC's based on our own engineering estimates. after we talked the wafer manufacturers into laser marking the wafers and developed inline wafer tracking software and developed the process to make those correlations, yield immediately stabilized, equipment and process problems stood out like sore thumbs and it was only using final yield feedback to the process level that we optimized the overall performance of the IC's. headway and the other battery manufacturers need to learn it too, just like all the other IC manufacturers did. the cost would be nothing compared to the legal/insurance protection they will receive at court in the future. imho.
We can only wish they are savy enough at Headway to put some better QC in place!
battery users could then track those cells for the remainder of their life too. not everyone will have a CBA to record the performance of their cells, but everyone who can should begin data logging their batteries, and we need to have a standard format for the data so that we can use large groups of data to establish or evaluate the life span of the cells and what really does cause degradation. set up a special file sharing format maybe, allowing people to record their observations, and maybe keep track of them as lots, like the current group buy where the cells will be manufactured at around the same time, to see if different problems are more common among individual lots.
I like your idea of a common place for us to accumulate well organised info for packs and cells, and not only headway. We would also need detailed testing procedure information so that reported results would have more of a chance of being accurate. Some kind of wiki maybe?...
RLT: Thanks for all the results you are posting, we will try to make sense of all this with a bit of time! Here's a few suggestions for your next tests:
- Check to make sure your analyser is not adding it's own voltage drop to the cell's own V drop (see above)
- Could you test the voltage drop of the cell under test during the same discharge cycle (instead of during three different discharges)? For example: measure the cell's voltage when discharging first at 3A, then stop discharge, then measure V at 5A, stop, then V again at 8A. Make sure you avoid the initial discharge voltage drop at the begining of the cycle (wait for a fairly stable voltage). Also try to do each of these readings fairly quickly before stoping discharge, then start up again at your next rate, wait a couple seconds for V to stabilise, then measure quickly again before stoping, etc...
) 
). While being off by .07V is sort of significant, due to the steepness of the curve after about 2.8V, does that throw off my AH results overall by a significant amount?
