I am realizing to my chagrin that while I can find bad modules through light bulb (high current) testing, such methods are basically useless for matching modules, in attempting to have a pack that doesn't code out under the stress of high drain/charge (eg accelerating uphill, which happens a lot in Colorado!) So, time for some additional investment. I already own a Prolong charger/discharger. All I am looking for is something affordable that can evaluate module capacity and hopefully IR. (I don't need perfect "absolute" numbers, just comparable measures for my modules ) Any thoughts? I see SkyRC B6AC v2 mentioned a lot. But that's not necessarily the best... particularly since it only supports one device at a time. My further questions on the B6AC: Does it actually measure IR on Prius modules? The docs only mention IR in conjunction with Lion batteries. The Mini version can supposedly be computer-controlled. Does this work? Seems handy
You need lab grade equipment to precisely measure and match modules. It costs way more that $10K to do it the right way. The SkyRC chargers aren't designed for such big modules so their MAH measurements are woeful, especially because they often shut off early and you lose the measurements sometimes when they do that. So it's really about how many measurements you can do DIY style with cheap gear... For example a decent thermal camera attached to your phone will show you bad cells that are not detectable any other way than slightly more heat. Also measuring how a fully charged pack loses voltage in each module after a few days or longer (self-discharge data) And as you already know voltage drop after 2 mins of a 12v / 50w bulb... I put all this Data on a spread sheet that looks like this in order to decide which modules to replace:
Too funny... I own a FLIR but never thought to use it for this. Nice. I'm not looking for "precise match"... just need close enough balance so I don't have one block dropping to 2 volt delta v Re-reading a bunch of stuff, and staring at your sheet (thanks, btw!) it would appear that a simple and good-enough strategy is: * Only measure topped-off modules * voltage change under timed load ought to be a good enough indicator of high internal resistance. Not exactly a "direct measurement" but who cares? I want results * In reality, getting rid of bad modules, and balancing on voltage change under pretty good load "ought" to give a good result And finally: be patient. Teasing out the hidden bad modules ain't for the faint of heart. Does that sound about right?6
Yes sir, you're correct! Measuring voltages when pack isn't fully charged gives confusing data. And the ultimate number I follow is "voltage difference" via Dr. Prius App. In general it's a mess, but I have seven chargers and seven lightbulb discharges so I can do a whole pack with 4 rounds of each, which speeds up the process. Of course I've come across a couple Green-tec rebuilt packs that uses lab grade testing gear and matches modules up with precise measurements and it's truly humbling to realize what's possible in terms of matching up modules near perfectly if you're willing to spend tens of thousands of dollars.
One question: for voltage drop... do you measure at 2 min just before disconnect? That's my assumption. Another supposed-difference-but-probably-doesn't-matter: "officially" you want to measure starting voltage immediately after the start. but i'm not convinced that does anything other than add uncertainty: voltage is rapidly dropping on initial connection. One thing that definitely causes some havoc: highest voltage drop is NOT necessarily associated with current maxed-out voltage. I finished an extended three-round discharge/charge... and at the end the module voltages (at full charge, and after hours of "relaxation" vary by up to 0.130 volts. So matching discharges may leave me with built-in min deltaV of 0.13 volts.
