Battery performance vault?

[johnrobholmes]

Who else would like to contribute to a battery vault around here? I have a good many lipo packs now, and would love to get a running list of batteries that perform well and batteries that don't. Real world logging along with test stand abuse would both be good methods for testing. Although we wouldn't have a continuous load on the packs, test stand results would weed out the cells that don't perform to claims of higher ratings.

[lincpimp]

Well I could give you a list of every cell that I have had poor success with, it is every one that I have purchased that has not been in a 4s or more configuration. I must be too hard on my setups. I have managed to ruin most of maxamps cells, except for their 2100 cells, which are correctly rated in my opinion. I have also had good success with their 4000 cells, 4s and up. I even run their 3000 cells in a 5s pack and have had good success with that too. I have a feeling that once you pull more than 15c from a 20c rated pack you may have issues. I treat all of my 20c lipos that way and have not had issues since then. I guess that I am really not stressing my batts now.

It would be nice to get some good, solid data on certain mfg cells. How do you suggest that we do that?

[johnrobholmes]

My plan is to simply data log with two methods. First method is in a vehicle or sled pull. Burst currents will show the true burst rating of the cells (assuming enough amperage is drawn). Second method is on a test bed with a variable pitch prop and appropriate motor. This will give an easy to find continuous current rating. There will need to be a voltage standard set, such as 3.3v/cell under burst for "good" performance. Lower is poor, higher is excellent, etc. Same for continuous rating. All cells will need to be tested at the same temperature.

We will need to set a voltage standard and temp, as well as sampling rate standard for burst amperage measurement. Seeing as how most eagleltree loggers are 10hz, that is a fine sample rate. 95-100*f would be a good pack temp to start with as well. As for voltage standards, I would like input on that.


What got me thinking about this is the wild variation in pack sizes and weights between brands. For example, between two brands of 20C 4000mah cell there is 25gm/ cell difference and about 20 cubic mm volume difference. One brand is smaller and lighter, the other is heavier and bigger. My gut is telling me the larger and heavier pack will have better performance.

[ClodMaxx]

i think it's a brilliant idea, and will be an invaluable source of information. of course, i would imagine at least one battery company that may not appreciate it.

[johnrobholmes]

There may be many battery companies that don't like it. I saw the fuss that was made on RCG about it. If they are going to mislabel cells, then maybe they are in the wrong business. I don't think my polyRC 10C cells are going to hold 10C, although to their credit they are only rated 12C burst so it is a very thin line. That is also why I think a standard temp is a good idea. If the cells are cold they won't perform.

[phildogg]

I'd love to see how well some companies amp rating holds up.
phil

[johnrobholmes]

I think 3.3v/ cell will be a good starting point for burst references, and 3.7v should be the continuous rated voltage.

[Serum]

yeah, though most companies use burst reference as the current that makes the cell drop way bellow 3V though.

I am still impressed with the enerland 5000 packs; 3.6V@110A load. (this load was in a burst) but still impressive.

[Serum]

I have some true-rc packs here; 4000mAh 4S delivers 1400 watts, and they drop to 3.1V per cell; not bad considering their price; and the fact that they are not their latest and greatest cells. Their new cells are way better.

for that price/weight they are, unbeatable.

[johnrobholmes]

That is the big problem I see, there is no standard reference for companies to rate cells. One companies burst rate may be a 2.5v cutoff, another may be 3v. For longevity I wouldn't want my pack dipping below 3.3v/cell.


The price/performance will be interesting to find out. With some hard data I think a price/watt hour ratio may be calculated.

[johnrobholmes]

I'm just gonna start piling my own graphs in here for now. For ease of interpretation a voltage vs amperage graph is best.

Hyperion 16C 2100 4s pack. Didn't ever get above 100*f, maybe not warm enough to perform? Hitting 3v/cell at 38amps at one point. Third run on the packs I think. I am going to shorten the wires and see if I can get better readings, there is too much wire on the rig right now.

[Just go Play]

Here is a chart of a pair of 25C trakpower 2s 4900 bricks that I'm returning for exchange. Much to my surprise they don't even perform as well as my previous 20C packs.

Results are from eagletree logs then dumped to excel and averaged of the 7 runs following 3 initial runs for break in. Temps starting from 70-80f peak at 115f consistently after 2200mA discharge. Voltage drop reaches 3.0v per cell @ 30C. I'll post results from new packs after I get them for comparison.

[Robert]

If you guys really want to do this I'll build us a web application that can hold this data in a database so we can archive it and search it. It will be much more useable in an database than just a bunch of text in forum posts. Pm me and we'll start to figue it out.

[Robert]

I can't believe no one jumped on this offer. I'm headed out to the pool and bbq. Pm me if you change your mind. If I've dirtied the thread feel free to remove the posts. :)

[johnrobholmes]

That sounds like a good idea to me! What kind of DB are we talking about here? I'm writing an article for a mag this summer that will have data on a ton of cells. So far I have come to the conclusion that no cell holds up to factory rated bursts in a Car- we just don't warm the packs up enough.