Lipo Battery Standards, open discussion

[Just go Play]

It comes as no surprise that the two of you would be able to work out some reasonable baselines for testing of this sort. It would surely benefit the consumers the most and help the technology gain acceptance more readily.

I have a couple of thoughts on this that I'll toss out for your consideration.

Since there is no current standards per say why not set the minimum per cell voltage to the nominal cell voltage under its continuous discharge rating? I understand that this would likely shift the relative C rating we are used to seeing current packs rated at. But since the current ratings are vague and somewhat random why not?

I think that testing and rating should be done at 2 levels really. The initial tests should be of the individual cells from the mfgr's and then packs from the various builders should be done also since materials and methods of pack building vary widely and will contribute to actual performance and thermal characteristics in actual use. I know that personally I could care less about individual cell performance if there is any change once it is assembled as a 2s4p pack for example.

One of the things I'm unclear on is the degree that variations in operating temperature will effect performance of a cell. I have heard that some cells work best at 110 - 120 df for example. I can only assume that this may be somewhat brand specific as there are various chemistries being used in different cells. This may be something that needs to be considered if possible during testing.

One of the packs I have lists only the following concerning thermal conditions.
During discharge do not exceed 60 degrees C.
Do not charge in freezing temperatures (assumed < 0 degrees C)
Optimal charging temperature 20-25 deg. C

I would like to see some sort of testing that would reflect use in 1/8 offroad vehicles as it seems to me that they present somewhat unique operating conditions compared to boats/planes/helis and the like. This may not be needed but I have never seen comparative graphs of packs under high constant load and repeated extremely high peaks like you see in offroad use.

I may have more ideas after some sleep but thats all for now. Hopefully some of this makes sense or is possibly useful to you.

[DrKnow65]

Seems to me that the most relative tests would mimic actual real world RC use. You guys have mentioned how spikes for heli-boat-plane-car are all different. Also the spikes are relative to vehicle weight and performance.

I would think that a massive collection of eagle tree data logs would be the best indicator of how to test the cells. An average of the data logs for cars would give the average need for the lipo. Burst duty cycle, amps between bursts, continuous, braking, ect. Devise a standard car cycle to run the lipo through, note the voltages and temps, calculate a "P" rating from there (performance). Devise a heli cycle, plane cycle, ect.

The trick of course is the test equipment, the environment, and the guy with the time and cash who has nothing to do with anything :)

[lutach]

Long term test would be good too. It might take long, but would help to see at what cycle the cell(s) starts to loose some capacity.

[hoober]

I personally like the CBA graphs at various "C" ratings. Various current ratings are even better. To keep it simple a single cell should be graphed at the various currents and a temp rise and max put in description at each current.

I like to look at th graph and make my own interpretation as to what I would "rate" the cell at.

Doing packs opens up another can of worms, but is far more realistic. If someone wants to discharge a certain pack at a certain "C" factory rating it is nice to see the results.

[Arct1k]

Question - Is it worth plotting charts under varying load - sign wave? - In real world you will hardly every draw constant ampage. I wonder how well different packs take this.

[BrianG]

Personally, a graph using the testing procedures found in my previous post is best. It's not perfectly ideal, but should encompass the worst-case use a lipo will endure. At the very least, it will rate all cells using the same conditions and procedures for a better apples-to-apples comparison.

[aqwut]

Originally Posted by BrianG

Personally, a graph using the testing procedures found in my previous post is best. It's not perfectly ideal, but should encompass the worst-case use a lipo will endure. At the very least, it will rate all cells using the same conditions and procedures for a better apples-to-apples comparison.

Yeah, that would be nice... Worst case scenario would be best.. I like it when batteries are underrated.. I usually underrate them myself by about 20% when figuring out speed, power, voltage & capacity.... It's always nice to have a battery pack that can perform whatever is marked on the sticker.. without any fine prints.. but I guess that would be hard to do..

Originally Posted by lutach

Long term test would be good too. It might take long, but would help to see at what cycle the cell(s) starts to loose some capacity.

LOL, I bet you'd love to be the person to perfom all these test... I know I wouldn't mind.. house full of LiPos, imagine that...

[johnrobholmes]

I think that both constant and burst testing will prove useful. The Constant rates will allow easy comparison between cells. The Burst rates will allow the true ability of a pack to be tested and empirically logged. Even 1/10th second bursts could be used to get an idea of how high a pack will spike before dropping out.

[lutach]

Originally Posted by aqwut

Yeah, that would be nice... Worst case scenario would be best.. I like it when batteries are underrated.. I usually underrate them myself by about 20% when figuring out speed, power, voltage & capacity.... It's always nice to have a battery pack that can perform whatever is marked on the sticker.. without any fine prints.. but I guess that would be hard to do..



LOL, I bet you'd love to be the person to perfom all these test... I know I wouldn't mind.. house full of LiPos, imagine that...

I can definitely tell which one is good . If they can handle our stuff, they can handle anything. A video of the test should also provide some form of where the data came from. A burst test should be done to confirm the pack can withstand the burst rating. Afterwards one should be able to establish a safe continuous and burst C rating on their own and not just go by what a factory puts on the cell. It all comes down to a little commom sense on the rating of cells and to avoid situations like the ones mentioned. All the batteries I've tested and owned does come with a factory C rating, but I like to beat them up a bit to see if they can live through it.

[BrianG]

Originally Posted by lutach

I can definitely tell which one is good . If they can handle our stuff, they can handle anything.

It just sucks that we have to do a live test to see. It would be nice to know the true capabilities before they puff. ...Kinda like testing to see if food is poisoned by tasting it.

[lutach]

Originally Posted by BrianG

It just sucks that we have to do a live test to see. It would be nice to know the true capabilities before they puff. ...Kinda like testing to see if food is poisoned by tasting it.

If I had a CBA, that's what I would do. Then people wouldn't have anything else to say. Who knows where those graphs came from. I can get you great graphs from Enerland, change a few things here and there to make it seem like it was done by me. Make a video of the test doing the 1C, 10C, 20C and burst ratings. That will definitely be the most effective way for me.

[BrianG]

OK, so who's gonna spring for the equipment and begin testing?

[DrKnow65]

How about a graph where the cell is loaded to what ever amps it takes to drop it's voltage to 3.2v. Then as the test progresses the amps will fall off to keep the cell at 3.2v untill the amps hit 0. Note the temp rise from an established 80*f. Cutoff voltage may need to be higher or lower for the cell to never pass the "maximum safe" temp ~140*f?


It would of course take multiple tests to place the proper cutoff voltage with each cell. Example cell#X test 1, LVC=3.8v (test begins) tester loads the cell to what ever is necessary to drop the voltage to 3.8v (say 60amps then decreasing, dictated by LVC) cell temp only rises to 115*f during the test. Recharge cell to 4.20v cool it to 80*f. Cell#x test #2, LVC=3.6v (test begins) tester loads cell (say 85amps then decreasing dictated by LVC) cell temp rises to 135*f during the test. Again at 3.55, 3.50, 3.45---

So on and so forth untill either the max safe temp is hit or the LVC drops below 3.0v. At that point a graph is made, showing the cells maximum constant performance.

The variables for air flow would need to be set, and a tester would have to be fabricated. We could use a data logger to show the volts, amps, time, and temp in a graph.

Wouldn't an ESC with a soft cutoff (proportional to lvc) be able to do this for us? We would just need a load that will be able to stay constantly above the required amp draw.

[BrianG]

Exactly. Do the test for several C points like how I originally stated. Eventually, you'll find the C rating that maintains the minimum voltage you are looking for. Some people may find 3.0v an acceptable cutoff, while others may accept only 3.3v as the cutoff. The graphs would show this and allow people to make informed decisions.

Then, compile all the graphs and make a little datasheet just like how electronic components are rated.

I don't know about other people, but my goal would be to provide the information for buyers and let them make the decision about what is adequate for their needs.

[johnrobholmes]

When finding the constant load on a battery with voltages as low as 3.2 or 3.0, we should find that the capacity does not hold up to the 90% standard. Drawing a battery at a rate that depresses the voltage to 3.1v/ cell average should make it fall out at a terrible capacity.