How far to discharge LiPo's

[JetPlaneFlyer]

In another thread Flyer88 attached the following quote from the hyperion web site:

LIPO CARE

Discharge - Analysis of damaged packs returned to us by users shows that nearly 70% have suffered from "capacity over-discharge" conditions. This simply means that the pack has been run too long, allowing resting voltage to fall too low. Such damage is visible upon cell inspection at the factory. Every type of battery has a minimum recommended discharge level. For example, even deep-cycle lead-acid batteries should not be regularly discharged down to less than 30%~40% of capacity. In the case of lithium polymer, it is best practice to always leave 20% of rated capacity in the pack at the end of a flight, with 10% as an absolute minimum.

To avoid capacity over-discharge, we recommend the setting of an appropriate cut-off voltage (LVC) AND the use of a transmitter flight timer. For LVC, the appropriate voltage cut-off depends in large on how high the max and average discharge rates are. For very high discharge applications, like F5B competition, something around 3.2V (or even lower) may be desirable. For standard aerobatic flight, start at 3.4V to 3.5V/cell, and set your timer to 4 minutes initially. Then note the capacity charged back IN to the pack on next charge. If the pack is 1000mAh, for example, there should be no more than 800mAh charged back in, meaning that 200mAh remained at the end of the flight (20% of capacity). If the charged IN capacity is less than 800mAh, set the timer a little longer, and check again the next flight.

So does that mean that you only take 80% of the rated mAh out, or does it mean that you only discharge them so that when you charge them 80% goes back in (the latter is what they seem to be saying)?


The difference can be quite large. For instance; I run my batteries so that they finish with a resting voltage of about 3.7V per cell, so that's 11.1V on my 2200mAh 3cell pack. LVC is set at 3.2v per cell and does not activate at any time. My Castle data logger says I've pulled 1650mAh from the batteries, that's 75% capacity, so in theory there should be 25% remaining.. On this count I'm safely within the 20% remaining guideline.

However when I charge the batteries they take 2065mAh from the charger, that's 94% of capacity, so only 6% short of fully discharged, on this count, according to the Hyperion recommendation, I'm certainly damaging my batteries.


So am i over-discharging or not. If I only run the batteries to the point where they will take 1760mAh from the charger (80% capacity) that's going to translate to about 1400mAh from the battery during flight, which seems like I'm being short changed when the battery is labeled 2200mAh


Steve

[flyer88]

This is a nice battery tester from Hyperion. It show percentage remaining and cell voltage, difference between highest and lowest cell, Lipo, Life, LiIon about $24.

At about 15-20 % my cell voltage is 3.67 -3.74

Flyer88

[rcers]

I shoot for 3.8v/cell minimum and generally fly to about that. I realize it is a very safe number. That is generally only a seconds to maybe a minute the way I size packs for flight. Just not worth squeezing out a few seconds when a slightly larger pack will do that for you.

I have really changed the way I handle LiPo's in the the last year or so. I am really good about the following now:

• Break-in, I try to do a couple of charge discharge cycles on my new packs before using them in-flight. This limits discharge levels to very low on those first cycles.
• Kindness (as best possible) on the first 3-5 flights i.e. light discharge in-flight.
• Leave battery's sitting discharged after flight. I used to charge them all for the next session. Now they sit until the night before or morning of before charging.
• Because of this 2c charging is common.
• Fly to 3.8v/cell cutoff.
• Better cooling in my planes.
• Purchasing less packs and using them more. I think once you start using them the clock ticks on timeline, but who knows...
• I stick to brands and sizes I know work well.

I don't know if that will make a huge difference. With the lower cost packs you certainly don't have as much at risk now.

Mike

[JetPlaneFlyer]

Originally Posted by flyer88

At about 15-20 % my cell voltage is 3.67 -3.74

flyer,

On that basis my batteries look ok.. Mine typically end up at 11.1V resting for a 3 cell pack, so 3.7v per cell.

But going by the published Hyperion guideline that you should not be putting more than 80% of the battery capacity back when charging the batteries appear over discharged. Charging from 11.1V I put 94% (+/1 1%) of capacity back into the battery.. I've checked this many times and figures are consistent.

Perhaps the Hyperion guidelines only apply to Hyperion batteries, I'm basing this on a selection of Turnigy Nanotech, Zippy and Loong-Maz batteries. Logically if I'm only discharging to 3.7V per cell and the batteries have a nominal rating of 3.7v per cell, then i shouldn't be doing any harm.. surely?

Steve

[slipstick]

You don't say what charger you're using but it might be worth checking how accurate it's capacity measurements are. I've seen some that are horribly inaccurate. When I discharge to approx 3.7V/cell resting I always find considerably less than 94% of nominal capacity going back in...in fact it's usually 80% +/- about 5% (and that's with my old Schulze charger and I've "calibrated" its current readings with different meters several times over the years).

BTW the fact that your pack voltages rebound to the nominal 3.7V/cell resting AFTER being discharged doesn't mean they're never been below that...in fact I can pretty well guarantee that they were well below that DURING the discharge, as I'm sure you're aware .

Steve

[JetPlaneFlyer]

Steve,

Yes, the accuracy or otherwise of the charger is open to debate.. it's Turnigy Accucell 6... These chargers are cheap but generally get good reviews and it's been fine for me so far in all other respects. I could verify by charging through a wattmeter but that would involve soldering up yet more adapter cables to add to the mountain I've already got

I appreciate that under load voltage drops considerably lower than resting voltage. Data logger says momentarily it's going close to 3.2V per cell under peak load when the battery is close to the end of it's run time.. see logger plot attached. I don't think I see anything to worry about in the logger output.

[kyleservicetech]

Originally Posted by JetPlaneFlyer

Steve,

Yes, the accuracy or otherwise of the charger is open to debate.. it's Turnigy Accucell 6... These chargers are cheap but generally get good reviews and it's been fine for me so far in all other respects. I could verify by charging through a wattmeter but that would involve soldering up yet more adapter cables to add to the mountain I've already got

I appreciate that under load voltage drops considerably lower than resting voltage. Data logger says momentarily it's going close to 3.2V per cell under peak load when the battery is close to the end of it's run time.. see logger plot attached. I don't think I see anything to worry about in the logger output.

This info on LiPos is quite interesting, especially the CC ICE printout. Being a dedicated A123 user, its hard to find solid data on the LiPos under active load conditions.

Now if the LiPos have greater "Losses" during the charge cycle, that would enter into just what percentage of total milliampere hour capacity was actually used when putting the milliampere hours back in during the charge cycle.

As for those A123's I've found that if you discharge 70% of the battery milliampere hours, it takes about 71 - 72% to recharge them. That, measured by a Western Mountain CBAIII battery test station, an Astroflight wattmeter, and a Cellpro Powerlab 8. All three of these units have been checked against my Fluke 87V meter, that is accurate to a fraction of one percent accuracy.

My models have my "datalogger" that shows total flights, total milliampere hours, total flying time for the flying season, and either ampere hours or amperes depending if the motor is running or not. That unit is accurate to a few percent, and it also agrees with the Cellpro Powerlab 8 charger.

This is stuff we couldn't do a few years ago. Times have changed.

[JetPlaneFlyer]

Yes, it's fascinating stuff!

I've checked the charger and it looks good. I put 1648mAh into a battery according the the charger a wattmeter in the charge circuit recorded 1637mAh.. That works out to an error of 0.7%

I'm starting to think it's specific to certain batteries, it's the 2200mAh Nano-Tech batteries that I've tested most and they are the ones that were taking over 2000mAh to charge from 11.1v to full. Charging an older 2200mAh Zippy from 11.1v only took 1648mAh (the battery i was using in the above test).. puzzling

Steve

[kyleservicetech]

Originally Posted by JetPlaneFlyer

Yes, it's fascinating stuff!

I've checked the charger and it looks good. I put 1648mAh into a battery according the the charger a wattmeter in the charge circuit recorded 1637mAh.. That works out to an error of 0.7%

I'm starting to think it's specific to certain batteries, it's the 2200mAh Nano-Tech batteries that I've tested most and they are the ones that were taking over 2000mAh to charge from 11.1v to full. Charging an older 2200mAh Zippy from 11.1v only took 1648mAh (the battery i was using in the above test).. puzzling

Steve

Interesting.

Wonder if different mfg LiPos have different "Efficiencies" during their charge cycle

[dumo01]

Another possibility is that a manufacturer has decided to label the batteries in terms of usable capacity to three volts rather than total capacity. It would make some amount of sense to do that but this far into the game I would hope they would have a fair amount of publicity regarding changing the game expectations.
Or more likely, there has been a change in the manufacturing process that gives them a higher charge density than they are labeled as, and as soon the manufacturer can figure out what it is and do it consistently, they will raise the price accordingly.

[Larry3215]

Steve,

There are a number of things that make your original question dificult to answer accurately.

1) Unless your using hi end batteries, the "rated capacity" is probably off - possibly a good bit.

The cheaper packs often get to a higher (dishonest) C rating by using for example a 3000 mahr pack but calling it a 2200 pack.

2) Even assuming they are being relatively honest - there is no "gold standard" that everyone follows for determining the exact capacity - so they can do it any way they want to get what ever number they want.

3)There is a certain amount of in-efficiency during both the charge process AND discharge process.

During the discharge for example, a certain amount of the batteries energy is lost as heat - before it gets to the Castle data logger. So there will always be some amount of discrepancy there.

The harder you push the packs, the greater the discrepancy between the logged discharge MAHRs and the actual total energy used - due to battery heating and the inefficiency of the chemical processes. Ive seen numbers anywhere from 20% to 40%

The charge process is also less than 100% efficient. I've seen numbers similar to discharge losses - anywhere from 20% to 30% losses during the charge process. Some of the variables there would be the charge rate, and how much balancing is going on during the charge as well as the pack Ir, the exact battery chemistry and the chargers internal losses.

Add in inaccuracies in the charger calibration to all that.

Its no wonder your numbers are not adding up

I personally dont like to ever let my batteries resting voltage go below 3.7. I prefer 3.75 - 3.8.

Thats a little conservative, but I dont buy cheep packs and I expect them to last several years and give several hundred flights.

[Larry3215]

Oh - most mfg's use the number of mahr put back in as the number to go by for how far to discharge.

That number is generally the safer, more conservative number to go by - as you pointed out.

Because of all the things I listed above, I prefer to go by resting voltage.

I use the Castle logger numbers to be sure Im not pushing too hard in flight.

[JetPlaneFlyer]

Larry,

Yes, I'm leaning toward resting voltage as being the most consistant and easy to measure criteria.

A possible source for innacuracy that I'd overlooked is the Castle data logger.. It's saying I'm pulling about 1650mAh to discharge a 2200mAh battery to 11.1V (resting)... I think it's possible that the logger is under reporting mAh pulled from the battery, which if true would explain the larger than expected discrepancy between what I'm measuring coming out vs. what I put back in. I'll check it by running a pack down with a wattmeter in place. If i use the same wattmeter as I used to check the charger then that should rule out errors between different instruments.
This test will have to wait a few days because I'm away visiting family...

Steve

[Larry3215]

Steve,

1650/.8=2062

All it takes is a 20% loss in the discharge/recharge cycle to account for your numbers.

Ive seen much higher reported losses.

[kyleservicetech]

Originally Posted by Larry3215

Steve,

1650/.8=2062

All it takes is a 20% loss in the discharge/recharge cycle to account for your numbers.

Ive seen much higher reported losses.

Hi Larry
Good to know, thanks for the info!