Testing for Amp Hours

[jeohearn]

I want to find out how many amp hours my battery has so that I can assure myself that I am not discharging too much during my commute. I am concerned that my ebikekit 12v 10ah LiFePO4 may no longer have 10 amp hours in it any more. My round trip commute uses 7 to 7.5 amp hours. I know the battery used to have about 10ah since once I forgot to charge and headed off to work the next morning and ran dry about 2/3 of the way to work. (My commute is mostly flat.)
How can I find get a more accurate measurement of how many amp hours my battery currently has? The only way I can think of is to hook up my simple watt meter (http://www.hobbyking.com/hobbyking/store/uh_viewitem.asp?idproduct=10080) and ride until the battery runs out and note the amp hours used. The catch 22 however is that when my battery runs out it will be 'off' and when it is 'off' the watt meter shuts off and (as far as I know) that info is forever gone. So, I would have to keep checking the meter reading (which is difficult because the meter is back with my battery on my rack) and hope my last check will be right before my battery runs out.
Are there other tools or ways to measure the amp hours of a battery?
Thanks everyone.

 

[neptronix]

Actually, you can power the watt meter off an external DC source, i believe it will accept 4.8-60v DC in. That will keep it on until your battery dies. Plug a 9 volt into it, i say.

http://boydhobby.com/Documents/Turnigy Watt Meter.pdf

The watt meter is not the most accurate piece of equipment though .. i know mine is off a tad.
Another option is a CBA, but those are spendy.

The watt meter has been out of stock at hobbyking for a long time unfortunately, but do check out this product, it should be equivalent, if not the same exact unit with a different sticker.

http://epbuddy.com/index.php?main_page=product_info&cPath=30_8&products_id=42

 

[John in CR]

Run it to cutoff, and then measure how much is put back in through the charger. All that requires is a multimeter to measure the current and a clock to measure the time. Current and voltage aren't constant, so factor that into the simple computation and take a number of current samples. Typical chargers are CC/CV (Constant Current then Constant Voltage with tapering current to the end).

 

[dogman dan]

I'd just ride to work and back, then maybe the next morning if saturday, ride around near the house.

Get an idea what around the block for 1/2 mile or whatever uses, then when you get cutoff you are near the house, and can estimate that bit since you last looked as you passed your house.

 

[theRealFury]

I think Neptornix has your best solution,

I have the same watt meter as well and it can be run off of an external power supply of 5-60v. so a 2s lipo (if your a RC guy) or a 9v smoke alarm type battery should do the trick (although the 9v battery wont last too long so check it soon after it cuts out). Once you have it working off the power source you can ride home from work on it to drain your usual 7-7.5AH out of it and then just go out for an extra ride after work around te block/street until it cuts out. Then you will know how much battery power you have left.

It wont hurt the battery going down to the LVC every once in a while so dont worry about it for testing purposes... better to find out how much power you have got than to be almost fully discharging your pack every day without knowing

 

[dogman dan]

Or you could look at your wattmeter more often. You'll know by the voltage and by the slow motor when it's getting ready to pop. Easy if you are not riding to a particular place, on busy roads. Find a low traffic place and ride around.

 

[dnmun]

to test it, first you need to buy a voltmeter so you can read voltages. you can use an ammeter on a separate voltmeter to measure the current, but it is best to use a watt meter like the watts up or turnigy watt meter or the cycle analyst. the turnigy is cheapest.

the ebikekit.com lifepo4 pack has the BMS folded inside the end cap where the charge plug is. you will need to expose it to get to the sense wire plug on the BMS so you can measure the voltages on each cell during the charge cycle.

charge it up to full voltage, and make your measurements on the cells while charging. pay attention to which cells are the highest, and how long it takes to cycle when the BMS shuts off the charge as the high cell goes to 3.9V. with that BMS, the cell has to drop back to 3.65V before the charge will resume, so it is really slow to balance and is the reason they do get so far out of balance with only short charge cycles.

we need to see those numbers to evaluate whether it is balancing.

then you discharge the pack into a load like a space heater, and keep track of the voltages on each cell as the pack discharges.

when the first cell drops below 2.1V then terminate the discharge and you can then read how much power was delivered by the pack on the wattmeter.

pretty simple actually, and it eliminates all the guessing which is worse than useless.

 

[dogman dan]

Good explanation of a proper load test. The only way to go, to diagnose a problem.

In a way, all he needs is a speedometer to find out what he's asking. It's not a guess. Ride to work and back, then keep riding. If your ride to work mileage is only 80% of the total, you are good. It's how I evaluated my battery for two years, till I got a CA. As long as I got the same total distance out of it, and it took the same charge time to fully recharge, I was happy.

Any changes, then time to get out the voltmeter and look at stuff. How could I have a problem if I get the full distance?

 

[amberwolf]

Another option is to get the Cycle Analyst, which saves it's data on shutdown, and has a speedometer as well. But it costs a lot more than the other options. I'd still recommend it over any of the others, for those two reasons, and the fact it's now got a large-screen version that is VERY readable under any condtions I've run across so far. (and it can be used up to at least 100V, and can be modified by it's maker ebikes.ca to run up to 200V).


I've also done what Neptronix suggests, to do capacity tests like this, and used a separate battery to run the WattsUp or Turnigy Watt Meter, while the regular battery ran down to cutoff.

 

[dogman dan]

Always nice to know your battery is doing well. I didn't mean to criticize the idea of a load test.

But it is possible to get good data on capacity with very simple methods. Doesn't really matter if it's measured in Wh, or miles or time. Provided you do a repeatable ride, in similar weather. You still get to find out, is it good, or is it going bad. Just looking at voltages of individual cell groups from time to time after a long ride is something I've done for years. I keep having no problem with my pings, so I don't do it so often any more.

 

[jeohearn]

Thanks everyone for all of your input. In the next few days I plan to use my watt meter to do the test (probably hook up a small battery so that it doesn't shut off when the battery runs to empty). Since I suspect I'm discharging close to 80% daily (which is why I am doing this test), I will probably begin charging at work -in addition to my charge at home- regardless of the results of this test. Then I can ride with peace of mind...and run an errand after work if need be.
Looks like a Cycle Analyst is a must have -I've been resisting it for a while because of the cost...but seems in the long run very worth while.
Is the Cycle Analyst fully functional with SLAs? My back up batteries are a set of 3x12v9ah (Powersonic) and I often wonder what % I am discharging when I use those on my commute (I do charge at work with those). I know I can't run those to empty in order to find out with out significant damage.

 

[amberwolf]

The CA doesnt' care what power source you use, as long as you don't exceed the 100V max rating.

Also, regarding voltage, it does have an auto-shut-off feature if input drops below 15V, by default, but you may lower this if you wish, with the caveat that the lower you go, the less time it has to save it's data, and the more likely that saved data/etc will be corrupted during shutdown. But that applies regardless of battery chemistry.

 

[jeohearn]

Well, I did hook up my watt meter to my ebikekit 36v 10ah LiFePO4, rode it till it was empty and found that it has 9ah in it. For one year plus I have commuted to and from work plus or minus 7.5ah per day before charging at home. From now on I will be charging at work too which will cut the discharge in half. Seems best for the battery not to discharge 7.5ah (and more if lots of headwind or if I run an errand after work) on a daily basis. Have procrastinated once again getting the CA. But the watt meter/drain the battery method worked well.
BTW, does having 9ah in this battery seem normal after fairly heavy use for one and a half years?

 

[whatever]

there is a thread which talks about a guy in china selling second hand a123 cells, there is a video and he has a meter that measures ahr of cells, find the thread, contact the guy, he will sell em

 

[theRealFury]

Well, I did hook up my watt meter to my ebikekit 36v 10ah LiFePO4, rode it till it was empty and found that it has 9ah in it. For one year plus I have commuted to and from work plus or minus 7.5ah per day before charging at home. From now on I will be charging at work too which will cut the discharge in half. Seems best for the battery not to discharge 7.5ah (and more if lots of headwind or if I run an errand after work) on a daily basis. Have procrastinated once again getting the CA. But the watt meter/drain the battery method worked well.
BTW, does having 9ah in this battery seem normal after fairly heavy use for one and a half years?

Yes that sounds about right to me, ~ 400 cycles over 1.5 years and its still got 90% of its capacity. I think most of the older LIFEPO4 is rated at either 70% or 80% capacity at 1000 cycles so your pack is doing ok

 

[jeohearn]

Thanks for the reassurance that my battery -currently at 90% capacity- is about where it should be. I will be playing it safe from now on by charging at work in addition to my charge at home.