how to test the battery in a just-purchased used ebike?

[DaveTheWave]

My wife just bought a used Hebb Electro Glide 500 ebike.

http://www.bernsonev.com/Hebb_Electric_Bike_Electro_Glide_500_p/43234422.htm

It came with a battery described in the spec sheet as a 37v 10AH Lithium Ion. (EDIT: According to the manual it is a LiMnO type.) The battery does not seem to hold a charge for very long, but I have no idea what to expect from it. Is there a straightforward way to test it to see if it is in reasonable good shape, or repairable, or is just shot and I need to replace it? At the moment, having just been fully charged, it measures 41.8V across the terminals with no load.

The bike was in storage before we bought it. Does anything need to be done to recondition a battery in a situation like this?

I have a standard Vellman multimeter with a current range up to 10A. I was thinking perhaps I could put a resister of sufficient rating across the terminals and measure the current? For example, if I put a 144 ohm 10 watt resistor across the terminals, it should allow a current of .25 AMP to flow, and the battery should last 40 hours at that discharge rate before the voltage drops across the minimum. Would that work? And how would I figure out what the safe minimum voltage is?

Or do I need to buy something like the "watts up" meter to check it out?

It looks like I can open up the battery case and take some pictures of the internals if that would help.

 

[dnmun]

can you post up a picture of the battery charger so we can read what is printed on it?

 

[DaveTheWave]

dnmun - I just edited the original post to include pix of power supply, battery and an illustration from the User Manual which shows the specific type LiMnO plus a wiring diagram.

 

[dnmun]

ok, your battery is gonna be a 10S limn2o4 so it should charge up to the full 42V.

can you get access to the BMS or the inside of the battery so you can measure the individual cell voltages?

you wanna be able to measure the cell voltages while it is charging, and post them up here. count from the bottom of the pack which is B-, measure between adjacent pins on the BMS sense wire plug or directly from the top of the cells in the battery. 20V DC scale on the meter. measure to .01V and post them up so we can tell if the battery is balanced yet.

or just open it up and post up pictures if you get spooked about going forward.

 

[dogman dan]

Or, if not ready for that just put it on the charger for 24-7 for about a week. This will at least get the battery as balanced as possible. It's a good sign that it takes a charge, especially if it holds that voltage more than a few min. Once it has had a chance to balance, a slow low amps ride can give you a clue if you really have a problem. If that battery won't go 10 miles at 15 mph, it's really got problems. It should go 15-20 miles riding very slow. Fast riding, it could die in blocks if it's really got a dead cell.

It should be expected to lose some voltage overnight, but not to drop into much below 40v if it's still OK. A lot of overnight self discharge is definitely a bad sign. At 37v, it won't have much left at all, and 37v could be because just one or two cells already dropped to zero v.

Opening up the pack to put voltmeter probes on each cell, or group of parallel cells is inevitable. Old batteries have a value of zero dollars. But it could have some ok cells and some dead cells, if you can use some ok cells for another purpose.

You really cannot expect that thing to be very ok after a long storage period, unless they stored it right, recharging the pack to restore balance monthly or more. In long storage, the bms can actually kill one or more cells, with a slow drain that keeps the bms powered up.

Nothing to lose really, by opening er up now, and follow Dnmun's directions to see what cells are good, and if the bms is functional.

Lastly, you won't regret spending less than 20 bucks on a cheap wattmeter. It will be well worth having with a new battery.

 

[DaveTheWave]

Thanks for the advice.

The battery is still showing 41.8V this morning, after sitting overnight, so that's a good sign.

Any suggestions what kind of watt meter to get, and where to get it for less than $20. The only things I'm seeing cost more than that.

Maybe this one?

High Precision G.T. Power RC Watt Meter and Power Analyzer 130 Amps at $26 at amazon?
http://www.amazon.com/gp/product/B00C596UIA/

or the "Watt's Up"? now on sale at $49 on Amazon.

or are you referring to something that looks more like test equipment, not something that mounts on the bike?

 

[e-beach]

Hey DaveTheWave! Welcome to ES where we are always taking thing apart! :lol:

Hummm...Sounds like you can go either dogmans way or dnmun's way but you have to do one or the other first. If you want to really know what is going on with your batteries, ride the bike to low voltage cutoff, carefully open the pack (do not scratch or puncture or short any cells!!!) measure each cell voltage and then measure the cells while charging as dnmun suggested. Keep a running tally of where each cell is, in at-least one hour intervals. And of course, measure each cell once the charge is completed. This will tell you which cells are accepting a charge best.

Or you could do as dogman sugested and put it on your charger for a week so the battery's will be as balanced as that pack can get, go for a nice ride and see how far you can go. Take your cell phone in case you need the wive to collect you if you happen to be too far out to pedal your bike home once the battery pack has reached low voltage cutoff. By doing this you will find out your real range that battery pack can take you under real conditions. I call it riding on dead-reckoning. Then you can do what dnmun suggested to fully test you battery condition.

Around here the theorem volts x amp hours = watt hours is used quite a bit, but I have found that to be a conservative figure, of course I live in L.A. where the hills are low and short. You should have at least 360 watt hours to burn on a good battery pack. The only real way to know is by using a watt meter to see how many watts you have burned on your rides.

You can get an inexpensive watt meter like this one out of Freemont CA USA which should arrive to you quickly...
http://www.ebay.com/itm/Boat-RC-Hel...Control_Parts_Accessories&hash=item2c76417a49

Or you could go so far as purchasing a cycle analyst from ebikes.ca which is the gold standard of e-cycle computers. http://ebikes.ca/drainbrain.shtml

Or you could find something in-between. You could learn to ride on dead reckoning by just taking the bike out until it dies and calculate the distance, but a watt meter is handy because you will know just how many watts you burned. And if you set up the meter properly you can also meter how many watts you put back into the pack. Down the line, this will let you know just how the pack is getting weaker over time as batteries don't last forever.

Let us know how it goes.

 

[DaveTheWave]

thanks, e-beach

I am going to first do dogman's approach: let it charge for several days, put on a watt meter, and measure how much power I am actually getting. That will tell me the best next step.

Pardon the newbie question, but you mention "low voltage cutoff". Is that a point at which the BMS takes over and cuts off the battery completely and automatically, or is that a point which I have to note visually on the watt meter, and then stop riding at that point?

Since I am in San Francisco and the hills are steep, I'm assuming I'll hit low voltage cutoff when I'm going up something, even if the battery might actually be capable of still delivering enough juice on level terrain.

best,
David

 

[dnmun]

yes, hi currents cause voltage sag so it hits the LVC sooner. the LVC is triggered whenever any individual cell drops under the 2.7V level or so. it may be 3V on your BMS.

 

[e-beach]

........Pardon the newbie question, but you mention "low voltage cutoff". Is that a point at which the BMS takes over and cuts off the battery completely and automatically, or is that a point which I have to note visually on the watt meter, and then stop riding at that point? Since I am in San Francisco and the hills are steep, I'm assuming I'll hit low voltage cutoff when I'm going up something, even if the battery might actually be capable of still delivering enough juice on level terrain.

Yes you might have enough juice to power yourself on the flats and it will still cut under load on a steep hill.
Your BMS should take over when you drain you battery pack to the level set by the manufacture, or if you hit 20 amps draw (according to the link you posted in the OP). I am not sure where your BMS should cut out. Generally it is about 31.5 volts under load for 36v systems, but that is no guarantee. Try to contact the manufacture to find the specifics of your controller. Maybe they have a manual you can download.

Also, if you are using a cheap watt meter, consider using 10AWG stranded wire so you don't have any problems with extra added resistance that might choke off voltage to your motor. The Cycle Analyst uses a shunt so that is not a problem.

 

[DaveTheWave]

I just ordered that little power meter off of ebay and will report what I find. Thanks again everyone for your advice and suggestions.

 

[GSW999]

I built my own 36v 15ah battery from laptop cells , no BMS , have to charge as 2 sets of 5 but not a big problem , gets me 15 miles with pedal assist but I pedal all the time and don't use the motor on hills, it just helps to have in the wind n rain as I have decided to only use my car once a week when I have to collect my son.

You may have a mix of bad cells and good cells , Don't junk it whatever you do you may need to take apart and get the best cells out.

The main thing you have to be very careful about is shorting the battery at the top , Ive wasted so many sets of 5 in paralell, if you short one it seems to brutalise every one of those batteries in the parallel chain.

I need a watt meter so will be pleased to know how your test goes??

Gav

 

[dnmun]

you could have bot the other one on ebay for $12.52 i think. i actually like them more because you can adjust the calibration on them, but not on the turnigy.

http://www.ebay.com/sch/i.html?_sacat=0&_from=R40&_nkw=60V+100A+watt+meter&_sop=15

 

[DaveTheWave]

dnum - that's the one I got. I paid a couple bucks more to get it from someone who already has it state side rather than in Hong Kong. Supposedly I'll get it by the weekend.

I'm thinking of making a simple dummy load out of light bulbs to apply a consistent test load to my battery and determine its actual capacity. I would take three 12v 25W landscape bulbs like these http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=281130802424 and put them in series (to divide up a 36V source voltage evenly) and connect them to my 36V battery through the watt meter to monitor what's going on.

Here's my calculation. Do I have it right? A 12v 25 watt light bulb uses just around 2 amps. Three of them in series burns the same 2 amps but 75 watts. So connecting this bulb array to my 37V 10AH battery should drain the battery in five hours, which I'm guessing is the equivalent of a very gentle ride. if I wanted to simulate a faster ride I could use 3 50 watt 12V bulbs in series which would drain the battery in 2.5 hours.

Or I could buy an 18 ohm 75W power resistor if I could find one cheap.

The advantage of this dummy load over a test ride is that it's absolutely consistent, doesn't vary with hills or wind or how hard I'm peddling. Of course the disadvantage is that it's not a realistic test of how the battery performs under real world conditions with occasional much heavier loads.

Have I done the calculations right? Would this test do any damage to a battery (assuming I don't accidentally short it out)?

 

[DaveTheWave]

OK, my little power meter arrived yesterday from ebay and it seems to work.

Now I have to cut the wires on the bike and insert the meter into the circuit.

What kind of connectors are best to use that can handle the current? (36v 10AH)

Would this one work? It's designed for 10 gauge wire.

Thanks!

David

 

[dogman dan]

That could work, but many of us prefer something better. 4 mm bullet connectors from a RC hobby web store are popular, and another commonly used one is called an Anderson powerpole. I like powerwerks for them.

But you want to get going now this weekend I'm sure. For that, you can just make temporary connections with wire nuts found at any hardware store. Splice in the meter with those for now.

 

[DaveTheWave]

Cool, thanks! I got lots of wire nuts.

 

[e-beach]

Here's my calculation. Do I have it right? A 12v 25 watt light bulb uses just around 2 amps. Three of them in series burns the same 2 amps but 75 watts. So connecting this bulb array to my 37V 10AH battery should drain the battery in five hours, which I'm guessing is the equivalent of a very gentle ride. if I wanted to simulate a faster ride I could use 3 50 watt 12V bulbs in series which would drain the battery in 2.5 hours.
.........

Have I done the calculations right? Would this test do any damage to a battery (assuming I don't accidentally short it out)?

Real world riding will give you a real world result. Bench testing will give you a bench tested result that will tell you techniclly what you have in your battery, but not what you will get riding in the wind, up and down hills pedaling or not.

So if you feel the need to bench test, yea if you are burning at 75 watt hours your battery, if in new condition should be drained in about 5 hours, give or take.....but, I wouldn't drain it fully as you may cause damage to the cells. Draining it to 80% of new capacity is as far as I would go without the controller hooked up to it to trigger the lvc if you hit it.

If you want to burn a theoretical 80% of your capacity, then I have the math like this:
36v * 10ah = 360wh / 75wh = 4.8 hours * .80 = 3.84 hours (3 hours and 50 minutes) * 75w = 288w on your meter....give or take a bit.
So 3 hours 50 minutes and 288 watts burned at 75wh is the theoretical 80%.

Please check my math.

Also, your new watt meter should tell just what you are burning.

 

[DaveTheWave]

The bike's service manual says the following, for my 36v 10AH LiMn battery:

"LED [on the controller] blinks six times as a warning to charge the battery. This is a warning signal when
battery voltage drops below 30.5V. Bike is functioning but user needs to charge the battery ASAP.
The blinks will stop once the battery is charged up to 34V. If voltage is below 20V, red LED light will be
off and unable to activate controller."

That seems to imply that the controller doesn't impose Low Voltage Cutout until the voltage drops to 20. Does that make sense? What should the LVC be on a 10S limn2o4?

 

[e-beach]

.......That seems to imply that the controller doesn't impose Low Voltage Cutout until the voltage drops to 20. Does that make sense? What should the LVC be on a 10S limn2o4?

This from wikipedia..

During discharge on load, the load has to be removed as soon as the voltage drops below approximately 3.0 V per cell (used in a series combination), or else the battery will subsequently no longer accept a full charge and may experience problems holding voltage under load.

Here is the link, it is worth your read. It's not long.
http://en.wikipedia.org/wiki/Lithium_polymer_battery

The quick answer is that if your batteries to hit 2v each under load then you have killed your battery pack. 2v * 10s = 20v.
So if the manual says 30.5v (3.05v per cell at 10s) is the farthest you should draw from that pack under load, don't go any farther then that because as you described it, the controller does not have a low voltage cutoff and you could ruin what is left if it's life.

 

[DaveTheWave]

Thank you. That is indeed interesting (and surprising.) What a crappy design not to put a LVC at 30V into the controller! I guess I will test this out and see if it they really were that careless. Maybe the manual is wrong. If so I will need to get another controller for it.

 

[e-beach]

...... If so I will need to get another controller for it.

Or when you feel the need to upgrade your battery pack just get one with a BMS or PCM. They have a lvc. Otherwise just watch the lcd's and make sure they are working.