Battery Load testing

11 posts / 0 new
Last post
zigalisr01
Offline
Last seen: 6 months 3 weeks ago
Joined: Thursday, September 16, 2010 - 14:47
Points: 59
Battery Load testing

I have read several posts, and reports in this group about battery testing. http://visforvoltage.org/forum/9723-instant-shelf-battery-capacity-measurement All the suggested methods are too complicated or involve $600.00 test instruments. I am looking for a simple way to test my battery pack. I have an XM-4000Li, it has 20 ThunderSky LFP40AHA, 3.3 volts, 40 Ah, I think they are LiFePo4 type. Anyway one exploded and I would like to test the others. The batteries are arranged as 5 packs, each pack has 4 batteries. They are connected in series for 66V total. I was reading in "Instant, off the shelf battery capacity measurement" thread the Load Tester, Advanced Tool Design Model ATD-5495 125 Amp Fixed Load Tester. I have a Cycle Analyst connected and a little before the explosion the voltage dropped to 50 volts. So, I can do a load test on each pack, 13.2 volts, but then how can I recharge the pack for another test? How can I charge a pack after it has been removed from the bike? I would like to remove the packs, charge them, and then do the test.

Spaceangel
Offline
Last seen: 6 months 2 days ago
Joined: Wednesday, July 15, 2009 - 15:49
Points: 500
Re: Battery Load testing

"I am looking for a simple way to test my battery pack. I have an XM-4000Li, it has 20 ThunderSky LFP40AHA, 3.3 volts, 40 Ah, I think they are LiFePo4 type. Anyway one exploded and I would like to test the others. The batteries are arranged as 5 packs, each pack has 4 batteries."

If one exploded after a voltage drop to 50 volts~ that would suggest a cell went reverse.

"They are connected in series for 66V total. I was reading in "Instant, off the shelf battery capacity measurement" thread the Load Tester, Advanced Tool Design Model ATD-5495 125 Amp Fixed Load Tester. I have a Cycle Analyst connected and a little before the explosion the voltage dropped to 50 volts. So, I can do a load test on each pack, 13.2 volts, but then how can I recharge the pack for another test? How can I charge a pack after it has been removed from the bike? I would like to remove the packs, charge them, and then do the test."

You can charge a 13.2 volt "twelve volt pack by using a good charger like Schumacher flooded, marine, AGM charger individually

KB1UKU

Mik
Mik's picture
Offline
Last seen: 8 years 7 months ago
Joined: Tuesday, December 11, 2007 - 15:27
Points: 3739
Re: Battery Load testing

How big was the "BANG"?

This information may be used entirely at your own risk.

There is always a way if there is no other way!

antiscab
Offline
Last seen: 1 year 4 months ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Battery Load testing

I have read several posts, and reports in this group about battery testing. http://visforvoltage.org/forum/9723-instant-shelf-battery-capacity-measurement All the suggested methods are too complicated or involve $600.00 test instruments. I am looking for a simple way to test my battery pack.

It depends on what exactly you are testing.

if all you want to know is whether a battery can supply a certain number of amps for a short period of time, than you can use an "instant" tester.
Thats all they are.

if you want to measure the capacity of the pack, don't use an "instant" tester.
It won't work (rather, it will lie to you)

The only way to measure capacity is to Ah count, and discharge all the way.

the simplest way to test your battery is to bypass the bad cell, fully charge the pack, and ride until it is discharged.
Your cycle analyst is already setup for that.

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

zigalisr01
Offline
Last seen: 6 months 3 weeks ago
Joined: Thursday, September 16, 2010 - 14:47
Points: 59
Re: Battery Load testing

The only way to measure capacity is to Ah count, and discharge all the way.

the simplest way to test your battery is to bypass the bad cell, fully charge the pack, and ride until it is discharged.
Your cycle analyst is already setup for that.

Yeah, that is what I was doing when the thing exploded. My battery pack is rated at 40 Ah, I was up to about 29 Ah according to Cycle Analyst. Oh and how loud was the BANG? Not very, it was like popping a small stick. The batteries are in there really tight, so the top pop off. The smell was horrible, after I rode home, you could not breath if you went into the garage. The scooter had to sit in the back yard for 2 days. The inside of the scooter, the battery box, was full of this gray spider web stuff. Some hanging out of the battery, yet still connected.

mf70
Offline
Last seen: 6 years 8 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Re: Battery Load testing

The thread you referenced was directed at finding out exactly what you are asking about, not instantaneous amp production. You would not have to have the rig on the road, and, assuming the inverter's cut-off was correct for lithium, would have some protection against reverse voltage.

The idea was to take the 12V input, then measure how much electrical work could be done before voltage dropped too far. The elements were:

  1. The inverter: this converts the 12V input to 110V AC, which in the "common currency" for the next several steps. In addition, and crucially, inverters have a fixed automatic low-voltage cutoff. This both provides a consistent measurement point and protects the batteries. HOWEVER, I DON'T KNOW WHETHER THE CUTOFF POINT IS CORRECT FOR Lithium CELLS.
  2. The Kill-a-Watt is the off the shelf capacity measurer. Readout is in killawatt-hours rather than Watt hours, but it is adequately accurate for this purpose.
  3. The 250 Watt heater is the load. Alternately, you could use incandescent lights to control the amp drain. The goal would be to drain at a rate similar to moderate road use.
  4. Finally, a standard VOM allows you to "hop around" and view the various voltages and amp draws involved in the setup.

As for safely charging each ~12V lithium pack, I have no idea whether a standard SLA charger would be appropriate.

Mark

mikemitbike
mikemitbike's picture
Offline
Last seen: 11 years 4 months ago
Joined: Saturday, September 20, 2008 - 13:07
Points: 310
Re: Battery Load testing

Hi, for the dischargetest you could use a CBA-unit. take a look at CBA
or at the manufacture website: West Mountain Radio
I use it for testing my NiMH cells (30 Ah- Vectrix and 78 Ah Pande elettra).

Greetings Mike

mf70
Offline
Last seen: 6 years 8 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Re: Battery Load testing

As was noticed on that thread, it seems to be 150 watt maximum. It lists 40A, but only at 3.7V! While that seems too limited, for the light e-scoots, it may be sufficient if you don't mind buying the dedicated unit. The output curve data would be quite useful.

I would want to test the battery capacity at discharge rates near road usage levels. My XB600 has an "open road" usage of about 650 Watts @ 48V. For one 12V piece of that, that would be about 162 Watts, so that device might do a pretty good job.

OTOH, I already owned the inverter, Kill-a-Watt, and 110V Load.

Mark

zigalisr01
Offline
Last seen: 6 months 3 weeks ago
Joined: Thursday, September 16, 2010 - 14:47
Points: 59
Re: Battery Load testing

I have put together an inexpensive test setup to do load testing. However, I am not sure what the results mean. I have the Advanced Tool Design Model ATD-5495 125 load tester. This does the "instant load test" referred to in some earlier posts. I have the battery packs removed from the scooter. They are 4 batteries each, 13.2 volts total. So, at a 125A load, which is 3 CA, the voltage drops from 13.2 to 11.0 and then returns to 13.1 when the load is removed. Is this normal? What type of voltage drop should I expect? What I noticed before the battery exploded was the voltage drop was much higher than normal, from 66 v to 52v. When I turn off the throttle, the voltage went up to 58 v but, never got back to 66 or 65 volts which for my scooter is normal. Does anybody know what the voltage drop should be for good/bad cells? I looked at the battery spec's and it shows the voltage drop based on Depth of Discharge. The DOD chart shows that when the battery charge gets close to 20% the voltage should drop to 3.1 volts. I think that the batteries are fully charged when I started this test. I am not sure how to tell when a cell is fully charged. I rely on the battery charger to turn on the green LED when it is ready. What does it mean it terms of volts and amps when the battery charger thinks it is done?

battery spec's: http://www.evolveelectrics.com/PDF/Thundersky/Thundersky%20LFP-40Ah.pdf

antiscab
Offline
Last seen: 1 year 4 months ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Battery Load testing

They are 4 batteries each, 13.2 volts total. So, at a 125A load, which is 3 CA, the voltage drops from 13.2 to 11.0 and then returns to 13.1 when the load is removed. Is this normal? What type of voltage drop should I expect?

yes thats completely normal.
how much sag to expect depends on the batteries production date, its history, temperature and SOC.

2.75v cell average at 3C is fine.

funny thing is, a cell thats sat on the shelf its whole life and still gives the full 40Ah will sag just the same as a cell thats been punnished and only has 25Ah remaining.

instant tests tell you nothing.

What I noticed before the battery exploded was the voltage drop was much higher than normal, from 66 v to 52v. When I turn off the throttle, the voltage went up to 58 v but, never got back to 66 or 65 volts which for my scooter is normal.

Don't use voltage as a replacement for an Ah counter.

my take is, the low cell got reversed (as in went below 0v).

discharge continued, resulting in both heat and pressure build up within the cell.

I would be curious as to why that cell was so far behind the rest......

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

antiscab
Offline
Last seen: 1 year 4 months ago
Joined: Saturday, July 7, 2007 - 23:55
Points: 1686
Re: Battery Load testing

As was noticed on that thread, it seems to be 150 watt maximum. It lists 40A, but only at 3.7V!

Thats perfect for this application.
though the practical limit is more like 25A @ 3.2v.

the lower discharge rate for the test doesn't matter so much, you don't encounter peukets until you are at 2-3C continuous.

without a BMS, discharging to completely flat when you don't know the capacity of the lowest cell necessitates discharing at the cell level anyway.

not doing that is why one cell was destroyed.

going below 1v is always bad with LiFePO4 (internal shorts develop).

so using a 10.5v cut off for 4 cells could well see: 3 x 3.2v + 0.9v.

especially if some cells are smaller than others.

also, if measuring the DC side, using a wattsup dc meter will tell you more (and its roughly the same price as the AC variant).

Matt

Daily Ride:
2007 Vectrix, modified with 42 x Thundersky 60Ah in July 2010. Done 194'000km

mf70
Offline
Last seen: 6 years 8 months ago
Joined: Friday, December 1, 2006 - 09:01
Points: 712
Re: Battery Load testing

You said:

I have put together an inexpensive test setup to do load testing. However, I am not sure what the results mean.

hee, hee, that's honest! Sorry to say, as others have commented, your results mean only that battery voltages sag under load. This is not unique news.

The thing you need to be examining is the battery capacity (AH) of EACH CELL. That will mean setting up a specified drain and timing how long each cell can supply that drain. If one cell can't match its neighbors, you can get toast.

To simplify the setup, you might do your pack in four ~12V sets. You will be able to spot which set is the weaker.

Mark

Log in or register to post comments


Who's online

There are currently 0 users online.

Who's new

  • eric01
  • Norberto
  • sarim
  • Edd
  • OlaOst

Support V is for Voltage