DIY Electric Car Forums banner

1 - 20 of 305 Posts

·
Registered
Joined
·
1,506 Posts
Discussion Starter #1 (Edited)
Ok this is the test thread for the battery I bought.

Right off the bat it appears I overcharged it for roughly 30 minutes. No swelling occurred but I'm attempting to try and understand the current voltage drop from my discharging. I believe I possibly damaged it permanently via overcharging, but so far I'm not sure if the 2.74v discharge at 10a right now is damage, needs recharging due to overcharging, or really the voltage drop



As you can see the charging for over 4 hours at 4.1v x 10 amps... this is the most I could constantly pull out of my DC supply without the voltage bouncing around.

Discharge charts will be up in a few hours once the testing is completed... Results are not looking so great at 0.25C as far as voltage drop

As I posted yesterday... the voltage drop is significant even when only draining at 10 amps of current out.... this is PRIOR to the overcharging sin I committed (battery was charged to about 20% capacity at this point)

 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #2 (Edited)
The battery is now discharged around 30AH... here's the chart (there was a 20A test that I stopped because he looked like it was going to be inaccurate)

Voltage has been consistent... voltage without a load is 3.44v.

At the moment I am simply attempting to drain the battery to recharge hoping the battery is made well enough 30 minutes of overcharging at 10a doesn't kill it. The battery's voltage was 4.17v before discharging began... it dropped very quickly to the 2.9v level during discharging

Ignore the jumps back up to rest voltage... these are me pausing it to adjust the terminals (as you can see the CBAII is not infinitely accurate voltage wise):






 

·
Registered
Joined
·
119 Posts
Right off the bat it appears I overcharged it for roughly 30 minutes. No swelling occurred but I'm attempting to try and understand the current voltage drop from my discharging. I believe I possibly damaged it permanently via overcharging,
Ack! :(

As you can see the charging for over 4 hours at 4.1v x 10 amps... this is the most I could constantly pull out of my DC supply without the voltage bouncing around.
I'm still learning but I see 3.44 Volts for just under 3 hours. But perhaps that is the battery voltage not the charger voltage? I'm not following. :confused:

As I posted yesterday... the voltage drop is significant even when only draining at 10 amps of current out.... this is PRIOR to the overcharging sin I committed (battery was charged to about 20% capacity at this point.
So it was crap before the overcharge? Well, 20% of capacity is pretty low, I would expect it to be down in the 2.something volt range at that state of charge.
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #4 (Edited)
Ack! :(


I'm still learning but I see 3.44 Volts for just under 3 hours. But perhaps that is the battery voltage not the charger voltage? I'm not following. :confused:
that is the battery's voltage, not the voltage from the DC supply... that's the battery's voltage changing over time (very slowly ... it ramps up to 4.2 volts very quickly once it reaches about 95%) ... I can't immediately access the saved chart I have where I charged it for over 250 minutes on the CBA... because I'm currently testing discharge I can't load it to show everyone... but just trust me it was over 4 hours :p


So it was crap before the overcharge? Well, 20% of capacity is pretty low, I would expect it to be down in the 2.something volt range at that state of charge.
I wouldn't say it's crap... I simply don't have enough data yet to make assumptions. The only thing I'm attempting to get out of this test since I overcharged it... is the total Amp-hours the battery has charged... if that ends up being really close to 90AH... then I will worry about voltage drop in following (much more closely monitored) charging sessions.

30 minutes of overcharging on Lifepo should not damage the battery... just so we're clear, but it might damage my abilities to test nominal voltage for this run. Voltage wise it seems OK that the nominal voltage is 3.4v and I'm discharging at 2.8-2.9v... it's not terrible... and to be expected until I get a good charging cycle out :mad:

One thing for sure is, lifepos are clearly not meant to be power dumped (at least this one)... voltage drop is substantial and sustained

Also it should be noted that this voltage drop could also possibly not even exist... the CBAII was not designed for high amperage discharges (has 12 gauge leads hard wired to it... they do get warm during testing). So far I've drained roughly 25 amp-hours out of the cell... it is staying at a very stable 2.9v (once I got the connections on there nicely)

Further tests will be able to verify... but applications where a constant 1-2C out of a pack is necessary probably shouldn't be looking to prismatic cells. Even at 20% capacity (and no overcharging) it was clear that had I dumped 90-150Amps out of this battery, it would have dropped below 2.5v

I almost forsee a "nominal discharge rate" of something like 0.3CA... which DOES vaguely follow many chinese lifepo suggestions
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #5
Ok here is the charging for over 4 hours (2 hours of charging prior to this wasn't saved). This is the voltage of the battery over time





This is the discharge... the battery has now drained 46.7AH from the initial start of testing. I intend to double the C rating discharge for the remainder. More to come.


 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #6
Ok, terminals were becoming too hot to sustain the draw for a significant amount of time. Will probably need to go and trim the wires down solder larger gauges and make hard post terminals.

1 foot of this 12 gauge wire they use isn't cutting it (I want to test up to 40A next charging)

Here is the voltage drop from 3.3v nominal at 20A of draw (holding steady at 2.61v):

 

·
Registered
Joined
·
3,219 Posts
...
One thing for sure is, lifepos are clearly not meant to be power dumped (at least this one)... voltage drop is substantial and sustained...
Hmmm... a 40A max load tester is not going to stress a 90Ah battery/cell enough to give useful data.

Furthermore, the whole point with LiFePO4 chemistry is to sacrifice some energy density for a much higher peak current capability, so these cells, if they truly are Lithium-Iron Phosphate chemistry, should be more than happy to dump power at a 3C or higher rate with little voltage sag.

I'm withholding judgment for now as it is entirely possible (but not likely) that you damaged the cell-under-test with that minor bit of overcharge (~ 5.6%), but for now it looks like the cells are NOT LiFePO4 chemistry or they are suffering from the "AA cell wrapped in a D cell body" syndrome... Not that a Chinese manufacturer would *ever* inflate their specs... :rolleyes:
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #8 (Edited)
Hmmm... a 40A max load tester is not going to stress a 90Ah battery/cell enough to give useful data.

Furthermore, the whole point with LiFePO4 chemistry is to sacrifice some energy density for a much higher peak current capability, so these cells, if they truly are Lithium-Iron Phosphate chemistry, should be more than happy to dump power at a 3C or higher rate with little voltage sag.

I'm withholding judgment for now as it is entirely possible (but not likely) that you damaged the cell-under-test with that minor bit of overcharge (~ 5.6%), but for now it looks like the cells are NOT LiFePO4 chemistry or they are suffering from the "AA cell wrapped in a D cell body" syndrome... Not that a Chinese manufacturer would *ever* inflate their specs... :rolleyes:
40A max load tester is fine for my application... Not to mention I can manually test anything further via a dummy load resistor (of which I have many to choose from).

It's entirely possible overcharging did something permanent, but very very unlikely unless the chemistry isn't LiFePO. 30 minutes is the max it was overcharging for... the time I took a shower, but it wasn't showing even 3.9v by the time I got in, so... again... unlikely it was even 30 minutes.

as you can tell though, initial testing prior to overcharging showed similar voltage sage proportions (and in this case proportions are all that matters).

Further tests will reveal what's going on. Overcharging lifepo's at all should only yield a overall drop in lifecycle as seen from a lower nominal voltage... that's really not what I measured at all (after letting the battery sit for 4-6 hours it still was at 4.15v prior to discharging).

Higher C rating isn't the reason I want lifepo btw... it's far lighter and less baby sitting than LiCoO2.

In general though, it's best to wait and see how subsequent cycles go I guess... reality is that there's no battery analyzer less than $4000 that can test 270amps and computerize the outputs... so I'm stuck with a max 0.5C test... hopefully with subsequent recharges it normalizes near the graphs the company posts (and Thundersky posts)

There's still the possibility that the leads from the CBAII are causing extra resistance via bad connections (though... still unlikely)... this would show as a voltage drop
 

·
Registered
Joined
·
1,858 Posts
You're planning on how many strings of these? Just curious how many amps you're planning to pull off your pack. My truck will draw over 400A with a fairly easy start until the rotor gets going and I know you've got in mind your super low cg creation with minimal weight but still, it will take some amps to get it moving, way more than 40 I'm sure.

Clue us in on your goals for this thread...
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #10 (Edited)
You're planning on how many strings of these? Just curious how many amps you're planning to pull off your pack. My truck will draw over 400A with a fairly easy start until the rotor gets going and I know you've got in mind your super low cg creation with minimal weight but still, it will take some amps to get it moving, way more than 40 I'm sure.

Clue us in on your goals for this thread...
No plans as of yet... the batteries are currently not useable from this test data under any circumstances.

I'm hoping the data for discharging is anomalous and after this charge cycle testing will normalize at something more acceptable (with a 1-2C range without much drop).

At the moment there's no way I'd purchase these batteries giving this kind of voltage drop... however the next test might be different.

I'm contemplating a parallel/series 60AH now actually... to cut down on loss of range through higher C pulls (even from really good lithiums) to do a 120AH x 106v or so set up... with whatever I choose. Hoping to not pull more than 60Amps off that pack sustained at highway or constant speeds (6kwh/hour tops).

My goals here are just to inform people as I test... that's it:p:p

I have no interest in pushing these batteries onto anyone... or even myself ;). I bought a sample to test as well as I could...
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #11 (Edited)
It appears that I was correct... the battery was merely giving false readings because of the overcharging...

recharging it is giving much more favorable results so far :D:D:D

0.25C discharge is showing 3.24v sustained (after dropping from 3.4v)



Here is the graph so far... I'm not prepared to say anything substantive yet... we'll see... but I shortened the leads by a lot and directly attached it to the terminals

Battery was only charged maybe 70-80%... we'll see where it goes.

Looking very positive.
 

·
Registered
Joined
·
43 Posts
I also wouldn't discount the fact that those leads might be hurting things a little bit also. If they are heating up, then there is energy being outputted there. You basically made them into a small heater, and heaters as we all know are pretty wattage hungry. As for the anodized terminals, shouldn't it be possible just to scratch it off in the right places with a knife before you put the leads on?
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #14
Ever hear of Nichrome wire? You can rig up a resistive load with it to test higher current discharges. Wind it around a non conductive form that won't burn or use glass bottles or such and play with that. Just a thought but not thought out! It's bed time...
Yeah I have a 4ohm 400w resistors though... 8 ohm 2 ohms etc from years ago (with heatsinks I can rig up a fan on).

Resistive wires (heating elements) is another alternative...
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #15 (Edited)
I also wouldn't discount the fact that those leads might be hurting things a little bit also. If they are heating up, then there is energy being outputted there. You basically made them into a small heater, and heaters as we all know are pretty wattage hungry. As for the anodized terminals, shouldn't it be possible just to scratch it off in the right places with a knife before you put the leads on?

I think between that and discharging the overcharge has helped... the load has stabilized dramatically this test run (far more than I predicted was possible)... at max testing output for the CBAII with this fuse in it (25A fuse... there's a 40A one that's swappable in)... with a bit of math I can use the CBAII to test any resistor load and figure out the output
 

·
Registered
Joined
·
4,958 Posts
Sorry, Technologic. I didn't realize you already started your testing thread. Better late than never.

12 gauge wire is certainly part of the problem. I got some funny readings with my tests as well even a slightly loose terminal that was barely warm to the touch was enough to drop the voltage by about 3%. If its actually producing heat off the wires, then there is likely to be considerable voltage loss there.

I think your results are comming in fairly clean now. Your discharge curve looks exactly like my spreadsheet plot for my 0.33C results.

If you can't get any heavier wire, even doubling up on the leads with the size you have will help.
 

·
Registered
Joined
·
308 Posts
Your leads are definitely your problem, it is obvious as you get different voltage drop every time you adjust the terminals. Also you should be using crimped on eylets bolted to the terminals, alligator clips won't do for this. Since the CBA does not have any sense leads I would oversize the wiring as much as possible to reduce the IR induced error.
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #18 (Edited)
Sorry, Technologic. I didn't realize you already started your testing thread. Better late than never.

12 gauge wire is certainly part of the problem. I got some funny readings with my tests as well even a slightly loose terminal that was barely warm to the touch was enough to drop the voltage by about 3%. If its actually producing heat off the wires, then there is likely to be considerable voltage loss there.
Yeah the 12 gauge certainly had me wondering as well. After shortening them and recharging the problem disappeared almost entirely.

Going to rig up some something like a heating element or if one of my dummy resistors works that will be more stable in order to give at least a 1C chart so we can proportionally figure out voltage sag when drawing it's rated AH capacity.

Here is the second charge... as you can see it's much more linear than the first:




Here is the Discharge at 23A (slightly over 0.25C) battery was only charged about 70% to full capacity.



As you can see it models for the first half or so at around 3.2v before dropping over somewhat sharply.

5 minutes after the test the voltage of the cell is 3.29v
 

·
Registered
Joined
·
1,506 Posts
Discussion Starter #19 (Edited)
Your leads are definitely your problem, it is obvious as you get different voltage drop every time you adjust the terminals. Also you should be using crimped on eylets bolted to the terminals, alligator clips won't do for this. Since the CBA does not have any sense leads I would oversize the wiring as much as possible to reduce the IR induced error.
Eyelets work fine for the negative, but the anodization makes it an issue for the positive... which caused a bit of an problem. I need more tools (I'm still in my apartment at my law school right now)

I wish I could swap out the wire gauge, but as I said the 12 gauge leads were hard wired into the CBAII so I'll have to look at it later and see if I can resolder them to the board or something with 10gauge or so...

For now what I did was tighten the bolt heads and flatten out large portions of the bare wire onto the post itself.

Eyelets just weren't doing it (but I might not have had properly thick ones here.

Suffice to say I got it to work much better... I'll keep trying to improve testing accuracy.
 
1 - 20 of 305 Posts
Top