LifePo4 voltagedrop, is this normal?

When fully charged, (48/52V, at 59/60V) I have to wait a few minutes before the voltage falls into the controllers operating range (any other workaround for this? Would be nice to grab my bike and just go when I'm in a hurry, I have the lcd3)


Anyways my question is, sooner than later the top charge of my battery goes down to about 52V, and when I ride opening the throttle all the way takes the voltage temporarily down to 42,41,40V, then back up to 52V when I let go, actually more like it goes back to 50 or 51V. But when it dips down to 40V the LVC kicks in..so I've lowered the LVC by 5% to prevent this happening and cutting my ride short for now..I know it's normal for the voltage reading to go down while accelerating, but how much is normal?
Thanks

Thats not normal at all. Hopefully someone with more bms experience can help out.

1. Assuming 16S, turn your charger down to 58.4V if possible, or buy a new charger. Most BMS will balance at 3.65V per cell.


2. 12V sag is excessive. See if one cell voltage is sagging worse than others indicating it is faulty or decrease your current draw. If the battery is made of the worst chinesium grade cells it may only be good for 10A.

LiFePo4 tends to lose power over time, rather than energy capacity. So it means that as they age, they sag more than they lose capacity.

12v is a lot for a 52v pack, which tells me that one or more of the following is true:

1. Your battery is too small for the application;
2. Your battery used cheap cells;
3. Your battery sat on a shelf fully charged for too long.

There isn't much you can do about it, except make the battery bigger so the voltage sags less.

As for burning off the surface charge, you can just hook up a light globe to the same connector as your bike has, and connect it for 5-10 seconds. Even if it doesn't light up, its still burning off that little bit of high voltage. You won't lose any noticeable capacity.

Agree with the above anwer. Lot of sag.... Seem like the internal resistance of your cells has gone to far up, explaining the sag....

Lifepo4 battery ? Where do you get the battery. How old is the battery? What are you using it with motor controller ? Yes you must answer all questions. A link to the battery.

If this is your 48v 30Ah pack in your other thread, we can do some ball park maths.

Resistance of battery = (open circuit voltage - loaded voltage) / Current.

You say in your other thread that you're drawing about 1750w at full throttle - Let's assume at about 42v - just before LVC kicks in. This means you are drawing about 42 amps (1750 / 42), rounded.

So 52 - 42 = 10v voltage sag.

Resistance = 10v / 42a = 0.238Ω, or 238 mΩ. Is this reasonable?

I had a google around for other 48v 30ah packs.

http://www.botepower.com/en/products/20160303023213.html

Promises <50mΩ - so your pack is nearly 5 times worse.

I'll put it in another context. I have a 150v LiFePo4 battery that is on its last legs. As in the guy who sold it to me (with a motorcycle) said that I should consider the price of the motorcycle to be without a battery. It was about 280mΩ. In case you didn't realise, the higher the voltage, the higher the internal resistance (since cells in series add to the resistance). So a battery that is more than 3 times bigger than yours, is considered "dead" at about the same resistance.

There is definitely something wrong with your battery.

Sunder said:

If this is your 48v 30Ah pack in your other thread, we can do some ball park maths.

Resistance of battery = (open circuit voltage - loaded voltage) / Current.

You say in your other thread that you're drawing about 1750w at full throttle - Let's assume at about 42v - just before LVC kicks in. This means you are drawing about 42 amps (1750 / 42), rounded.

So 52 - 42 = 10v voltage sag.

Resistance = 10v / 42a = 0.238Ω, or 238 mΩ. Is this reasonable?

I had a google around for other 48v 30ah packs.

http://www.botepower.com/en/products/20160303023213.html

Promises <50mΩ - so your pack is nearly 5 times worse.

I'll put it in another context. I have a 150v LiFePo4 battery that is on its last legs. As in the guy who sold it to me (with a motorcycle) said that I should consider the price of the motorcycle to be without a battery. It was about 280mΩ. In case you didn't realise, the higher the voltage, the higher the internal resistance (since cells in series add to the resistance). So a battery that is more than 3 times bigger than yours, is considered "dead" at about the same resistance.
There is definitely something wrong with your battery.

Click to expand...

Yes you are right! I am fully aware of that... At the time I had that battery, I was in what you describe as category 2 : "2. Your battery used cheap cells"...

My old pack was a piece of sh*#t made out of 10 years old LAPTOP CELLS (lithium cobalt chemistry not made for high amps).
238 milliohm is definitly NOT reasonable for my applications (pulling 30A on them is too much for 9P), but I wanted to learn about batteries and do the experiments anyways (rather than destroying good quality cells).

OLD PACK MATHS :
View attachment 5

NEW PACK MATHS :

Now that I've learned more...
I'm now planning on building a pack (not solderless) with quality used cells (mainly 2014 and 2015) Sony VTC4 with less than 100 cycles of life (NMC chemistry, high power cells). Each cell is in the 20 milliOhm ballpark.... While in my old laptop pack my cells were in the 100-200milliohm ballpark which is atrocious...

Here's my new cell test spreadsheet (ongoing testing)... I bought 290 sony VTC4 (each rated 2000mAh, 30 amps max, 12 mOhm AC 1kHz impedance).
View attachment Makita VTC4 cells from doctorbass.xlsx

I will solder on bussbars this time... In a manner that is similar to what doctorbass did back in the days with Sony V cells (see this pic) :


I will reuse the Makita Packs holders and slide my bussbars into cells tab and then solder quickly the tabs to bussbars (here's a preliminary testfit I did):
View attachment 1

Nota bene : 1750W was fresh off the charger : not 52V, but rather 58.8V... Anyhow, still atrocious resistance (voltage sag still 10V on the old battery 58.8 --> 48.8V). Also, I'm not sure my cheap watt meter shunt is well calibrated : It was this at the time https://www.amazon.ca/gp/product/B01EUISSRE/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1 ...

Back to you.... The cool thing about the motorcycle is that your battery is 120V.... So at the same 40A, voltage sag is STILL 10V.... So in your cas the RELATIVE SAG is less noticable (120V --> 110 V less noticable than 60V --> 50V) performance wise. BUT it still dissipates too much heat (P = V x I or P = R x I^2) which could harm the battery in the long run.

With a 14S10P of Sony VTC4, I can be in the "Less than 30 milliOhm pack" club, assuming neglectable resistance in connections... Far better than 250ish Laptop cell Experiment I made last year.

238 milliohm ballpark is a good estimation of what my old Laptop pack was Sunder....

See my first excel file above "Li-Ion.xls"... In the sheet called "Pairing 9 Sets", and inside the Excel Cell "W35"....

As you can see I did theoretical calculations before assembling my pack.... Which yielded a 216.585 milliOhms whole pack estimation (14S9P)...

I wasn't far from your 238 milliohms...

But my calculation didn't account for connection resistances (screws, springs, bussbars, wirering...), and also... 10 years old cells deteriorate very quickly ! (just look at the capacity lose after 200 km in my excel file!!)

Thanks for all the detailed info guy's. I've decided to sell that battery at a low price and wipe my hands clean of it.
Would a battery with these specs suffice, do you think?
Thanks

Pwnlord187 said:

Thanks for all the detailed info guy's. I've decided to sell that battery at a low price and wipe my hands clean of it.
Would a battery with these specs suffice, do you think?
Thanks

Click to expand...

A 20Ah pack rated for a maximum 30A discharge rate (only 1.5C MAX !!!! ). That' the typical cheap chinese low quality battery that might last you one year if your lucky. Honestly, I think a can do better discharge rating with scavenged laptop cells. :|

Get some proper cells or you will just waste your money and wont be happy.

Ok, thanks. Starting to wonder if maybe my battery was fine, and just didn't like my controller set at 50A output..Regardless, done some more homework, I might just get a 48V/20AH Li-ion that has 60A max output and 2000W BMS. Can get this pack at a good price too. Apparently lithium offers better speed at lower wattage, and with overall less voltage sag than LP4.. Any objections to this please let me know! (Otherwise might place the order) Thanks again!

Pwnlord187 said:

Ok, thanks. Starting to wonder if maybe my battery was fine, and just didn't like my controller set at 50A output..Regardless, done some more homework, I might just get a 48V/20AH Li-ion that has 60A max output and 2000W BMS. Can get this pack at a good price too. Apparently lithium offers better speed at lower wattage, and with overall less voltage sag than LP4.. Any objections to this please let me know! (Otherwise might place the order) Thanks again!

Click to expand...

A few words of advice :
1 - If the seller cannot tell you a reputable brand (Samsung, LG, Sony, Panasonic, Sanyo, Molicell) for the cells that make up your pack, the quality will not be good. DONT BUY IT ! What are the cells in you pack ?
2 - Know the exact cell model. Not just "it's a Samsung". You need to know brand and model. For exemple, Samsung INR18650-30Q is awsome (NMC chemistry, 3000 mAh capacity, 15 A current)-rated, but Samsung ICR18650-22F (LCO chemistry, 2200 mAh but just 4.4A current-rated) is not adequate for the majority of eBikes.
3 - Dont cheap out on the battery. You'll just end up being sad. Lithium cells are expensive. Almost as if "there is no such thing as getting an good deal on lithium batteries". Well not exactly, but if the price sounds too good to be true, it most probably is. If you want your battery to last 3-4 years buy quality cells. If you want to pay half for a battery that will last you3-6 months, well...
4 - The battery is just too important to just blindly trust the seller that swears it's a good battery... What are the cells ??? Can he tell you ? If he doesn't know, he doesn't deserve your business.
5 - Get to know the specs of these cells. Which one are you getting ? Here's a Quick ref sheet : https://endless-sphere.com/forums/viewtopic.php?f=14&t=26383&start=1400#p1250239
6 - Brands like Trustfire, Crapfire, Shitfire, [insert word here]-fire, etc, are rewraps of very low or fake cells. DO NOT BUY

When you say your 20Ah pack delivers 60A juice, it means your pack is rated at 3C (low-end, but not necessarly crap).
20Ah probably means between 6P and 10P.... That tells me are cells are capable of either 10A (quite acceptable) each or 6 amp each (quite on the low side for an eBike)

So the real question is : What are the cells in the pack you're interested in buying ??? ---> BRAND, MODEL NUMBER and from there we can find Chemistry type (NMC, NCA, LTO, LCO...), Capacity, Current rating.
Is you cannot answer this simple question. The pack is probably a scam. DO NOT BUY.

Pwnlord187 said:

Apparently lithium offers better speed at lower wattage, and with overall less voltage sag than LP4...

Click to expand...

Lower wattage is more gentle on the cells. But lower wattage usually means LESS speed, not better SPEED.
Yes at lower wattage you pull less amps, hence there is less sage since the internal resistance of the cell is a constant. That's ohms law : Voltage sag = Amperes pulled x Internal Resistance or delta V = R x I

was thinking one from this site, such as this one..
http://www.lifepo4.in/app ebike battery/48v 20ah lifepo4 app Electric bicycle ebike battery.html
with 2000W BMS