strange LiFePO4 behavior

[JayPrab]

I'm well into converting a classic air-cooled VW beetle to electric!
I've got it running on LiFePO4s (32S, 106V nominal voltage, 206Ah).
The system is 15KW @ 96V, so call it 150A peak, which it NEVER will see.
But the batteries are dipping even at just 30A draw!
The internal resistance per the battery data sheet is 0.2 MILLIohms... So that's 6.4 MILLIohms for the 32 batteries in series....
It should in theory only drop like 0.2V due to internal resistance at 30A....
HOWEVER...
When my car takes off, the battery voltage dips mightily, down to 50 to 60 V causing the controller to brown out and reset....
QUESTION:
Do EVs typically have voltage regulators between battery and motor controller?
Do you have any ideas on how to prevent this from happening?
Thanks for your help!!
Jay Prabhakar

[David Hess]

Do EVs typically have voltage regulators between battery and motor controller?

No, because a full power regulator would be non-trivial, expensive, and waste more power.

Do you have any ideas on how to prevent this from happening?

What was the source of the batteries?  Could the batteries be used or counterfeit?

[Ice-Tea]

What does the rest of your system look like? Cables? Busbars?

[Siwastaja]

Sounds like a bad connection or bad battery cell. In either case it means a huge amount of power is dissipated somewhere, therefore a lot of heat and significant risk of fire. You need to find the fault.

[3roomlab]

you should check if the 0.2m ohm is real
try checking 1 cell

I'm well into converting a classic air-cooled VW beetle to electric!
I've got it running on LiFePO4s (32S, 106V nominal voltage, 206Ah).
The system is 15KW @ 96V, so call it 150A peak, which it NEVER will see.
But the batteries are dipping even at just 30A draw!
The internal resistance per the battery data sheet is 0.2 MILLIohms... So that's 6.4 MILLIohms for the 32 batteries in series....
It should in theory only drop like 0.2V due to internal resistance at 30A....
HOWEVER...
When my car takes off, the battery voltage dips mightily, down to 50 to 60 V causing the controller to brown out and reset....
QUESTION:
Do EVs typically have voltage regulators between battery and motor controller?
Do you have any ideas on how to prevent this from happening?
Thanks for your help!!
Jay Prabhakar

[ejeffrey]

Start off by setting a ~100 mA load.  Then go over the pack with a multimeter looking for higher than expected voltage drops.   You should check every cell and every joint in the interconnect.

If that doesn't turn up anything, load test individual cells at a higher current. 

Ideally it is preferable to do this all in reverse.  Test each cell/module individually to make sure it is working and fully charged, then assemble into a pack and test the pack to make sure all the interconnects are low resistance, then connect to the motor controller.  A weak or discharged cell connected in series in a pack may end up reverse biased when the pack is under load.  This can cause more damage or cause a fire.  Likewise if your problem is a bad connection, a 1 ohm resistance like you have will cause the controller to brown out, which is fine.  A 0.2 ohm resistance would instead dissipate a couple hundred watts and risk damage or fire.

[dietert1]

Of course you need a power controller between battery and engine. It you turn on the engine with the car at rest it will be a short with current limited somewhere else, maybe by the cables. Depending on what kind of engine it is there should be some means of controlling engine power, e,g, a switching power converter. How do you control engine power?

Regards, Dieter

[Siwastaja]

Of course you need a power controller between battery and engine. It you turn on the engine with the car at rest it will be a short with current limited somewhere else, maybe by the cables. Depending on what kind of engine it is there should be some means of controlling engine power, e,g, a switching power converter. How do you control engine power?

If you read the opening post, it is obvious there is a motor controller there, and it even has a brown-out feature. Current and controller input voltage needs to be measured and reason for excessive drop found. These tests should be very short (<1 second, as quick as the OP can read their meters), not to heat up things.

[dietert1]

Of course you need a power controller between battery and engine. It you turn on the engine with the car at rest it will be a short with current limited somewhere else, maybe by the cables. Depending on what kind of engine it is there should be some means of controlling engine power, e,g, a switching power converter. How do you control engine power?

If you read the opening post, it is obvious there is a motor controller there, and it even has a brown-out feature. Current and controller input voltage needs to be measured and reason for excessive drop found. These tests should be very short (<1 second, as quick as the OP can read their meters), not to heat up things.

If you read my post, why don't you let the OP answer the question instead of adding more "general wisdom".

[David Hess]

Of course you need a power controller between battery and engine. It you turn on the engine with the car at rest it will be a short with current limited somewhere else, maybe by the cables. Depending on what kind of engine it is there should be some means of controlling engine power, e,g, a switching power converter. How do you control engine power?

The OP says he has a motor controller.  The question was does the existing motor controller require a separate regulator between it and the traction battery?  The motor controller should be designed to run directly off of the traction battery.

[IanB]

Ideally it is preferable to do this all in reverse.  Test each cell/module individually to make sure it is working and fully charged, then assemble into a pack and test the pack to make sure all the interconnects are low resistance, then connect to the motor controller.  A weak or discharged cell connected in series in a pack may end up reverse biased when the pack is under load.  This can cause more damage or cause a fire.  Likewise if your problem is a bad connection, a 1 ohm resistance like you have will cause the controller to brown out, which is fine.  A 0.2 ohm resistance would instead dissipate a couple hundred watts and risk damage or fire.

I think I would second this. There really is a science to assembling battery packs. Each cell should individually be fully charged and tested for electrical characteristics such as total capacity, transient power delivery and internal resistance, and any duds rejected. All cells going into a pack should, as far as possible, be electrically identical. Then the selected cells should be fully charged before assembly, so they start out in a consistent state of charge. (Obviously, working with fully charged cells is dangerous, so due care must be taken to avoid inadvertent shorts.)

[mzzj]

Likely bad connection or whole batch of total junk cells. Or you have fooled yourself and the battery is not fully charged in the first place. 
Depending on where you sourced the Lifepo4 cell the likelihood of getting junk is from 0.01% to 99%

And internal resistance tells only part of the story. Internal resistance is typically measured at 1khz. Substantially longer perioid of load currect and you start to have other effects like ion migration/overpotential.
IIRC high capacity Lifepo4 has thicker electrodes and worse ion migration/overpotential issues. Having said that 50% voltage drop at 30A current out of 208A battery is too much which brings us back to:

Likely bad connection or whole batch of total junk cells.
Depending on where you sourced the Lifepo4 cell the likelihood of getting junk is from 1% to 99%

[IanB]

I'm well into converting a classic air-cooled VW beetle to electric!
I've got it running on LiFePO4s (32S, 106V nominal voltage, 206Ah).
The system is 15KW @ 96V, so call it 150A peak, which it NEVER will see.
But the batteries are dipping even at just 30A draw!

[Edit: post deleted due to mistaken belief about cell sizes.]

However, I think this question still is appropriate:

If you test one of the cells individually, how does it perform? Does its capacity measure up to the label? And if you try to draw 30 A from it, how does it stand up?

[mzzj]

Wait. I hope you have lots of cells in parallel here? There is no way a single LiFePO4 cell has a capacity of 206 Ah!

Lifepo4 cell sizes are not limited to 206Ah

Most common ESS storage battery at the moment appears to be 280-320Ah per cell with 560-600+Ah per cell getting also more common. Winston has 1000Ah to 10000Ah cells but I’m not sure if these are just several cells in parallell without even internal connection.

EV cells tend to be bit smaller, BYD for example commonly using 138Ah cells.

https://en.winston-battery.com/static/upload/file/20220507/1651889071195617.pdf
Lots of terminals but what you’d want anyway if the battery can supply 10kA constant current and 200kA bursts. 

[IanB]

Wait. I hope you have lots of cells in parallel here? There is no way a single LiFePO4 cell has a capacity of 206 Ah!

Lifepo4 cell sizes are not limited to 206Ah

Most common ESS storage battery at the moment appears to be 280-320Ah per cell with 560-600+Ah per cell getting also more common. Winston has 1000Ah to 10000Ah cells but I’m not sure if these are just several cells in parallell without even internal connection.

EV cells tend to be bit smaller, BYD for example commonly using 138Ah cells.

https://en.winston-battery.com/static/upload/file/20220507/1651889071195617.pdf
Lots of terminals but what you’d want anyway if the battery can supply 10kA constant current and 200kA bursts. 

I see, those are big things! I didn't know that kind of cell/battery was a thing. Don't they need to have internal cooling or something?

[Siwastaja]

Quick Googling reveals that this weird-looking 206Ah number is now a typical available size for Chinese cells, so probably OP did not parallel anything, just a series string of single cells. It would be easier if the OP disclosed what cell they are using.

I see, those are big things! I didn't know that kind of cell/battery was a thing. Don't they need to have internal cooling or something?

Those Chinese LFP bricks have no considerations for cooling, or pretty much other type of engineering either. For some people, they work for years without complaints. For others, they burn the house down. But cooling is not an issue if you keep sensible charge/discharge currents. Efficiency of li-ion batteries approach 100% when current approaches zero.

[5U4GB]

What was the source of the batteries?  Could the batteries be used or counterfeit?

Alternatively, could the batteries have protection circuitry that prevents excessive current draw?