Trying to get max AC inverted output from 44W sony NPF li-ion batteries

[alex ezorsky-lie]

Hi, I'm a filmmaker and while many filmmakers are kinda obsessed with sony NPF batteries I am on the extreme end, trying my DIY best to make them do everything (I know they can't really but it's fun trying!)

Anyway the highest realistic capacity one can expect from something like this sony NPF battery is 44Wh at 7.4V about 6A max current (though many claim more). So I assumed if I put two in series and then put that 14V pair in parallel with another two I could get ~15V and 12A or 176Wh and a max 176W.

To stress test this I got a 500W AC inverter (extra to be safe). When I tested the batteries directly to the inverter I triggered the overvoltage on the inverter which is set at 15V and my fresh batteries were putting out 16V (so close). So then I bought this buck converter to bring it down. I set the max voltage to 15V and the max current to be as high as possible (12A) which hypothetically limits me to 180W which should be fine.

I ran this new 4NPFs -> buck converter -> AC inverter -> and plugged in a high powered 250W LED into this watt metered plug into the inverter. The LED light has a dimmer so I could test various watt draw. I started low and the light came on! I ramped up the watt draw until it hit 90W at which point I noticed the buck converter was showing 9A and the voltage was now dropping to 11.2V at which point the AC inverter's low voltage trigger went off and shut things down.

So I expected my janky setup to have some legit efficiency issues but 90W vs 180W is much worse than I thought. Even with the buck converter's %98 efficiency (180W x .98 = 174W) and the AC inverter's %86 efficiency (174W x .86 = 150W) I'm thinking I should be able to squeeze out at least another 50W.

So I have come to you fine folks to see if we can increase this NPF to AC efficiency. I guess I don't exactly know enough about battery science to know why the voltage drops so significantly and if that's the unfixable part of my system? Maybe these batteries just have an inevitable issue as having high internal resistance?

What do you think, are you unsurprised at my result? Do you think I've hit the ceiling with NPF>AC output? Can we do better?

Thanks so much!

 

[Harald Kapp]

I see a few issues here:

1. Putting batteries in series and parallel is possible, but raises issues with even distribution of current when discharging and especially when charging. Li-based batteries need to be balanced when charging, otherwise they can quickly deteriorate due to unequal charging.
2. It is quite normal that battery voltage drops with increasing output current. Imagine a small series resistor within the battery that drops more voltage when current increases (although the underlying effect is cause by the battery's chemistry, not a physical resistor). Nothing you can do about that but using more powerful batteries which do not drop so quickly the output voltage when loaded.
3. "Can we do better" Yes, I think so, definitely. The chain of conversions you built is horribly ineffective: From DC1 (battery) -> DC2 (buck) -> AC (inverter) -> DC3 (LEDs). Each conversion incurs losses. A better way is to go straight from the battery to the LEDs using a LED driver that can work from the battery as input and generate the constant current for the LEDs as output. Or maybe the videolights have a DC input (I have one that operates directly from batteries or a DC plug - way less than your 250 W, though)?

Another venue to circumvent the issue with high current and the voltage drop while keeping the inverter in case your LED has no DC input is this:

• put the 4 batteries in series.
• use a buck regulator to step the ~ 28 V battery voltage down to 12 V matching the inverter's input requirements
• use the inverter to generate the AC mains voltage for the LED

By putting all batteries in series you reduce the current required for the same wattage, thus you also reduce the voltage drop. -> edit: this is rubbish, see my next post.

 

[Harald Kapp]

I think I goofed - a bit.
I overlooked the matter of you putting two stacked batteries in parallel. While it is true that at twice the voltage (all 4 batteries in series) the current is 1/2 of that with only 2 batteries in series, in the latter case the doubled current is split in 1/2 for each leg of the parallel circuit. So each battery delivers the same current as in the full stack of 4 batteries. Some imbalance in current may occur due to differences in the battery state (charge state, internal resistance etc.)
The advantage of a 4 cell series connected stack is that it allows for much more voltage drop at the input of the step down regulator while still allowing the regulator to output 12 V.