Boost converter with battery issues

[guitardenver]

I have a battery application. With an 11.1 V battery input, I need to maintain a 12 V, 2 A output as the battery discharges.

I bought a 1500 mAh, 11.1 V LiPo battery pack with a 35 C rating and I fully charged and balanced it.

https://www.amazon.com/FLOUREON-1500mAh-Battery-m4-fpv250-Shredder/dp/B06Y5VJ13Y/ref=sr_1_4?s=toys-and-games&ie=UTF8&qid=1539483472&sr=1-4&keywords=floureon+11.1v+3s+lipo+battery+1500


I also got an LM3481 SEPIC mode dev board.

The test setup

Battery feeds the input of the boost converter, the output of the boost converter goes to an electronic load.

Load is a 1.2 A constant current load.

The problem

The test starts off fine, but quickly becomes out of control. The battery input current keep going up and up, within 7 minutes and battery current reaches almost 4 A and rising. The output stays at 12 V at 1 A.

What I think is the issue

I think this is due to the internal resistance of the battery. Once the current starts to flow, the voltage drop across the internal resistance will cause the input voltage to drop, therefore the boost converter will have to draw more current to keep the output current going, then the resistance would drop even more voltage, causing more current to be drawn. This just keeps happening in a "runaway" unstable system.

Possible Solution

Regulate the input current. Put a max input current based on the current load. If the input current gets too high, control the boost converter via a DAC to lower the output voltage, and keep the input current at that max value.

But that max input current draw will depend on what the load is. If the load requires an input current of 500 mA, I would want the input current to limit to be around 500 mA, but if the load changes, so does that set point.

Questions

1) Is this a common problem using boost converters with batteries?

2) Did I just buy a bad battery?

3) What would work for this application?

4) This battery that I bought seems like it has really high internal resistance. Like 1.65 Ω (for 3 cells in series). I put on a 3 A load and the battery voltage drops very low!

Sorry I don't have exact numbers on that, I didn't write it down. What is a normal internal resistance for a LiPo battery? For a 35 C rated battery, there is no way it would be able to provide 35 times its capacity and have any voltage left over for the load. It is a 1500 mAh battery.

[PTR_1275]

Sounds like a bad battery.

You mention input current increasing, what’s the battery voltage at this time?

[guitardenver]

I connected to my charger to get some metrics.

When I put it in balancing mode. I get this reading for each cell
Cell 1 = 4.06
Cell 2 = 4.06
Cell 3 = 4.06

Then I put it in discharge mode.

Open (no load) voltage 12.06V

Then it goes to 500mA load
The voltage immediately goes down to 11.18V

While discharging at 500mA it also gives each cell voltage
Cell 1 = 3.48
Cell 2 = 4.03
Cell 3 = 4.03

Cell 1 is not balanced at all, even though the balancing mode reads that it is.

Should LiPo battery drop that much voltage at 500mA load?

12.06-11.18 = 880mV
880mV/500mA = 1.76 Ohms of internal resistance.

I tested both batteries. They came in a two pack. Could both be bad?
Is Floureon a bad brand?

[MiDi]

Bad battery, 0.3C is nothing for LiPos, you should go for refund or replacement.
If they are rated for 35C this gives ~50Amax.
For this pack if you short them out would give <7A with 1.76Ohm internal resistance.
This LiPos should have internal resistance in mOhm range to be rated for 35C.

[SiliconWizard]

Just a couple remarks.

If you look at a typical LiPo battery's discharge curve, the loaded battery cell voltage will drop rather quickly from the full charge voltage (4.2 V) to the nominal voltage (3.8 V) under significant load (with 1 A load on a 1.5 Ah battery I would expect maybe a couple minutes max, probably less). The discharge curve is not linear and the initial part is much steeper. Once it has approx. reached the nominal voltage, the drop is almost linear (and much less steep) until it gets close to the cut-off voltage, and then it will drop rapidly.

So in your case, a 3-cell battery, the fact you get ~11.1 V rather quickly after you load it doesn't surprise me a whole lot (although from your description, it seems to do so almost instantaneously, which is indeed fishy).

What happens after this initial drop is more interesting though. The battery voltage seems to drop too quickly for this load. Overall, it looks like it's certainly not a 35C battery as stated on it.

From Amazon's comments, most customers seem satisfied, but there are a few who say it's just crap. So there may be a definite quality issue.

As a side note, and as you unfortunately experienced (probably because of a quality issue on the batteries, unless there's some info we have missed), I'm warning you with this kind of setup: a battery followed by a step-up regulator. Whether the battery is dead or is just reaching its fully depleted state, with a step-up regulator, the input current will keep rising to keep up, up to levels which may be unsafe for the battery and its environment (heating issues and possibly worse). Even in the case of a fully operational battery, this is what would happen every time it approaches the fully discharged state. Not good. IMO, you should add a voltage supervisor that will shutdown the step-up regulator when the battery voltage is too low - don't just rely on the protection circuit on the battery, as you don't have much information on it if you buy off-the-shelf batteries targeted at RC toys, and as the cut-off voltage on it (if it has any) may be too low to prevent the battery from heating up in your case. In your case, a threshold of 9.5 V to 10 V should be adequate.

Just my 2 cents, that applies to primary batteries as well. Never use a step-up reg. without a means of shutting it down when the batt. voltage is too low. An 1.5 v AA alkaline battery may severely heat up when it reaches EOL without this protection. The worst case with primary batteries as far as I've experienced is with zinc-air batteries, that can litteraly explode if treated this way.

[guitardenver]

Thanks for the replies.

I will get another battery and see how it does. I defiantly have a way to cut off the battery completely from the input of the boost. I don't plan on using amazon batteries in the final product, but needed some quick cheap ones just for testing. Very unfortunate they are bad.

Could you explain why the following logic is wrong? You guys have kinda indicated there should be no issues using a boost converter with a battery input.

In a boost converter, if you want to maintain a 12V output at 2amps or 24W, you must draw at least that much power on the input. More due to efficiency.

So the lower the input voltage goes, the more input current must be drawn to maintain that 24W on the output.

Batteries have internal resistance. So once current starts to flow, the input voltage to the boost lowers due to the voltage drop on the battery resistance, which means more input current is needed to maintain the output. Drawing more current drops more voltage on the internal resistance even more, which makes for an even higher input current. This cycle just keeps going like a "runaway" system until there is not input voltage left on the battery or something fails.

Is this scenario incorrect? Even if I didn't have bad batteries, they should do this. What am I missing there?

[SiliconWizard]

You're correct as long as you don't draw more than the battery can handle, but as I said, you shouldn't just wait until the battery provides no current anymore. On its own. Some batteries' chemistries DON'T like too much current draw when they approach their fully depleted state and can go bonkers. That situation naturally happens with a step-up regulator.

That's why you should add provision to automatically shut down the step-up once the battery's voltage reaches a low-voltage threshold which will depend on the battery's chemistry and number of cells.

[ogden]

Is this scenario incorrect? Even if I didn't have bad batteries, they should do this. What am I missing there?

RC cars draw high current only briefly - during acceleration. This is why cheaters manufacturing "miracle 35C batteries" get away - users rarely notice how much current that piece of crap actually [does not] deliver.

You shall check RC hardware review journals/channels/blogs/forums, find popular and reliable battery brand, size battery for discharge rate that does not exceed 1C (3000 mAh?). Don't buy anything with "fire" in the name Also get battery protection board, search your favorite shop for "3s bms battery protection"

[Audioguru]

Maybe the overpriced Chinese battery is an old one that was on the shelf at Amazon for years. Amazon knows nothing about batteries.
I buy my Li-PO batteries at my local hobby store that knows all about how important the age of a Li-PO is and they sell so many that they are on a shelf for only a week. I will buy more with a reduced price on Black Friday next month.

[SiliconWizard]

Amazon knows nothing about batteries.

True. The only time I bought one at Amazon was a laptop battery. It was branded as genuine (and it turned out to be). It was DOA though. Not because it was defective or bad quality. Just because it had been stored for too long without being recharged. There's an internal protection circuit that will prevent you from even recharging such a battery (Lenovo) when it has gotten under a certain threshold. It becomes unusable. Amazon's model has definitely a problem with goods that can't be stored for too long and that are not marked as food.

Never buy a battery from Amazon.

[Audioguru]

Many of my name-brand Li-PO batteries bought at my local hobby store are 3 years old and have had hundreds of charge-discharge to 3V per cell cycles and storage charge at 3.7V per cell over winters. They are losing their capacity noticeably this year.

[sleemanj]

Batteries have internal resistance. So once current starts to flow, the input voltage to the boost lowers due to the voltage drop on the battery resistance, which means more input current is needed to maintain the output. Drawing more current drops more voltage on the internal resistance even more, which makes for an even higher input current. This cycle just keeps going like a "runaway" system until there is not input voltage left on the battery or something fails.

Is this scenario incorrect? Even if I didn't have bad batteries, they should do this. What am I missing there?

Do not think of it as a "current is drawn then voltage lowers then more current is drawn then  voltage lowers more", it is simultaneous not sequential, an equilibrium is reached.

Also, remember, wires and connections are resistors too, often unexpectedly high value ones.

[ejeffrey]

So the lower the input voltage goes, the more input current must be drawn to maintain that 24W on the output.

Batteries have internal resistance. So once current starts to flow, the input voltage to the boost lowers due to the voltage drop on the battery resistance, which means more input current is needed to maintain the output. Drawing more current drops more voltage on the internal resistance even more, which makes for an even higher input current. This cycle just keeps going like a "runaway" system until there is not input voltage left on the battery or something fails.

Is this scenario incorrect? Even if I didn't have bad batteries, they should do this. What am I missing there?

I'm not sure what you are asking.  Are you asking if that scenario can happen?  Yes, it can.  Are you asking if it will happen every time you hook a boost converter up to a battery? No, it won't.  It indicates that the battery in question is either discharged, improperly rated for the job, or worn out.  For a properly rated and functioning battery with a usable state of charge, the internal resistance is so tiny that increasing the current draw causes negligible voltage drop and increases the power delivered.  Once the power delivered equals the power demanded the battery and boost converter reach equilibrium. 

[guitardenver]

Turns out, all this was due to some shotty cables. They had some bad connections and had huge resistance at higher currents. They were prevent the the supply from sources the needed power.

I swapped them out and everything is right in the world. This was also causing a huge efficiency issue with the boost converter. Back to 89% efficiency.

Always have good cable connections.

Thank you for all the help and insights. I still learned something.

[MiDi]

I connected to my charger to get some metrics.

When I put it in balancing mode. I get this reading for each cell
Cell 1 = 4.06
Cell 2 = 4.06
Cell 3 = 4.06

Then I put it in discharge mode.

Open (no load) voltage 12.06V

Then it goes to 500mA load
The voltage immediately goes down to 11.18V

While discharging at 500mA it also gives each cell voltage
Cell 1 = 3.48
Cell 2 = 4.03
Cell 3 = 4.03

Cell 1 is not balanced at all, even though the balancing mode reads that it is.

Should LiPo battery drop that much voltage at 500mA load?

12.06-11.18 = 880mV
880mV/500mA = 1.76 Ohms of internal resistance.

I tested both batteries. They came in a two pack. Could both be bad?
Is Floureon a bad brand?

So the cell 1 voltage was measured through power lead?
Normally it is measured through the balancing connector to avoid the drops in power leads...