Replace your car battery with capacitors!

[hayden55]

Have y'all tried the 16v Maxwell modules yet? I think they are only $150 assembled and self balancing. They have been used on the Ford Focus forums.

[serialk11r]

Originally Posted by hayden55

Have y'all tried the 16v Maxwell modules yet? I think they are only $150 assembled and self balancing. They have been used on the Ford Focus forums.

I would highly recommend against them (500F) for a light vehicle:

The leakage current goes up with capacity, but the ESR of even a small ultracap is way lower than a battery. After some point, you are wasting the power density of the ultracaps because you'll never need it, and paying extra for its poor energy density, while increasing leakage current.

I really don't think you should ever need more than 100 farads on a car, even if you're starting a Dodge Viper. 100 farads keeps the leakage current to a reasonable 1ma, while providing 7kJ of cranking energy. There's a reason those 500 farad packs are used on semi trucks!

Put it in parallel with any cheap battery to provide reserve capacity. The caps do the cranking only.

[teoman]

You are right. But coupling the caps to a battery also has voltage problems.

A lead acid sits at 13 ish volts. So you are essentially wasting delta 3V.

In my un investigated case, the capacitors are not sufficient to start the vehicle (i probably need to do a better connection to the capacitors). I am thinking of a small boost converter to top the caps to 16V for startup.

[hayden55]

Originally Posted by teoman

You are right. But coupling the caps to a battery also has voltage problems.

A lead acid sits at 13 ish volts. So you are essentially wasting delta 3V.

In my un investigated case, the capacitors are not sufficient to start the vehicle (i probably need to do a better connection to the capacitors). I am thinking of a small boost converter to top the caps to 16V for startup.

Didn't mazda solve this with programming in dedicated circuits/timing for the start stop regen portions?

[teoman]

I recall they had a higher voltage system. But I dont have a mazda.

[jakobnev]

I've been pondering on a hybrid setup like in the picture.

The idea is that instead of buying 4 small LiFePO cell to make a car battery, you buy one big one, put it in series with caps and let a voltage booster keep the total voltage constant, and high enough to decrease alternator load.

The caps are meant to be big enough to provide peak amps during cranking so the boost converter can be smaller and cheaper.

When the car is parked you can bring the single cell with you and recharge it in a nice warm place. (Just a LiFePO likes it)

By using just a single cell you can buy/carry a much bigger one that can easily handle the cranking amps, that way it will last longer and the voltage it provides let's you skip one cap in the system. And you save money on a BMS.

The diode at the bottom of the picture isn't meant to let you crank the engine without the battery, it's just there to keep your clock and ECU-memory happy.

[serialk11r]

Originally Posted by teoman

A lead acid sits at 13 ish volts. So you are essentially wasting delta 3V.

In my un investigated case, the capacitors are not sufficient to start the vehicle (i probably need to do a better connection to the capacitors). I am thinking of a small boost converter to top the caps to 16V for startup.

The capacitor's lifespan at maximum voltage is shorter. You lose some energy capacity but does that really matter for something that weighs under 1lb?

Leakage current means it's not optimal for long term storage anyhow. Mazda i-ELOOP uses a 24V cap bank and 24V alternator with a DC-DC converter to drop it down to 12V. There is a 12V battery to provide standby power as well. The purpose of the 24V is more about increasing alternator power and efficiency, and the capacitor is 25kJ (so most likely 80 farads, 10 800F caps in series ish).

If you're not using a mild hybrid system, you have no need for high power energy storage. You only need one second of high-power discharge to crank the engine, which a supplementary cap will provide very nicely when paired with any cheap battery (that provides reserve capacity).

[redpoint5]

Self-discharge is only a problem near the rated voltage. Back the voltage off just 0.2v per cap and there's practically no discharge.

[serialk11r]

Originally Posted by redpoint5

Self-discharge is only a problem near the rated voltage. Back the voltage off just 0.2v per cap and there's practically no discharge.

Er, I'm pretty sure self-discharge is proportional to voltage, so you still have substantial leakage current at 2.4V. The car's parasitic drain is still higher though.

One useful application for capacitors is cars with longer wiring runs to the battery. The extra resistance of that wiring can reduce starter power even if the battery has plenty of capacity. You can wire a capacitor bank in parallel to the alternator output (which should be very close to the starter and has a nice fat terminal you can tap into) to avoid routing all the current through long wire runs.

[redpoint5]

Originally Posted by serialk11r

Er, I'm pretty sure self-discharge is proportional to voltage, so you still have substantial leakage current at 2.4V. The car's parasitic drain is still higher though.

All of my caps are wired with LED balancing, so there's no easy way for me to measure self-discharge. That said, I thought self-discharge was non-linear with voltage? I could be wrong.

Either way, you're right that parasitic loads tend to exceed the self-discharge. My motorcycle can sit about 3 days before the voltage dips too low to start the bike.