Capacitor used to smooth voltage?

[TGP37]

What is the most common way to smooth DC voltage in a car?

Using a capacitor wired across the alternator + and ground?

The fluctuation is only 0.3v under load and 0.5v at idle, nothing problematic. But I am wanting to regulate the voltage a little better for fine tuning.



What size capacitor is optimal (if appropriate) and can I use a resistor to further stabilize the voltage?

Or am I experiencing radio noise from ground looping or spark plugs?

I would like to keep the voltage variation no more then 0.1v

This image describes what I am thinking.......but I don't know what size capacitor and what size resistor to use.

[robertisaar]

yes, caps smooth out DC ripple. the resistor would cause the cap to charge and discharge slower.

what size cap? big. like, throw in a 1 farad cap, that should smooth out the voltage swing considerably.

main differences between a battery and a cap: speed of charge/discharge and capacity. caps suck for capacity, but can fully charge and discharge in ridiculously low times(assuming nothing is preventing them from charging/discharging at full speed), batteries have excellent capacity, but comparatively slow charge/discharge rates.

i made a capacitor bank out of 100 470uF caps, comes out to ~1/20th of a farad, should have awesome reaction time and makes quite a spark when bridging the terminals.

[TGP37]

Fantastic, thanks!

I have two 1,000uf caps on hand, IIRC that is 0.001F /ea. I just might need a bigger capacitor.....or leyden jar lol

I'm gonna try using a variable resistor to play with resistance levels and see if I can find a good value to use.

[robertisaar]

it looks like what you're getting at is building a low-pass filter. or at least, mostly. in a normal lowpass, all current goes through the resistor, not to the alt an accessories as well.

if you can figure out what frequency the ripple is happening at, i can simulate it to see what kind of values to start with.

assuming it's the alternator's regulator doing it, i believe the CS130D operates at a frequency of 400Hz. using a single 1000uF cap, you would need a resistor value of .398 ohms to form a lowpass filter to filter out the 400Hz ripple. with 2 1000uF caps, it's .199 ohms. since it's only feeding a cap, you can probably get away with a lower heat rated cap, but even with a pot, those values will be hard to get.



and even then, i'm not so sure 2000uF is going to produce that much of an effect.... but i'm not an EE, just a guy who constantly tries to learn from them.

[TGP37]

I can see a triangle wave in the voltage via hpt scanner. It appears to be 4-6 Hz but I'm sure there are many frequencies involved. Since I don't have an oscilloscope I'm probably going to build a diy capacitor board to filter and smooth a broad range of frequencies, if that is possible. I did find a diagram online, posted below, of a multiple capacitor circuit for broad ranges. But when I run it in a circuit simulation it gives me an open loop error.

You'll see the voltage difference isn't much at all. I'm thinking a host of frequencies are dashing along my power supply. Making the wideband reading a bit jumpy. Maybe I am looking at the wrong approach? How about using zener diode(s) approach?

This is what I logged with only 6 pid's loaded (best resolution possible).


And this is the circuit I found...

[robertisaar]

the diagram looks to be functional.... i wouldn't have any idea on what kind of frequencies it would filter, but i don't see any kind of flawed connections/etc.

i'm not so sure i would use zeners(or any other kind of diodes) for something like this...

[Superdave]

You need a real oscilloscope to see what's going on although there is always going to be a voltage variation in a system running on an alternator. 13.6-14.4V depending on RPM and load.


Your variation on the wideband looks normal, that's just the exhaust pulses passing by.

[TGP37]

You need a real oscilloscope to see what's going on although there is always going to be a voltage variation in a system running on an alternator. 13.6-14.4V depending on RPM and load.


Your variation on the wideband looks normal, that's just the exhaust pulses passing by.

Interesting about the exhaust pulses, I thought the voltage variation was partly the cause. That explains why it levels out at higher rpms.

I do need an oscilloscope, planning to buy one someday soon. I walked into radio shack to get a few blank circuit boards and asked if they had the scope. The guy looked at me strange and said what's that? Radio shack is falling apart over the years.

I know my power supply system is working normally with only 0.3v fluctuation at most. I do want to stabilize even further mostly for the electronics that don't regulate like the fuel pump, fans, ignition coils.........there is no kill like overkill