help with PSU filter design

As I am better with coding and digital electronics, I am having trouble figuring out the best way to filter power coming from my 12V 30A UPS.

I am using a camper converter/charger to maintain a large deep-cycle marine battery. the battery will be the primary power source, and the converter/charger will only kick on when needed. This UPS is to power several generalized circuits, BTW.

I will be using a AttoPilot Voltage and Current sensor to monitor current draw and battery voltage from the charger side of the circuit. this is so I can check voltage to see when the battery needs charging, and current draw to shut off the converter charger when the current draw levels are low enough.

I haven't scoped the converter/charger yet to see how much electrical noise it produces, but I suspect it is substantial due to it age (1988 ish).

I currently have a large cap of the car amplifier sort, that is listed as 500K uF. is this enough, you think?

I will scope it as soon as I find an extension cord to power it...

...Well, actually, it looks fairly clean with the cap installed. I attached a quick screen cap of the output and settings.

So I guess now I just have to design and build the Arduino monitoring component.

Thanks for all the help. :smiley: :smiley: (can I give myself Karma for answering my own question?)


I don’t see a need for a big cap at the battery, the capacity of the battery already exists and is even higher.

If ever, you can add caps at the 5V voltage regulator input, in addition to or instead of C2. Ceramic caps protect against fast spikes on the input line, while big (electrolytic…) caps protect against slow dropouts, eventually together with a diode to prevent current from flowing back into the battery.

Why does C3 say 500k uF Amp capacitor? The Amp bit makes no sense.
Batteries have a high impedance so I think you are right to put a capacitor there but I would just use a 100uF in parallel with a 0.1uF ceramic.

22pF is close to nothing on power lines, you better get another bulk of 100nF caps.
The diode is uncritical, as long as it supports >20V and enough forward current.

If you want to power more 5V devices, possibly with more linear (7805 type) regulators, an upstream step-down regulator from 12V to 7V may be fine, to reduce the power dissipated by the following 5V regulators.

An 7805 with 35V input will produce 6 times the heat of the attached 5V device. Even with 15V input it produces twice the heat of the attached 5V devices. That's why I'd reduce the input voltage as much as possible, just to keep the heat sinks small and to obtain the full output current. With a 7.5V input the regulator consumes only half the power of the attached devices, can eventually be mounted on a PCB without an additional heat sink. Even if there exist 78(HC)05 types or switching regulators of higher output currents, I'd go with distributed regulators, just to supply detached modules with "clean" power, and individual short circuit protection.

Personally I'm more confident in old school linear regulators, have no special experience with nowadays switching regulators. But I'd not hesitate to use upstream switching regulators, for their much better efficiency.

I meant it as in "distributed system", where multiple controllers in different places have to be powered. When lengthy lines to a central power source are involved, it's a good idea to add a voltage regulator to every controller.