I have learned the hard way the value of adding capacitors to the rails of my bread board circuits. I’m now in the habit of putting a 100nf capacitor between Vcc and Gnd whenever I’m building a new circuit particularly when I’m using 74LS/74HC series chips.
I’ve chosen this value based on vibes. But as I’ve copied more circuits online I see all sorts of caps chosen as “voltage smoother”. Sometimes polarized sometimes not. Some up to 10uf.
I’ve googled “guidelines for choosing a voltage smoothing capacitor” and all the blogs and forums seem to be mostly concerned with AC voltage. Not too much out there for DC.
This is my circuit but I’ve picked these values and types completely at random. What should I consider when making a more educated choice?
Your substitution of L7805 for the module has highlighted the real issue! If it really was a L7805 then C1 is recommended as 0.1uF for ‘transient response rather than stability’. So in that case it is not a smoothing capacitor on the rail (which is what it looks like) but a component directly related to the performance of the circuit. Transient response and stability are two common uses for capacitors.
But you are using a module and the 5V output is a rail and might benefit from additional smoothing. I use 0.1uF ceramic with low voltage circuits . Without extensive testing and circuit analysis the distribution is somewhat arbitrary, but is based on the devices used, their sensitivity to transients, the likelihood that they are inducing transients on the rail, and the length of traces and wires. Where smoothing is required because of how the load affects the rail the capacitors are usually placed close to the load. A 10uF (probably electrolytic) would be considered part of the PS filtering and would be placed close to the regulation.
Just to add a little bit here, I recommended reading the data sheet for each chip/device in your project. For example its common for a 0.1uF cap to be as close to each Vcc pin on the chips; but that does not mean all chips need one, or of the same value;
Sometimes caps are used to filter out unwanted frequencies and/or to stop them being introduced, as such the value will depend on the frequency you are trying to get rid of.
In power regulator circuits is not in common to have more then one capacitor as their could be a range of frequencies you are trying to filter out.
Hi Pix
Firstly the L7805 and “this” are 2 entirely different devices. As Jeff says the Data sheet for the L7805 suggests 0.1µF on the output. The sheet also suggests the 7805 does not like too much of a capacitive load so I would certainly NOT put an electrolytic here. In fact if it is driving a capacitive load, especially above 6V, a diode should be fitted across the 7805 to short this capacitance to ground in the event of a short to ground on the input. I would suggest you read this data sheet (all of it).
Regarding “this”. If you care to look at the schematic you will see that this device already has 7 10µF caps (total 70µF) installed in this position on the board.
Cheers Bob
I think this is the answer I was looking for.
Thanks Jeff.
I totally appreciate everyone’s thoughts.
In this particular case I’m trying to learn some guiding principles for the general cases rather than one specific case. Understanding the tradeoffs I’m making with my choices.
What I’m basically learning here is: if the value of the capacitor is going to affect the circuit I’ll have to measure, reading the datasheet for guidance where applicable.
Hi Pix
Jeffs notes are valid but if using an L78XX the electrolytic cap should be placed at the load. If driving a PWM controlled device like a motor speed controller and similar the inductance of the wire between the regulator and PWM switch is the bogey man here. Particularly this bit
The cap negates the effect the inductance has on the supply. The rapid switching of the load will cause a “charge pump” effect at the supply and if you have a longer length (a couple of metres or even about 1metre) you could be able to measure something like 6V or more at the output of the 7805. If you put any larger cap at the 7805 output you could enhance any charge pump effect and introduce problems that are not there normally. Don’t know for sure as I haven’t had reason to measure this, I just don’t do it.
I had a problem with this once while trying to set the trip point for a low battery shut down add on I had made for a friends golf buggy. It was impossible to measure the trip point with a variable supply as the 1metre or so length of the supply wiring caused the apparent voltage at the supply to rise by about 1.5V or more. The solution, Very short leads (150mm or thereabouts) or several hundreds of µF temporarily connected right at the controller input.
Cheers Bob
