Hello,
I am getting the ERC error of my input power pin not driven by output power pin but I have put a power flag on the output so I am unsure the issue? Thank you
Just to say that ERC is often a pain in the neck. OW!!! Many of us dismiss ERC errors (one at a time is safer) once we determine that there is no real problem.
One common sort of issue is when I use a filter resistor between a power input and an IC supply pin.
These days I do not think I am bothering to run ERC. But I am assuming the small risk as a result.
I donât see any PWR_FLAGs in your screenshot. However the marker points to pin 1 which is a Power Input according to the symbol definition, and RAW probably isnât connected to a Power Output, so a PWR_FLAG is needed on pin 1. And the SolderJumper will prevent the 3.3V line from being connected to the Power Output of pin 8.
Have a look at the FAQs - specifically this one - you need a âpower flagâ which is a âKiCad specialâ to sort this out. Once you understand it, it makes good sense but catches lots of people out.
Thank you. I thought I just had to put the power flag on output. I am trying to modify the arduino and atmega328p schematic. This is the part it is on:
Thank you. I put a flag on input as well tho I thought I wouldnât need to but it fixed the error. Putting the flag doesnât change the schematic right? Itâs just a kicad thing?
Add a POWER_FLAG to your RAW input wire - KiCad ERC needs to understand that this wire is the source of power for the MAX604.
Yup, just a KiCad thing. ERC isnât that clever but helps to avoid major errors. The power flags just show how a circuit is driven and donât alter functionality. They are there to try to catch some fundamental design errors but, unfortunately seem to catch out lots of people. If you run ERC, you can give it a bit of latitude as long as you understand why it is flagging an error. Generally, you need to add a flag to show where power is coming from if there is a power input pin connected to something else. If there is a fuse or a solder bridge, KiCad doesnât understand that that is still a continuous connection (it kinda assumes all fuses are blown & solder bridges, unabridged).
Like this right?
grut: What Iâve done with some of my schematics is pick a corner, and for each ground and voltage used, place a PWR_FLAG symbol with a short wire to the respective voltage symbol(s). ERC gets satisfied and thereâs no visual clutter on the main drawing.
Like so:
But the power flag should be used to indicate where the power comes from. Putting them into the corner is useless and just silences the warnings. In that case, I just would disable that ERC check altogether.
Iâve been using power flags as Meterman2026 indicates, assuming that the power flags are attached to the net and that the ânetâ result is the same no matter where they are located. Now I am questioning my reasoning.
If thereâs only the one connector and the power bus pins are labeled on it, then it should be considered self-explanatory, but when there are multiple connectors, then the flags go on the one âinputâ connector?
Itâs a matter of personal taste. Some place the flag close to where the power comes from, be that the output of a regulator, or a power input connector. Others place them in a separate part of the schematic, like shown above.
Both are fine - thereâs no need for dogmatic statements about which is more right. The former way shows the logical way where power is coming from, the latter does not. Both keep ERC happy. The key thing is that every net which has a power input also needs a power output or power flag to satisfy ERC.
To counter some of those who never use it (again fine, but I wouldnât recommend that), the key thing is that ERC, like DRC, will save your bacon if you use it from the start, and recognise that you may need to tweak things a bit (eg pin types, power flags, etc). This is especially true as you move to more complex designs where itâs easy while editing to make a small mistake that is easy to miss visually.