I am having quite a bit of difficulty getting wires to connect in my schematic. The wires seem to be visually connected (with green circles), but ERC gives wire not connected errors. That said the given errors are intermittent. If, for example, I delete a wire and reconnect, that specific error may not show up again until the third running of ERC later (see the attached animated gif).
Based on looking through the forum for answers, I have tried the following with no success:
Made sure the symbols in question use the same grid as the schematic (50 mils for both).
Checked that everything is annotated.
Deleted my symbols from the schematic and recreated the symbol with new pins and the 50 mil grid.
Connecting pins with one long wire and (alternatively) with a specific wire from each pin.
Open your schematic, right click on any symbol, select âAlign Elements to Gridâ, then try your ERC again.
Also make sure you have 50 mil grid selected before aligning.
Those DRC errors look more like Power Input not driven by Power Output errors. To ensure you are reading the correct error click on the error marker and see which message is highlighted.
I was waiting for jmkâs answer, that that is not the problem.
There are some mis understandings here. From the ERC report in your screenshot I see: Error: Pins of type Power output and Power output are connected. And that is a completely different error from what you describe in the text. So apparently there is a fault in the way your SK6812 LED is designed. Both VDD and VSS pins must be Power Input types and not Power Output. To fix it, you have to change the pin type of this symbol in the Symbol Editor.
I do have a pretty decent sized 107cm monitor, but that is not really the point. This forum software scales images and when you click on it, it first scales bigger pictures first to your monitor size, and with the next click to the native resolution. With a small cropped part (See below), there is no scaling and you can read the text more easily.
I had the âwire not connectedâ error earlier in the same schematic and fixed it by matching the grid of the symbol to the grid of the schematic, so when I saw the other errors I neglected to pay attention to what the error actually said and assumed it was the same error I had received earlier. Lesson learned.
As to the actual error, it was caused by changing the negative pin on the default led symbol from âpower inputâ to âpower outputâ (which still seems way more intuitive to me because I think of current as having a direction). Looking at the matrix of permitted connections, by default KiCad calls connected power output pins errors. Iâm very new to schematic and KiCad, so if anybody has time to explain the reasoning behind that default configuration, I would love to hear it.
It says power not current. The criterion is whether the pin is a generator or consumer of power. On a net a consumer must have exactly one generator. More than one generator cannot coexist on a net. KiCad knows nothing about direction of current (but Spice does).
Thanks for the explanation. Iâm afraid I donât know enough about electronics to make sense of all of it, but I am going to do some googling to see if I can discern the crux of the distinction between power and current. Having one generator on a net makes total sense to me.
Nothing to do with Kicad but, itâs a choice between raking leaves or cleaning-house⌠@retiredfeline Short of having that coupler, this Movie âcat people 1942â may motivate you
(a favorite of mine )
It wonât hurt my feelings if this gets deletedâŚ
Have in mind that KiCad also donât knows electronic and learning electronic from ERC rules would be wrong idea. ERC is a simple rule checker that can help to find typical beginner errors. I donât like to have extra things at schematic (like not used pin markings, PWR_FLAG after each ferryte bead) just to satisfy ERC so I donât use it.
To 12V car bulb current is going in from accu + terminal through one wire and is going out through second wire back to accu (electrons run opposite). But that bulb is powered through two wires so power is going to bulb through its both pins.
If you have two 5V power sources each of them limited with current to 1A and you need 1.5A nothing very wrong will happen if you connect both this outputs together. The one with little higher voltage will reach its current limiting giving 1A and the lacking 0.5A will be sourced by second one. From electronic point of view - acceptable, for ERC - not.
Thatâs not a good example. In fact if it works itâs an exception to the rule. In general the power sources could conflict. E.g. 12V and 6V batteries (or regulators) shorted by accident. So ERC says: I think you may have a problem here. You can always silence the warning if you are sure about what you are doing.
I didnât found the better one
But I said that it is only âacceptableâ.
But truly if connection is by mistake it will be typically defferent sources connected together (like 5V and GND) what is clearly a bug so it is good ERC finds it.
The ERC is not smart. The ERC cannot tell you if a circuit will work. The ERC cannot tell you what a component/symbol does. It cannot even tell you if it is connected correctly.
All the ERC does is check to make sure all the pins on all the symbols are connected.
Basically, it checks to make sure you have connected everything. If something is not supposed to be connected, you have to tell the ERC or ignore the ERC.
If certain pins have certain definitions, they must be connected to other pins with other certain conditions or the ERC will say there is an error.
If this happens, first check your wiring, and if satisfied, you can change pin properties to satisfy ERC or you can tell ERC there is no problem or you can ignore ERC.
)
