No, I do not know why you believe diodes are somehow backwards. Unlike other polar components, diodes, including zeners, are intended to operate with both forward and reverse voltages within their rated limits. Diodes have many applications and I don’t know what you consider most common. When used in a rectifier you would measure the positive voltage on the cathodes. When used as clamping diodes the positive most voltage would be measured on the cathode during normal operation. And I assume you are referring to the typical shunt regular application for a zener when you talk of them being “upside-down”, in which case you would measure the positive most voltage on the cathode. All diodes have a reverse breakdown voltage but most diodes are not intended to be operated at or above this voltage and are likely to be damaged if they are. Zeners function just like regular diodes with the exception that they are intended to operate in the region of their rated breakdown voltage, as long as they also remain within their rated current. Zeners also have a much better controlled breakdown voltage.
The idea that the anode is a “positive” terminal only holds true for LEDs and only because they only emit light when forward biased.
Edit: Part of the problem is probably due to people like Sparkfun publishing information like this.
Not to muddy the waters, but I will… It is valid to connect an LED in reversed bias configuration. LEDs will act as light sensors in that orientation. I first heard about this from Forrest Mims’ write up in Popular Electronics in the 70’s. Not when it was published, but back-issues. While I was alive in the 70’s, I was too young for electronics then. SparkFun has also done a write up and video on this. Yes, Sean should have said “reverse bias” instead of “reverse voltage”.
One reason for calling the anode of a diode “plus” (and conversely the cathode as “minus”) is because of the doping of the ends of the silicon die. The anode is P-Doped (positive doped) and the cathode is N-Doped (negative doped).
That is true, I did not intend to imply that it wasn’t valid, it’s just that LEDs are typically only used for one purpose and they only serve that purpose when forward biased. In fact LEDs behave just like any other diode although with different voltage characteristics.
So too will any glass encapsulated diode such as the 1N914 but their currents are much smaller than that of most LEDs.
You are of course correct and that is a valid argument against my statement that diodes aren’t “polar”. I have probably opened myself up to criticism with that statement. Strictly speaking the doping of the semiconductor material does give it “poles” thereby making it “polar”. The main point I was trying to make is that in terms of the voltage that can be applied to it, a diode will function correctly with either polarity. It will allow current to flow with a voltage polarity that forward biases it’s pn junction, or it will block current flow with a voltage polarity that reverse biases the junction. Essentially turning either on or off depending on the polarity of the applied voltage.
I also must take blame for taking the conversation to super pedantic levels. In my (relatively limited) defense, I did admit to wilfully muddying the waters.
Me thinks we have really wandered far away from the topic at hand.
Welp, when one does not have the polarity correct for a Light Emitting Diode it does not illuminate; positive voltage need be applied to Pin 2 (unlike every other device where Pin 1 is positive).
Welp, when one does not have the polarity correct for a electrolytic capacitor it may very well explode; positive voltage need be applied to Pin 1 (just like every other component(except the oddball above)).
I’ve been an electronics nerd since high school, and a professional technician since the age of 18. It wasn’t until I started to use KiCad that simple parts required Pin numbers.
And, apparently recently, it has been decided that it is a good idea to give “Tabs” Pin numbers.
@1.21Gigawatts My honest personal experience of the issue.
Yes, I did mention that earlier. But applying reverse voltage does not destroy it. (See tri-color LEDs)
Electrolytic capacitors are not being discussed here, diodes are.
Why do you keep mentioning your experience? I don’t keep mentioning my 40+ years of experience as if it somehow trumps yours. It’s irrelevant to the discussion.
But you did, and I fully expected that you would ridicule my experience and state that yours was bigger than mine; you do it every time.
Nope, Pin 1 numbering in relationship to polarity of parts that normally don’t have a Pin number assigned on the DataSheet; I don’t care what part it is.
KiCad requires that Pin 1 of a diode be assigned to the cathode of the part for expected functionality of the device to carry from the schematic to the PCB; and that is in fact opposite of the majority of other simple components.
I do not know how old you are or how long ago “high school” was for you, so I wouldn’t know if my 40 years exceeds yours or not.
Really, got an example?
They may not appear on every datasheet but they are standard in the industry. The industry also has a long standing tradition of identifying the cathode of a diode.
KiCad “requires” no such thing. The libraries are created that way because, and I repeat, it is industry standard. You are free to create your libraries how ever you like, KiCad won’t mind.
And your point? There was an incident in the past that caused diode symbols not to match the footprints. That was a library issue, not a KiCad requirement.
And as you can see by reading posts in that topic, KiCad won’t mind if you name the pins ‘A’, ‘C’, or ‘A’, ‘K’, instead of ‘2’, ‘1’ so long as your footprint library matches.
What I am certain of is that we have derailed this topic quite enough and your last three posts have not added anything to the discussion, you’re just out to prove me wrong, about anything it seems.
Second: It is not KiCad requirement. You can name pins differently.
Third: In my typical design I have about 10 transils and 1 LED. Transils cathode to work as expected need be connected to positive. So thanks to that (thinking like you, I have never thought that way) I have only 1 diode opposite instead of 10.
Fouth: If you have something against discuss rather with ONSemi than with KiCad.
Fifth: Take 10uF/25V electrolityc capacitor. Put +10V at its pin 1 and 0V at pin 2 - nothing happened. Put +10V at pin 2 and 0V at pin 1 - device damaged.
Take diode. Put +10V at its pin 1 and 0V at its pin 2 - nothing happened. Put +10V at pin 2 and 0V at pin 1 - device damaged.
For me both bahve similar. If you invert diode pins numbers than they will behave opoositelly.
If you have voltage drop of 10V across the diode then it is no longer a diode.
My main objection here is to the idea that the cathode of a diode being pin 1 is somehow “backwards” and that zeners are an exception. The industry standard is and has been for a long time to identify the cathode of a diode. On the other hand, pin 1 of an electrolytic is the positive pin, but it is often the negative pin that is identified on the component, yet no one is calling that backwards.
Even if you want to associate pin 1 with the “positive” pin, when it comes to the most common uses of diodes the cathode is usually the most positive pin. The anode can generally never be more than 0.5V - 1V more positive than the cathode. (With the exception of special purpose diodes and LEDs)
Damaged diode from electronic point of view is no longer a diode but if someone asked me what it was I had in my hand I would rather said ‘I have a damaged diode’ then ‘I have no diode’
I have never thought of Zeners as being exception - they also have their cathode marked and numered as 1.
In all switching circuits cathode is more positive pin than anode so when they decided to mark positive pin as squere then cathode is better candidate then anode.
I’m happy that I am using SMD elements and the question square or round is not applied
SparkFun has also done a 
In my (relatively limited) defense, I did admit to wilfully muddying the waters.