Badwolf wrote: ↑Wed Oct 20, 2021 3:31 pm
Genuine question as I'm not an electronicsist: what's the frequency response like of a zener diode? Is it good enough to allow (say) a 6v zener to be placed across the 5V rail to clamp such spikes?
My naive physics background suspects once it's 'on' it takes some time to turn 'off' so would just stop everything working if it actually started clamping, plus if it could breakdown quickly enough to be of use here, presumably every PSU would have one already.
That was my thought some 25 odd years ago as well
In reality something like a 5.6V zener will start conducting around 5.2V. You can easily pull something like 100mA or a lot more. As you slowly increase the voltage you are basically getting into a few amps region.
If there are only going to be something like a couple spikes a second, then a zener might actually work. You would need to know the voltage and current to work out the wattages to find a suitable zener.
When you get into switchmode power supplies where those spikes can happen thousands of times a second.. You would most likely need a 25Watt zener to clamp and withstand that sort of abuse.
I think I experimented with all this and ended up with a 50V zener on a 5V line before it would stop catching on fire.. But of course that is rather pointless
Generally the correct way is with "transorbiant suppressors" or "transorbs" for short.They can be called many things like "Metal-oxide varistor" etc. These are basically voltage dependent resistors. In that at a 5V rated "varistor" will probably pull just a few milliamps. But as the voltage rises, the resistance drastically drops and thus clamps the spikes. Of course depending on the overall wattage per second of the spikes, you need a suitable wattage varistor. I cannot remember offhand, but I think generally they are a couple of watts.
But these are not really designed to clamp continuous voltage spikes like on badly operating power supplies etc. You should fix the power supply instead for that
