I'm trying to think of ways I can protect my arduino from my powersupply malfunctioning and causing a surge in output voltage due to lightning strike. I do not mean a direct lighting strike.
I read here,
After reading im not sure what zener diode I should use. My output voltage is 14.0vdc. The input can tolerate 20v. I have 10 amps available current.
To be clear. Im trying to filter surges on the 14vdc. This 14v runs to a buck step-down that runs the arduino
Google: TVS suppression circuit
I read a little bit about them prior and I came across this,
Fro. That article it dont sound like a good idea
Those sound like different things. Either your power supply malfunctions, causing a voltage transient. Or there's an EMI spike due to lightning. They require different kinds of protection. The former is in practice not a huge issue since it's extremely rare for SMPS units to fail in a high-voltage situation. It does occasionally happen that an SMPS fails and takes devices connected to it with it, but surprisingly infrequently. I never really bother protecting my circuits against this apart from designing appropriate fuses into each board connected to the power source. Usually polyfuses btw.
Protecting against lightning related issues logically brings you onto a TVS path. Given the wiring situation in your cash-crop grow-op it would be wise to adopt something like that, although I'm not sure if it will be entirely fail-safe.
But I hear TVS may burn up too fast. If i use Thyristor I most likely just have to change a fuse.
I never mentioned anything about a smps And actually I think its not a smps. This psu uses a large ac transformer. But im not 100% sure because it is adjustable after all.
It will most likely overvolt if hit by a surge. A fuse isn't going to protect much on its own nd the TVS may not even blow the fuse because it could burn up before the fuse blows?
Maybe I could use a tvs before the crowbar to filter the transients so I'm not blowing fuses when the psu kicks on. Then for any long surge protect im going to crowbar circuit after the tvs. Does this sound logical?
Again, determine the problem you're trying to solve and then select the proper mitigation strategy. So far it seems you're mostly googling around and finding fairly random solutions to a variety of problems. There's no focus in your approach yet.
There's indeed a 0.1% chance you got yourself a linear power supply that somehow was left on the shelf since 1985.
So you don't know what kind of power supply you have, but you do know its failure mode?
Or after it. Doesn't make a difference after all. A TVS diode protects a line. Doesn't matter all that much where on that line it's placed.
???
Not really, no.
Again, what are the specific failure modes you are trying to hedge against, where do they originate from and what are the requirements on the protection, i.e. what is the fault margin on the side of the modules and components you're trying to protect?
So far we've got a fear of your power supply failing and outputting a high voltage but without further specification or a clear reasoning why this might be the case, let alone some insight into what kind of protection the power supply itself might have against this.
Seriously, if you are concerned about the power supply you're using, get a unit from a reputable manufacturer with the necessary safety systems built into it.
Hi,
What is your power supply?
Can you post link to specs/data?
Thanks.. Tom... 
here is what im using now,
here's what i will have next week but its only 5 amps,
I think it would make a difference where the TVS is placed. If before the crowbar circuit then the tvs could filter transients and short spikes that would otherwise falsely trip the crowbar circuit. correct me if im wrong.
Both of the products you quoted are workbench power supplies not intended to power finished projects. They do fine at that BTW, but it's massive overkill.
They're both SMPS units btw. The main condition under which they will supply too high a voltage is if you mistakenly set it too high. With a fixed power supply this problem isn't there.
Show me how that would work given the fact that in a crowbar circuit the only thing in series with the load is the fuse. That's the whole idea behind it. Hence, it doesn't matter where you place the TVS diode; before or after the fuse, before or after the junction of the voltage reference, before or after the junction of the thyristor, etc. etc.
Would the TVS not be parallel to the supply circuit? therefor the surge/transient would meet the TVS that's in parallel first and it would block the transient before it arrives at the Zener diode in the the crowbar circuit?
if switch on psu that set at 14v, the transient is lets say 20v. it arrives at the 18V TVS and is suppressed before it gets a chance to un-intentionally- accidentally trip the crowbar.
Yes, but you're overseeing several things, including, once again, what a TVS is intended for, which is NOT protection of a surge voltage from the power supply. A TVS forms a low impedance path for a surge on the line to the negative or the positive pole of the power supply. So a TVS can not protect against a surge from the power supply itself.
All this is very clear if you take a glance at the datasheet of a random TVS diode where the application circuit is sketched and the intended purpose is explained.
Furthermore, the value of a TVS diode is in that it is generally faster than a crowbar circuit.
I understand it is a Transient Voltage Suppression diode. I understand that it creates a low impedance path once the breakdown voltage is met it will quickly clamp but only for a short time. How are you suppose to prevent from inadvertently tripping the crowbar from transients?
Yes, and to what...?
Yes, to what..?
Just checking if you really do understand or you just know which words to use in this context.
Transients are not created equally, and of course you can design the crowbar in such a way that it will only trip under certain conditions.
A TVS is typically used for a very short, very high-voltage but relatively low-energy transient. A crowbar is generally used for a longer-lasting, low-to-moderate overvoltage situation with a lot of current carrying capability. So design your crowbar in such a way that it won't freak out over the spike of 'static' from your woolen vest or that lightning strike 2 miles away.
Btw, for most typical 'TVS-type transients' a typical crowbar circuit will probably not trigger anyway since it's such a momentary spike. And for a lightning strike really close by or another VERY fast, VERY high energy pulse, there's pretty much nothing you can do anyway.
I think what once the clamping begins, the low impedance path would be the common where the diode is clamped at. it would quickly clamp common and vcc together?
these are very good points i had not taken into consideration
so what this means is it wont do anything for the crowbar if it clamps to the vcc rail?
I just want to do what i can to protect the devices that are powered by this feeder cable. it would be awful to have to rebuild them all. Im a bit confused now because i thought the TVS would prevent the positive polarity transient from rushing into the circuity ?
What i want to do is create a little box that connects between my feeder cable and my psu. this box should protect from over voltage spikes from no defined source other than the power supply overvolting for whatever reason . The one im most worried about is lightning. i just want to do what i can to protect it.
Worst case scenario, surge comes through and don't blow a fuse but fries my mosfets and shorts them all on
couple of things i do understand,
What do you mean by 'it'? The TVS won't affect the crowbar, no. Well, that is to say, that very fast transients will be clamped by the TVS before the crowbar can respond to them, but like I said, a crowbar generally doesn't protect very well to very momentary/short transients anyway.
As I said earlier, you need to figure out what risks to protect against. If there was a catch-all protection mechanism, everyone would simply hook a 'magic protection black box (tm) (R)' to their circuits...
It does that. At least for very brief, low-energy but high voltage spikes. That's also why you generally don't see them on power supply lines. Those are generally low impedance lines to begin with, relatively short too.
Be sure to patent it; I've given a name to it for you already; see above.
Most of what could be done, you've already done - that's the shielding of your circuit and the cables between modules.
In terms of lightning, the main problem is NOT the PSU you're using. It's the cables to sensors, motors etc. that exit from the shielding and that are relatively long, and go into high-impedance inputs on your modules and/or connect to HV-sensitive devices such as mosfets. Those you could (1) try to keep as short as possible, (2) shield if feasible and (3) protect with TVS diodes.
Item (1) is particularly important. The damage of lightning, at least the kind of damage you can protect against, is due to induction. The shorter you keep your wires, the smaller the magnitude of inductive spikes. Lightning very close to your device will likely do damage whether you protect it or not; lightning still is a force of nature. Those can be pretty brutal.
To a certain extent...it depends on the design and component choice. In human reaction terms, it's still fast. But a TVS will generally be faster.
Are crowbars ever used to protect from lightning? I guess this give me time to build backup devices. I thought if i put a crowbar circuit on the feeder line and the voltage spiked the crowbar would disconnect the power. but i am thinking that the lightning may be so fast that the crowbar wont catch the surge in time anyways.
sounds like it might be a better idea to protect the power supply instead. Maybe a backup UPS power supply to run the psu? chain a few backup ups'S together 
although that would provide me some protection it wouldn't protect me the induction as you mentioned
Thanks you have given me a lot to think about
Don't think so!
Yeah, well, if it worked that way, and if it did, it still wouldn't to anything about the unprotected control wires picking up a few kV of stray inductance here and there. You'd still have a fried Arduino & friends.
Well, of sorts; amazingly these days they survive a lot of abuse; SMPS's have come a long way. I'd just invest in a decent unit to power this project and then take the one you've got there back to the workbench where it belongs. It's a nice toy to keep around when making something new or testing something. Quality is so-so; if you scope the output, it's kind if (very..) noisy. Frankly in terms of cleanliness you're better off powering your project from a suitable fixed voltage SMPS; usually they're much better filtered than the bench power supply you're using now.
As to protection; not sure you can do much against a direct lightning strike, but you could hook up some kind of circuit that sends you a text on your phone when everything has gone offline. I have a feeling you might have something like that running anyways to signal, let's say, people running through your place etc. Just a wild guess.
