EMP-proofing my system with tvs diodes design idea

[Maast]

After several days of digging through white papers and research articles I've concluded that the best solution is a large capacitor like a Delta CA302 surge suppressor paralleled with a TVS Diode array which is in turn paralleled with a large MOV based SPD like the Midnites or Siemens FS140.

Even though the TVS diodes and actually the MOVs do react in nanosecond time, they don't reach full conductivity quite fast enough so the capacitor (which reacts in picoseconds) is there for whats called "waveform shaping" which is basically slows down the E1 rise time to where the TVS/MOVs can begin conducting and once the E1 and E2 is done and the big-but-slow E3 surge comes in will continue to conduct enough amps to trip the main upstream breaker.
E3 is still a concern because even though the pole transformer core saturates the secondary windings are still subjected to rising voltage because the strengthening magnetic field outside of the core. Core saturation soaks up a lot of energy but it doesnt soak up all of it by any means.

So in summary:
- The capacitor handles the E1 nanosecond rise time
- The TVS diodes handles the trailing edge of the E1 and the beginning of the E2 from 10ns to about 1ms. The TVS diodes also handle the oddball transients that come through and prevent wear on the MOVs, even though the capacitor will do this too however the capacitor doesnt do any voltage clamping just amp absorbing and voltage rise slowing
- The MOVs handle the rest of the E2 amperage surge and the beginning of the huge E3 surge long enough for the:
- The main breaker disconnects the house from the body of the E3 surge.

 

[Hedges]

- The MOVs handle the rest of the E2 amperage surge and the beginning of the huge E3 surge long enough for the:
- The main breaker disconnects the house from the body of the E3 surge.

Try some math based on the current/time to trip main breaker.
Identify what MOV (or array of MOV) can absorb enough to allow that.

I haven't tried. Curious to see what you come up with.

Would be much easier for a device which clamps to zero volts rather than one which clamps at 200V or 800V.

 

[A.Justice]

Its a good idea to keep back-up parts in a F cage or EMP proof your system somehow.

Yes, it is. People laughed at me on Reddit for trying to prep my system for an EMP. They called me crazy to worry about it, but I'm more worried about a Carrington type event, which becomes more and more statistically probable everyday.

I keep all of my old, upgraded-from parts in a diy Faraday cage. I have an old charge controller, inverter, BMS, old cell phone and charger, 18650 charger (I use 18650's for everything), and a few other random electronics that I don't use often, that could be used to charge batteries if my main system is damaged.

It's one of those things I hope I'm wrong about, but will definitely come in handy if I'm right.

 

[Hedges]

... which becomes more and more statistically probable everyday.

Can you explain?

(Is my winning the lottery more and more statistically probable every day?)


Faraday Cage - good for shorting out AC EM fields. Would also short out DC.
What comes from a Carrington event that would have an effect on an unpowered electronic gizmo measuring about 1' in each direction?

 

[Cdkipp]

I was going to say this was a silly idea. Buy, you know, nothing is a silly idea anymore. Excuse me while I go make a massive dehydrated food purchase.

 

[Hedges]

I was going to say this was a silly idea. Buy, you know, nothing is a silly idea anymore. Excuse me while I go make a massive dehydrated food purchase.

Last time there was a Carrington event, telegraphs worked without needing batteries connected. Think what it would do to modern networks.
When a Russian hack took down Ukraine's power grid, it also wiped out Maersk's data on where all their cargo containers were supposed to be delivered.
Who knows, maybe a winter storm could shut down power, so gas stations can't pump diesel for semi trucks, and no food gets delivered.

So yes, store food. Have a wood pile, or bottles of propane. Also invest in gold and lead.

 

[A.Justice]

Can you explain?

(Is my winning the lottery more and more statistically probable every day?)


Faraday Cage - good for shorting out AC EM fields. Would also short out DC.
What comes from a Carrington event that would have an effect on an unpowered electronic gizmo measuring about 1' in each direction?

I just meant that things like that happen now and then, and the longer we go without an issue, the more likely it becomes that it will happen. Sort of like playing the lottery every day; your chances aren't great, but your odds of winning technically increase every time you buy a ticket for the same jackpot.

You are correct about an event like that not doing much, if anything at all to smaller electronics though. All the stuff in my box (I didn't go all out and build a faraday cage, I use a big metal ammo can, and a thick mylar bag. It lowers signal strength about 35 db) is unused, and would otherwise be sitting on a shelf.

I'm not super concerned about something happening, but it's possible, and I feel better knowing that if shit did hit the fan, I can at least charge my phone and radios. I would rather be wrong, and not need it, than be right, and not have it.

I also shamelessly prep some food and water, I've been in situations where there is no food, and people loose their minds. I don't have a bunker with an armory or anything, but I can take care of my family for a few weeks, without relying on civilization working.

When covid hit and there were mask and TP shortages, I had a big pack of TP and a box of masks in storage, same with a bottle of isopropyl alcohol. The stores near me ran out of flour, milk, eggs, and other staples; I had a bucket with things my wife wanted to bake with.

It's not always about "the apocalypse", it's about being ready for natural disasters, man-made disasters, or people panic buying things they aren't going to use anyway.

Sorry for the prepping overshare, I got banned from Reddit and can't post stuff like this anymore.

 

[Maast]

For me its primarily a Carrington event and/or a Cascadia 9.0 quake. With both of those its not a matter of 'if', its a matter of 'when'. Either will kill power and gas for months if not years. Maybe decades. My rationale for a EMP E1/E2/E3 event is I'm already rigging for lightning and with a bit of a stretch and I can cover that too.

Yeah, I had to chuckle at the panic buying at the beginning, and just didnt go to the store for the first 3 months. A bidet is a wonderful thing. Got tired of making bread though - never realized just how much we go through.

I've simply assumed I'll be feeding most/all of my immediate neighbors after the first month (of which there aren't many in my rural/far suburbia AOR). I hope they like beans and rice with a can of roast beef hash thrown in for fat, salt, and flavor because I've probably topped at least 800lbs of one and 600lbs of the other and several hundred cans. I'm keeping the good stuff for me and mine though, especially the lead and iron.

Dont worry about the prepping overshare. I'm somewhat evangelical on the subject of you get me going. I bet after covid and the winter storm there'll be a lot more of us. It only makes sense - shit happens. Usually at the worst possible time.

 

[SpongeboB Sinewave]

After several days of digging through white papers and research articles I've concluded that the best solution is a large capacitor like a Delta CA302 surge suppressor paralleled with a TVS Diode array which is in turn paralleled with a large MOV based SPD like the Midnites or Siemens FS140.

Even though the TVS diodes and actually the MOVs do react in nanosecond time, they don't reach full conductivity quite fast enough so the capacitor (which reacts in picoseconds) is there for whats called "waveform shaping" which is basically slows down the E1 rise time to where the TVS/MOVs can begin conducting and once the E1 and E2 is done and the big-but-slow E3 surge comes in will continue to conduct enough amps to trip the main upstream breaker.
E3 is still a concern because even though the pole transformer core saturates the secondary windings are still subjected to rising voltage because the strengthening magnetic field outside of the core. Core saturation soaks up a lot of energy but it doesnt soak up all of it by any means.

So in summary:
- The capacitor handles the E1 nanosecond rise time
- The TVS diodes handles the trailing edge of the E1 and the beginning of the E2 from 10ns to about 1ms. The TVS diodes also handle the oddball transients that come through and prevent wear on the MOVs, even though the capacitor will do this too however the capacitor doesnt do any voltage clamping just amp absorbing and voltage rise slowing
- The MOVs handle the rest of the E2 amperage surge and the beginning of the huge E3 surge long enough for the:
- The main breaker disconnects the house from the body of the E3 surge.

View attachment 37505

I doubt this does anything at all. Take one apart.Looks the same as their LA302R lightning arrestors that do absolutely nothing at all. Just 3 wires stuck in sand is what they used to be. I heard that Delta started using real MOVs sometime recently due to Midnite giving them crap about their "silicon oxide varistors". But I think that any of the 302 series is just sand.
Beware.

 

[Hedges]

It may actually be a capacitor. Thought I tested one, should go back and try again to be sure.

I confirmed the LA 602 ain't worth $#!+

Delta LA 602 Lightning Arrestor "Hey! Is this thing on??"

I've been using Delta Lightning Arrestors on my system since they were spec'd by Real Goods as part of the package I bought from them. I now have a surge capacitor on the AC input as well, which is supposed to slow rising edge of voltage so it doesn't get as high before the Silicon Oxide...

diysolarforum.com

Hmm, they do have MOV

Delta Lightning Arrestors

Not that I trust them.

 

[SpongeboB Sinewave]

It may actually be a capacitor. Thought I tested one, should go back and try again to be sure.

I confirmed the LA 602 ain't worth $#!+

Delta LA 602 Lightning Arrestor "Hey! Is this thing on??"

I've been using Delta Lightning Arrestors on my system since they were spec'd by Real Goods as part of the package I bought from them. I now have a surge capacitor on the AC input as well, which is supposed to slow rising edge of voltage so it doesn't get as high before the Silicon Oxide...

diysolarforum.com

Hmm, they do have MOV

Delta Lightning Arrestors

Not that I trust them.

Well, everything including wires in sand will have some capacitance. Right ? Have you tried measuring the capacitance ? I have a feeling that these are just re-marked LA302Rs. Not sure though. I would not put it past them. If so, then the capacitance will be very very low. PicoFarads ? Why doesn't Delta put the capacitance rating on the price sheet ? Hmmmm... What is the dielectric constant of sand ?

Also, you see the ETL listing on the Deltas ? That is (was ?) not for any type of UL surge ratings. It had to do with some kind of EMI specification as I remember.

PS. I just looked for the Delta TLC250 (MOV ?) arrestor. There is not picture of it and they are priced at $1400 each. What the heck is that ?

 

[Maast]

The CA302 is a capacitor, the LA are their MOV based arrestors - but I'll be using the Midnites for that.

If you do a search on Mouser or Digikey for 4000v capacitors you'll find motor surge capacitors so I may use those instead.

 

[SpongeboB Sinewave]

The CA302 is a capacitor, the LA are their MOV based arrestors - but I'll be using the Midnites for that.

If you do a search on Mouser or Digikey for 4000v capacitors you'll find motor surge capacitors so I may use those instead.

Thank you. What is the capacitance of the CA302 ?

 

[Maast]

Thank you. What is the capacitance of the CA302 ?

From here it should be 0.1 to 1.0 µF (typical values: 0.13, 0.25, 0.5 µF). Its referring to protection of rotating machines but the principal should be similar enough. I can't find any actual capacity figures for the CA302 or anybody who has actually tested it.

If I get my hands on one I'll put it on our capacitance tester at work.

 

[Hedges]

LA are their MOV based arrestors

The LA series Delta refers to as "Silicon Oxide Varistor"
The MO series say "MOV"

So far, I haven't come across any independent verification of the clamping voltage Delta claims for their LA series.
My tests with a HyPot showed no breakdown of LA 602 up to 7000Vpeak.
Midnight did a similar test with an unlabeled "competitor's product" with similar results, also a test at a UL lab with high voltage/current pulse.
Midnight cut open the unnamed device and something like sand poured out.

Thank you. What is the capacitance of the CA302 ?

I just measured mine. It has one white, three black wires.
I get a little over 1.1 uF between any two wires.
From one wire to two others about 1.5 uF
From 2 wires to two others about 2.4 uF.

Such values would be an actual capacitor with thin dielectric, not just grains of sand.
Impedance vs. frequency would depend on internal construction. Maybe I'll measure that some day.

Normally I don't expect much over 1 MHz out of uF capacitors. Even an SMT component 1 mm long (for 1 nH inductance) and 1 uF has self-resonant frequency (where its impedance is lowest) of 5 MHz. We use a hierarch of different capacitor values from uF down to pF if we want to suppress high frequency noise. Putting a thousand 1 nF capacitors in parallel would total 1 uF but frequency response would be 32 times higher. Only with careful layout/routing do you see the benefit.

A capacitor like CA 302 will have a few inches to a few feet of wire connecting it. Those behave as a transmission line not pure inductance, which has interesting behavior (for those who enjoy using Smith charts.)

Delta's literature says surge arrestors might respond within 5 ns but the surge capacitor is immediate.

https://deltala.com/download/SurgCap2.pdf

I would expect wiring leading to house to limit rise-time coming from a surge, but also isolate any surge arrestor or capacitor somewhat from the wires it is trying to clamp. Those should be kept as short as possible. Even if just 3" of wire to get inside the package, that could be 75 nH and would lower resonant frequency of the capacitor under 1 MHz.

 

[SpongeboB Sinewave]

OK then they must be placing a capacitor inside. I wonder what the voltage rating is ?

May have to take one of those apart sometime. If I do, I will post a picture here.

The name "Silicon Oxide Varistor" was actually coined by the guy from Delta. It's not a real industry component anywhere but Delta.

Their "SOV" does not clamp to any voltage until it blows up, if there is enough current behind it.

 

[Maast]

I just measured mine. It has one white, three black wires.
I get a little over 1.1 uF between any two wires.
From one wire to two others about 1.5 uF
From 2 wires to two others about 2.4 uF.

Well, crap. Those values are FAR too large to deal with a 1ns rise time unless as you said they're from stacked capacitors. They'd be okay for a typical 1ms lightning rise time. Off the top of my head you'd need something in the range of 200-400pf for 1-5ns.

HEMP frequency goes up to ~3ghz with the bulk of it below 1ghz. I'm starting to think capacitors aren't the answer and I should use a Wurth Star-Gap ferrites instead which have a peak at 2.5ghz.

 

[Hedges]

A different physical configuration can work better.

Any time you stub off to a low impedance, of course it doesn't do much above some frequency.

A "feed through" capacitor used to pass a signal into a shielded enclosure can be 1 uF but effective to a GHz and beyond. It is basically 4-terminal device and any signal has to pass through it. Capacitance value is inversely proportional to the voltage that will develop, for a given current, so better to have an effective 1 uF than an effective 400 pF.

If you have something to clamp voltage like a spark gap, after that you can have a series element that is high impedance at higher frequencies. That slows rise time and limits current. An additional parallel low impedance downstream then doesn't have to snub as much. Similar to a low-pass pi filter.

Ferrites? Any magnetic material works great for frequencies within its range up until it saturates. Then it isn't there anymore.
Air-core coils don't saturate. They should hold off voltage so long as dielectric doesn't break down (do calculate the parasitic capacitance).
Resistance helps too. That's the only thing that may save breakers (and people operating them) in the case of a dead short.

A small common mode choke, for instance, could be rated 4A (differential) but about 1% or 40 mA common mode current saturates it.

 

[Maast]

The Wurth Star Gaps are specifically made to not saturate at high currents. They actually have a calculated gap in the toroid. Kind of a neat product.

The goal is to blunt the rise time until the TVS and MOVs can begin conducting. I shouldnt need too many stacked 10kv caps to accomplish that - but I really don't want to build that too but if I do it'll probably be smd capacitors bridging a couple traces on a board. I'd much prefer to buy a premade product. I'd love to tear apart one of the Delta CA302 or 303s to see whats inside it.

Huh, it just occured to me that at this point I'm not too far off from making capacitors out of twisted wire pairs, but the wires would be way too long and produce too much impedance at such a high frequency. I think. More reading

 

[Hedges]

If you took a pair of wires and twisted them, that would have some distributed capacitance (quite small).
If you made an air-core coil of of the twisted pair, it would have common-mode inductance but that would not be seen by differential mode, wouldn't interfere with 60 Hz AC.
In fact, a spool of 2 or 3 plus ground cable would work. Don't want more than one layer for thermal reasons.

This would serve as an inductor to resist a common mode spike and would provide some shunt capacitance to resist differential voltage being developed. What common mode signal does get through, when it encounters different impedances, will try to generate differential.

TVS and other clamping devices before and after will help. The ones before will experience the full transient, but after only have to clamp what gets through, primarily the lower frequencies.