Surge protection wonderings

Erik Friesen

I have the below schematic designed into a working system. It finally got its workout the other day from power problems, and here are the results.

Fuses were 10amps, tvs - www.mouser.com/ProductDetail/Vishay-Semiconductors/SMCJ188CA-E3-57T/?qs=sGAEpiMZZMsYiK5PgaDog3CiZPoBePnsj2ef5SeAKrI%3D mov - www.mouser.com/ProductDetail/Littelfuse/V130LA20AP/?qs=%252bb5%252bmZrzU7HQXhj4tF7fmA%3D%3D

MOV's fine. TVS on the hot leg, finished. These fail as a dead short.

Am I in delusion to think that any value of fuse will blow before the TVS dies?

Here is a 2A fuse I am thinking of trying - www.mouser.com/ProductDetail/Littelfuse/0216002HXP/?qs=ar9f0rk5DXB8E0LvYopR9w%3d%3d

kwinn

Even "fast blow" fuses take a bit of time to blow. The MOV's and TVS's are there to short the excess voltage to ground or neutral and cause the fuse to blow. The MOV's and/or TVS's may make the ultimate sacrifice to protect the rest of the circuitry if the voltage spike is fast and high enough. On rare occasions even that sacrifice is not enough.

Erik Friesen

Being as I designed the circuit, I am in question about the validity of my idea. I have considered using Tspd's instead of tvs's as their on resistance is very low, therefore they can handle much more current. I haven't seen them used in any AC line circuits, though.

It isn't a big deal if the fuse blows, I just don't want to be replacing the tvs every time. These are 1500w smc devices by the way.

The last surge happened from a shorted cable. I think the breaker is around 50amp, so I would guess there was a pretty good jolt to the line, as it got two systems on the same transformer.

aercon.net/

kwinn

Erik, the only difference I can see between your schematic and what I use and encounter in the field is a noise filter (usually part of the input receptacle) in front of the fuse, and for 110V only the line is fused.

A few years back lightning struck the transformer feeding the building of one of my customers. The surge destroyed every surge suppressor in the building, fried every PC power supply, most of the motherboards, and quite a few disk drives, as well as the spectrometer I was looking after.

Every logic board in the spectrometer was fried as were several of the low voltage interface and driver boards. This was typical damage for all the other equipment that was connected to AC power. Luckily most of the equipment is powered down by shutting off the main breakers at the end of each day.

If the surge is big enough nothing will save the surge suppressors or the circuitry they protect. Thankfully surges of that magnitude are rare.

bill190

The following is of note...

"The lowest three levels of protection defined in the UL rating are 330 V, 400 V and 500 V. The standard let-through voltage for 120 V AC devices is 330 volts."

From...

http://en.wikipedia.org/wiki/Surge_protector

Erik Friesen

I did a little poor mans testing today. I hooked up 220 to this circuit, and used progressively higher fuses, also using 10w 1ohm resistors in line. Results -

3.15 amp fuse with and without one ohm resistors - Fuse blown only.

4 amp fuse with and without one ohm resistors - Fuse blown only.

10 amp fuse with resistor - Resistors now measure 200 ohms, fuses did not blow, tvs seems ok, mov blew hole in side.

Note - this was done in open, concrete area.

Erik Friesen

Decided to dredge this up, rather than start a new thread.

I tried a 190v sidactor in place of tvs8,33, and 34 http://www.mouser.com/ProductDetail/Littelfuse/P2300SCLRP/?qs=MQaP9csZowJE%2fRy4PR3TZQ%3d%3d

I thought that a sidactor would release each AC cycle, however I am finding that small transients set it off and it clamps until the 2 amp fuse blows. Am I misunderstanding sidactors/thyristors here, or do I just need to up my voltage?

bee_man

I don't think I understand what you are trying to do. I don't use sidactors but I think it would clamp until the current stops. All these components are sacrificial to hopfully save the main circuit. So if you experience a lightning strike you will have to replace the fuses & MOV. If it works the PCB traces are usually gone too.

Jim

LoopyByteloose

Hmmm.....

120VAC-rms actually has peak voltage of 1.414 times that figure or 169 volts at the peak. Are you forgetting this fundamental fact and have your design limits set too low?

Regarding surges.....
There may come a day when a solar flare up will just knock out nearly all and everything that is plugged in. Some times Mother Nature just seems to demonstrate that we are not gods. Canada already suffered a major grid failure from just such a surge.

Nonetheless, I suspect redundancy may enhance a design in ways that higher rated or more sophisticated components cannot.

Another afterthought.... I spent quite a bit of money on a good surge supressor for my home setup and to my dismay I have found that the building offers an extremely poor ground. So any design that relies on a 'neutral' leg may be subject to a poor assumption.

Erik Friesen

Yes, I learned early on about the rms voltage. Granted, 190v is hitting fairly close, although TVS exhibit no issues at that voltage.

The biggest problem with most surge protectors out there is that they rely on the resistance of your house wiring to protect your equipment. Placing a small value resistor in line before allows them to do their job much more effectively, although the resistor is like to burn up in a power surge.

I talked with an old electronics engineer that had worked with a number of different ways to protect systems from lightning. He has used resistors extensively to protect quite a few different things. At one facility he worked at, they were losing a motor a month to lightning(large electric motors). He found some large, low value resistors, and since that point they have replaced a few resistors, but about one motor in five years. He says that the resistor is a block to the high frequency signals brought in by lightning. The only drawback to this method is that it has to be calculated based on load, which is why you can't just put them in a consumer surge protector.