2
\$\begingroup\$

I have a pickup truck with a 3kw AIMS inverter-charger. It is wired with 4/0 welding cable (for both positive and dedicated neutral back to the alternator), and I installed a block with a 300A fuse in the engine bay as close to the alternator as possible. My concern is that most of the conceivable events causing a short circuit would have a current path with resistance such that the short would be much less than 300A and therefore not blow the fuse.

What is the recommended practice for wiring such systems? I did a search for "DC ground fault protection" but none of the products that came up appeared to be appropriate for my application. If a 300A 12VDC GFCI device was available I would gladly purchase it. Thanks in advance.

Edit: I am talking about a live-to-ground short, not live-to-neutral. Suppose the insulation surrounding the live wire to the inverter (which is routed along the truck's frame) is somehow sliced open and the bare wire makes a weak short against the frame. This is the event I want to be protected from. To be clear, answers pertaining to the physical protection of the live wire are not appropriate here; I have already taken steps to make sure it is as far out of harm's way as possible.

Edit 2: These are the events I foresee possibly occurring as a result of a live-to-ground short (and want to prevent): 1) batteries being drained, 2) batteries exploding due to a short of up to 300A, 3) a fire being started due to the resistive heating from a short of up to 300A through components of the body or frame.

Edit 3: I emphasize that the inverter has a dedicated neutral that connects directly to the alternator housing. Therefore, ALL return current must pass through this wire unless there is a live-to-ground fault somewhere. In such an event there will be a difference in current between the live and neutral wires, which is the condition upon which I want my mystery device to trip and disconnect the live wire.

Edit 4: Here is a concrete example of a situation I want to prevent from which the fuse offers no protection. Imagine that the vehicle is involved in a collision and the body of the truck pierces the 12V wire insulation. The 12V wire will now have a weak (the body of the truck is painted and likely dirty) short against the truck body, resulting in finite current that could quite possibly be way less than 300A but still sufficient to cause a fire due to resistive heating and/or battery explosion (or even just drain the battery and make the truck unusable).

\$\endgroup\$

closed as off-topic by PlasmaHH, ThreePhaseEel, laptop2d, PeterJ, Harry Svensson Dec 4 '17 at 18:53

This question appears to be off-topic. The users who voted to close gave this specific reason:

If this question can be reworded to fit the rules in the help center, please edit the question.

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$ – Dave Tweed Dec 2 '17 at 0:33
1
\$\begingroup\$

If you wanted to do your own "bake-at-home" GFCI, you could.

What you need is:

  • a way to measure the current rolling through each of your feeder lines -- Search "300A Shunt Resistor" (you'll need 2)
  • a special way to pre-amplify the measurement on the 12V line -- search "high-side current measurement amplifier"
  • a special way to pre-amplify the measurement on the "neutral" line -- search "current measurement amplifier"
  • A way to compare those two measurements -- I suggest a micro-controller search "Arduino"
  • A way to control the disconnect the line -- Omron has some big relays you might find interesting.

You'll need a bunch of other glue circuitry to make all those things work together, but this is not a design service.

As a side note, You might have better luck not trying to find a GFCI for this application; but rather design a way that your feeder lines are much less likely to short out, 3M makes high-temperature glass-based electrical tape you can wrap your lines in. And honestly, I doubt 4/0 welding cable insulation is going to fail you any way, but you can tape up the connection points.

\$\endgroup\$
  • \$\begingroup\$ Finally, a useful answer. Thank you. I have wrapped the hot wire with hard plastic wire loom as a preventive measure and I am very confident that it will not fail, but I figured if a reasonably priced ground fault protection device was available I would purchase it as well. That said, your solution is way beyond my capabilities -- it seems like I will just have to roll the dice. \$\endgroup\$ – Josh Dec 1 '17 at 14:04
  • 1
    \$\begingroup\$ Honestly, I would roll the dice. Just protect the connection points. GFCI'ing the AC side /w off-the-shelf stuff is not a bad idea if your inverter doesn't already do it. \$\endgroup\$ – pgvoorhees Dec 1 '17 at 14:19
  • \$\begingroup\$ Yeah, the inverter has a built-in GFCI outlet and I ran 10/2 SOOW to the truck bed where I mounted an auxiliary 20A GFCI and 30A generator-style (NEMA 14-30) outlet. My worry was hitting some road debris which might kick up and cut through the hot wire and possibly cause it to short against the truck's frame. Given how I've routed the wires and the amount of physical protection I've afforded them it seems an extremely unlikely scenario, but if some off-the-shelf hardware was available to eliminate the possibility I would probably be willing to invest in it. \$\endgroup\$ – Josh Dec 1 '17 at 14:52
1
\$\begingroup\$

The solution is to protect each cable with a circuit breaker or fuse of appropriate rating to prevent the cable from melting in the event of a short circuit.

Ground fault protection is to safeguard humans against tiny currents flowing from mains to earth through a high resistance person. Such a current wouldn't trip the cable protection breakers, so the GFCI trips. It's only useful for leaks to earth, but this is a common way people are shocked. It provides no protection from live-to-neutral shorts.

Your installation should be safe against a short to ground at any point. If you have a large gauge cable with a large fuse, that's fine, but if you then take a smaller cable from that, the smaller cable needs a smaller fuse where it starts. This is usually done in a distribution board, for mains, and a fuse box, in a vehicle. Small fuses plug into the big 12 V distribution panel, and the thin wires go from there.

\$\endgroup\$
  • \$\begingroup\$ With respect, this is simply inaccurate. The assumption that a short circuit would melt 4/0 welding cable is not reasonable, and I explicitly stated in my question that my concern is short circuits with resistance such that the current through the short is well below the fuse or breaker rating of 300 amps. Such shorts are common in automotive applications due to the low voltage. Also, the short circuit I am describing is live-to-ground, not live-to-neutral. GFCI definitely provides protection from such events, the issue is whether such a device exists for this application. \$\endgroup\$ – Josh Dec 1 '17 at 11:50
  • 1
    \$\begingroup\$ If your source is limited to produce less current than your cable can handle, why put in a fuse? Anyhow, you said you were worried about smaller devices being damaged, not the thick cable. In that case, protect them with smaller fuses, eh? \$\endgroup\$ – tomnexus Dec 1 '17 at 12:09
  • 1
    \$\begingroup\$ I said nothing about protecting devices; I'm concerned about the following: 1) batteries being drained, 2) batteries exploding due to a short of up to 300A, 3) a fire being started due to the resistive heating from a short of up to 300A through components of the body or frame. \$\endgroup\$ – Josh Dec 1 '17 at 12:13
  • \$\begingroup\$ I see where you are going, but I think a big fuse is all you need. If the inverter draws a full 300 A, then a short circuit will draw more. If the body is somehow not able to carry 300 A, then it needs protection too. How about a 100 A fuse in the link between battery negative and body? \$\endgroup\$ – tomnexus Dec 1 '17 at 18:57
  • \$\begingroup\$ A weak short could definitely draw way less than 300A. Imagine the vehicle was in an accident and the body of the truck was deformed such that it pinched and sliced through the insulation of the hot wire, barely making contact with it. Now we have a 12V wire with a ground fault against the painted truck body, which is also dirty. This is not a zero-resistance short. The current will be finite, and very possibly less than 300A. Anyway, the neutral line is not the issue since my concern is with a ground fault. \$\endgroup\$ – Josh Dec 1 '17 at 19:05
1
\$\begingroup\$

A ground fault interrupter is NOT what you need. You need an appropriately sized fuse or circuit breaker.

Ground fault interrupters are an entirely different thing. They require three connections: a hot, a neutral, and a ground wire. The current normally only flows through hot and neutral. Current flowing from hot to ground causes the ground fault interrupter to trip.

Your DC system has hot and ground. No neutral, so a ground fault interrupter is not possible.


You mention an inverter. I take it that means you are making 120VAC from the 12VDC supply.

It might make sense to have a ground fault interrupter on the AC side, if it has ground, hot, and neutral. But, that solves a different problem - it would protect you from certain faults on the AC side.

\$\endgroup\$
  • \$\begingroup\$ This answer is a duplicate of the one below by @tomnexus and my comment on his answer applies here as well. This does not solve the problem described in my question. Also, there is hot/neutral/ground: the 12V wire, the dedicated neutral wire (through which all return current should pass), and the frame/body of the truck is ground. My 300A fuse does nothing to protect the electrical system and batteries if a weak short was to occur between the live wire and the frame of the truck, as I described in my question. \$\endgroup\$ – Josh Dec 1 '17 at 12:10
  • \$\begingroup\$ Not a duplicate - I explained a bit about what a ground fault interrupter does. A gcfi still wouldn't help. How would you measure the current flow through the entire body and frame? And be able to tell if it is caused by a short in your high current system or a problem (or normal operation) in the regular 12V system? You can't really tell. So, proper fusing and maybe an ampere meter in the high current line. \$\endgroup\$ – JRE Dec 1 '17 at 12:16
  • \$\begingroup\$ There is no need to measure that quantity at all. The GFI would be installed across the live and neutral connections (I reiterate that there is a dedicated neutral for the inverter -- its return path is not through the frame); any difference in current between the live and neutral wires indicates a fault, which should trip the mystery device that I seek! \$\endgroup\$ – Josh Dec 1 '17 at 12:18
  • 1
    \$\begingroup\$ Your ground is not a separate wire, it is the entire truck. You cannot measure current "across" two connections. You measure current "through" a conductor. To do that, you must insert the amperemeter in series with the conductor. You can do that with the neutral wire. How do you intend to measure current through the truck body and frame? \$\endgroup\$ – JRE Dec 1 '17 at 12:44
  • \$\begingroup\$ You could measure a voltage difference between neutral and ground, but you can't be sure of what it means. The truck uses the body and frame as the return. So, you could measure a voltage difference between your high current "neutral" and ground, and it might only mean that there's a normal current through the body and frame. You can't tell. \$\endgroup\$ – JRE Dec 1 '17 at 12:47
1
\$\begingroup\$

I have to agree with @tomnexus. Fuses are typically used to prevent meltdowns or continuous arc flashes, whether it be a cable melting down, or a battery exploding, etc. GFCIs are used for human safety, because even a small amount of current can cause serious harm at the right voltage. That said, I see your issue and what you are trying to do, and there's no easy solution. In your case I would recommend simply hoping a short doesn't happen as any reasonable solution would be either prohibitively expensive or complex.

I have recently encountered this issue with one of the batteries I am working on. It is designed to output 300kW continuous, and has overload protection on the order close to 500kW. A 200kW arc flash is quite powerful and very dangerous, however this would be seen as normal operation by the electronics. In the case of this battery there are a fair number of sensors, including UV for arc flash, electrolyte sensor to detect ruptured cells, etc. However it is still a very real problem and has no easy solution.

I'm afraid the best solution is to simply protect the live wires as well as possible and hope nothing happens. It seems you have taken a fair number of precautions to prevent it, so it is very unlikely and I wouldn't worry to much, but is always a possibility and there's not much more you can do.

\$\endgroup\$
  • \$\begingroup\$ Fair enough, I am willing to accept that there may be no cost effective solution. \$\endgroup\$ – Josh Dec 1 '17 at 17:14
  • \$\begingroup\$ Just re-read this -- out of curiosity, what sort of batteries put out 300kw? I'm assuming this thing must be about the size of my truck. \$\endgroup\$ – Josh Dec 2 '17 at 9:35
  • \$\begingroup\$ @Josh Its about 5'x5'x2' in size. It has a nominal voltage just north of 1kV, so its quite scary. \$\endgroup\$ – Redja Dec 4 '17 at 14:57
  • \$\begingroup\$ That is pretty cool. What is that sort of technology called and where can I read about it? How do the kJ/kg and kJ/m^3 compare to, say, fossil fuels like gas and diesel? \$\endgroup\$ – Josh Dec 4 '17 at 22:03
1
\$\begingroup\$

This answer does not answer your question directly, but does provide a more standard, and possible better alternative approach.

OK so, from what I understand, your root problem here is that you are concerned about a situation occurring where a large current drain will occur from the vehicles electric system caused by a short that is not sufficient to trip a fuse.

In order to resolve this you are considering using a two fault protection mechanism, in the form of a GFI. That is, in order for the situation you are trying to avoid to happen, you must have two faults. 1. A short somewhere, and 2. A fault in the GFI.

Although that approach is a technically valid one, the costs and commercial availability of 300A D.C. GFIs is prohibitive. Creating a home grown solution is possible, but that is no minor undertaking and you would still need an expensive contactor to break the circuit, or a crowbar element to blow the fuse.

So how do you provide two fault protection without adding an active breaker?

This is traditionally done by double insulating the cables and even triple insulating where there is a cut risk. If you have extra protection on the cables such that the outer protection must fail as well as the cable insulation itself, you effectively have double fault protection.

The first thing I would do, is add conduit to the cable. Polyamide tubing is really tough stuff, readily available, and easy to use. At 10-20 cents a meter it is also very cheap.

enter image description here

Further, wherever the tubing ventures through holes in bulkheads etc, add appropriate grommet strip material to protect the tubing from any sharp edges. enter image description here

Finally, tie everything down. Wear from vibration and movement will be the long term failure mode. The less things move, the longer it will survive. Use tie down and clips where you can and possibly even calking through holes in bulk-heads.

enter image description here

Ultimately, by properly protecting your wiring this way, you not only provide the two fault protection you desire, but you also significantly reduce the possibility of the primary fault you wanted protection from in the first place.

\$\endgroup\$
  • 1
    \$\begingroup\$ All good suggestions. I used wire loom and the insulation is a good 1/8" thick anyway. I used a strain-relief clamp connector in the hole I drilled in the back of the cab, so that should be good as well. Thanks for the input. \$\endgroup\$ – Josh Dec 1 '17 at 19:23

Not the answer you're looking for? Browse other questions tagged or ask your own question.