diy solar

diy solar

Grounding my van electrical.

For inverters they say same size or 1 awg less than the cable powering the device (my manual does). Good rule of thumb to follow.

Can anyone explain what the ramification would be of not grounding the DC negative bus bar? Why should that be done?
To answer your question about the bus bar:

You wouldn't want to not "ground" it to the chassis. It improves the (negative) connection between your alternator and starting battery and is another link in the chain where you want good connections. If your charging system is otherwise not connected then it's kind of irrelevant however.

BUT.

If you're using an inverter and have your inverter case ground wire connected to that bar as you should, then grounding the bus bar actually properly then is actually true ground situation which would allow safety devices such as a GFCI to function properly via the borderline black magic that makes them work.

While I can't tell you if it would or would not trip without it, I do know that it's designed to be at the same or very similar potential as any unfortunately involved person to ensure it detects the fault and stops it before you get hurt or die.

If you don't do that you can wind up with a hot case at 120/240 volts ac, or worse other things in the system become electrified.
 
So continuing on the topic of grounding:

Given a choice between having a gfdi on the negative bus vs having the negative bus grounded in the traditional manner (with a low resistance connection directly to ground), which offers more/better protection?
gfdi? Did you mean gfci? Want to be sure I'm not thinking of something else.

In that case I would argue that it is *grounding* and you're safer 10 out of 10 times to properly attach your chassis and inverter ground to your common negative bus. That's how these things are designed and what the install schema show.

There's no reason you can't put your negative and inverter's AC grounds together and many reasons why you should, and the inverter ground needs to be connected to the chassis no matter what.

Since the negative stuff can and usually is connected to the chassis also, it just makes sense to just put them together.

This also satisfies galvanic corrosion issues simultaneously as some of the marine guys keep pounding into my head and I should just accept.
As a car stereo guy a single ground point is good but hasn't been necessary for our equipment in years, but I'll admit it's good practice.
 
To answer your question about the bus bar:

You wouldn't want to not "ground" it to the chassis. It improves the (negative) connection between your alternator and starting battery and is another link in the chain where you want good connections. If your charging system is otherwise not connected then it's kind of irrelevant however.

BUT.

If you're using an inverter and have your inverter case ground wire connected to that bar as you should, then grounding the bus bar actually properly then is actually true ground situation which would allow safety devices such as a GFCI to function properly via the borderline black magic that makes them work.

While I can't tell you if it would or would not trip without it, I do know that it's designed to be at the same or very similar potential as any unfortunately involved person to ensure it detects the fault and stops it before you get hurt or die.

If you don't do that you can wind up with a hot case at 120/240 volts ac, or worse other things in the system become electrified.
Ok. So Given a GFDI device in which the DC negative is bonded to ground through, and if that device trips it still maintains a high resistance connection to ground through a resistor bridged across that GFDI's ground pole, which is safer?
 
Ok. So Given a GFDI device in which the DC negative is bonded to ground through, and if that device trips it still maintains a high resistance connection to ground through a resistor bridged across that GFDI's ground pole, which is safer?
So then a dc ground fault detector and not an ac gfci. Got it. They work somewhat differently.

I'm unfamiliar on whether or not this would even function in a mobile application without isolating the cases through it. Every one I've seen requests an earth bonded ground and thus is only suitable for residential or otherwise fixed applications. They work by detection of negative bus bar current to earth ground. With a resistor as you've stated. This high resistance shouldn't cause any issues with safety however in a mobile application this earth ground isn't available.

I believe the whole point of this device is to go between the negative bus bar and ground. This way the ground path cannot bypass the gfdi.

I take it you're not using this mobile then?

If you are then I don't believe the device would function, as the "ground" side of the gfdi is at exactly the same potential as the negative/input and thus cannot trip.

Maybe if all of your negative connections are totally isolated from the chassis but since that's not ever a scenario I've found myself in I can't say it'll still function even then.

Got a manual for your specific device?
 
So then a dc ground fault detector and not an ac gfci. Got it. They work somewhat differently.

I'm unfamiliar on whether or not this would even function in a mobile application without isolating the cases through it. Every one I've seen requests an earth bonded ground and thus is only suitable for residential or otherwise fixed applications. They work by detection of negative bus bar current to earth ground. With a resistor as you've stated. This high resistance shouldn't cause any issues with safety however in a mobile application this earth ground isn't available.

I believe the whole point of this device is to go between the negative bus bar and ground. This way the ground path cannot bypass the gfdi.

I take it you're not using this mobile then?

If you are then I don't believe the device would function, as the "ground" side of the gfdi is at exactly the same potential as the negative/input and thus cannot trip.

Maybe if all of your negative connections are totally isolated from the chassis but since that's not ever a scenario I've found myself in I can't say it'll still function even then.

Got a manual for your specific device?
In a mobile application, chassis is ground. The idea of the device is that if current gets detected on the negative bus, it trips. In my case, the device I am considering is a 3 pole gfdi. 1 pole for SCC1+, 1 pole for SCC2+, and the third pole is the ground pole. one side is connected to DC negative, other side is connected to ground (chassis). if it senses 1/2a or more of current flowing from negative to ground, it trips all poles severing the solar system from the dc bus. A high resistance connection would still exist across the ground pole since it is bridged with a resistor.

Before you go there, I will say this replaces a traditional dc negative low resistance connection to ground. This would effectively be the _only_ galvanic connection to ground from the DC negative bus.

Is it better protection, less, or no different than using a traditional ground?
 
That sounds like the outback gfdi I'm familiar with, which is defined for an earth ground. I've attached their schematic for your reference.

As stated, chassis cannot be ground for a gfdi if it's at the same potential as the bus bar. It would not function. You stated this would be isolated. I'm with you so far.


There is always current on a negative bus during normal operation, so that's not what it wasn't to detect, but rather fault current between this negative bus and ground. You'd also need each device case/frame isolated from the negative and connected to the chassis as well. If you do all of that and truly have an open circuit between all negative wires and the chassis then you will have created a ground path which should make the device function.

In this manner any accidental positive connection to these device cases or frames would result in current flowing through the chassis and GFDI and to the negative bar, thus tripping it.


If you are isolating everything from the power system such that this is the only connection in the entire isolated system to the chassis then I believe it might work that way but I disagree with the decision to isolate them from the chassis in this manner.

I'm of the opinion that all DC negative leads in a mobile situation should be tied to the chassis because it allows one to utilize a common connection point for the AC voltage circuit ground as the manufacturers suggest.
 

Attachments

  • Screenshot_20210803-125420_Drive.jpg
    Screenshot_20210803-125420_Drive.jpg
    79 KB · Views: 6
Regarding the resistor, I'm actually not enough of an EE to be able to tell you if that resistor is specifically sized for an earth ground application and if it has any negative effect if connected with a low resistance chassis ground as you've described, because what I'm familiar with simply isn't designed for that.

Control wiring is typically far less theoretical than what's required to make any recommendations beyond the design and testing (and thus recommendations) of the engineers who created the products I'm using, and I'm a mech E student to boot.

Most of what I do is interfacing/programming robots, presses, and safety guarding with the occasional "high" voltage retrofit 480v and under.
 
That sounds like the outback gfdi I'm familiar with, which is defined for an earth ground. I've attached their schematic for your reference.

As stated, chassis cannot be ground for a gfdi if it's at the same potential as the bus bar. It would not function. You stated this would be isolated. I'm with you so far.


There is always current on a negative bus during normal operation, so that's not what it wasn't to detect, but rather fault current between this negative bus and ground. You'd also need each device case/frame isolated from the negative and connected to the chassis as well. If you do all of that and truly have an open circuit between all negative wires and the chassis then you will have created a ground path which should make the device function.

In this manner any accidental positive connection to these device cases or frames would result in current flowing through the chassis and GFDI and to the negative bar, thus tripping it.


If you are isolating everything from the power system such that this is the only connection in the entire isolated system to the chassis then I believe it might work that way but I disagree with the decision to isolate them from the chassis in this manner.

I'm of the opinion that all DC negative leads in a mobile situation should be tied to the chassis because it allows one to utilize a common connection point for the AC voltage circuit ground as the manufacturers suggest.
It is the outback GFDI.

All of my devices are mount-isolated from the chassis and have their chassis' EGC-bonded to the ground bar (which is in turn bonded to the chassis). This includes the solar panels.

I should also disclose that this chassis ground is common to the chassis ground for the dc negative of the chassis battery system. (yes, there will be alternator charging).

I don't want to beat a dead horse. "It could work" doesn't sound like "it works." I want, "it works." If you think this is somehow questionable, then there is no reason I can't stick with a traditional DC negative ground. If I can step up protection a bit in a simple manner, then I would like to do so.
 
Eh. I enjoy the understanding but this should be it's own thread if I were a mod.

So your device cases are connected to the chassis but your solar system negative is not, aside from the factory charging system?

If you utilize an isolated DC to DC charger you *might* be able to make the gfdi work but it's such an unusual configuration for me I just can't answer it.

I don't know which gfdi you intend to use but my experience with them is that they're not designed for anything other than a true earth ground.


If/since your mounts are isolated from the chassis you could just run all the device case grounds to the gfdi (or an isolated case ground bus bar then the gfdi) and then tie that output to the negative bus bar. Any positive wire touching any case ran through the gfdi should trip it, but I can't answer that resistor issue.
 
I will say however that a gfdi is not a human safety device like a gfci.

They (at least this specific one) require half an amp of current which is enough to harm or kill a person if it's high ac voltage causing the fault.

In this regard, normally just fusing things appropriately for the wire size used at any point is sufficient to obtain more or less the same protection.

That said, the outback gfdi is really only intended for the solar panels and not anything else.

The same result can be had with fuses as close to the panel as possible, such as in line mc4 fuses. Fusing solar panels is it's own whole discussion though and involves knowing the fused circuit rating of the panels.
 
Regarding the resistor, I'm actually not enough of an EE to be able to tell you if that resistor is specifically sized for an earth ground application and if it has any negative effect if connected with a low resistance chassis ground as you've described, because what I'm familiar with simply isn't designed for that.

Control wiring is typically far less theoretical than what's required to make any recommendations beyond the design and testing (and thus recommendations) of the engineers who created the products I'm using, and I'm a mech E student to boot.

Most of what I do is interfacing/programming robots, presses, and safety guarding with the occasional "high" voltage retrofit 480v and under.
Duly noted. Your input is appreciated.
 
I will say however that a gfdi is not a human safety device like a gfci.

They (at least this specific one) require half an amp of current which is enough to harm or kill a person if it's high ac voltage causing the fault.

In this regard, normally just fusing things appropriately for the wire size used at any point is sufficient to obtain more or less the same protection.

That said, the outback gfdi is really only intended for the solar panels and not anything else.

The same result can be had with fuses as close to the panel as possible, such as in line mc4 fuses. Fusing solar panels is it's own whole discussion though and involves knowing the fused circuit rating of the panels.
Thanks for the input. Now we're definitely deviating from grounding discussion. I'll take a step back. Your input has been helpful and I don't want to hijack OP's thread.
 
Thanks for the input. Now we're definitely deviating from grounding discussion. I'll take a step back. Your input has been helpful and I don't want to hijack OP's thread.
Your question however exactly the reason why I hate using ground to describe anything at the same potential as negative in mobile situations though.
 
Back
Top