Relay panel bracket electrified ... 12Vdc!

[nkemp]

While trying to fix another problem, I noticed that sparks flew when a jumper connecter to the cooling fans touched the relay panel metal frame. To see how much current was there I touched one end of the wire to the fan terminal and the other to the bracket and enough current flowed to run the cooling fan.

I clamped on a meter and watched it as I moved stuff and it worked....I found the problem.

The trouble is that the metal jacket on the accessory relay goes hot ... sometimes. It depends on the position of the internals (they move depending on how pressure is applied to the wires). Once that metal housing is hot, the whole electrical panel goes hot since it is not grounded (which is probably good in this case in the event that the 12Vdc source may not be fused).

For now the solution is to isolate the relay from the panel. An identical replacement may not be the best idea since it may be prone to the same type of problem. A plastic housed relay is the better choice but it needs to have the current carrying capacity.\

FYI everyone

Nick

I checked the main relay and it sorta acts the same. When I move the wires, the meter readings jump around sometimes looking like they may have hit 12Vdc. That relay is the same type relay as the accessory.

NOTE: this could be the reason for some electrical compartment fires. The power to those relays is a direct feed (non fused) from the battery. So in the right condition where the hot relay exterior/panel touches a ground it could get very hot and "sparky" before it disconnects. From what I'm seeing, this should be a service upgrade! We need to find a suitable replacement relay where the housing cannot get eletrically hot.

[Bitsyncmaster]

I split some vacuum hose and put it onto the edges of the panel where wires lay. The metal is not normally grounded. Some things to check are relay pins that have pushed out.

[Patrick C]

Very weird. I have that relay panel grounded to the main firewall ground and use that panel as a "ground buss" so to speak for all of my aftermarket electrical add-ons. No problems whatsoever here.

[nkemp]

Very weird. I have that relay panel grounded to the main firewall ground and use that panel as a "ground buss" so to speak for all of my aftermarket electrical add-ons. No problems whatsoever here.

From what I've seen here, I would not ground the relay panel. The reason being is if the Main or Accessory relay develops an internal short from the hot lead to the relay housing it will electrify the relay panel bracket via the mounting screw that secures the relays to the panel. The wires feeding the Main and Accessory relays are not fused or breakered as best I could see on the wiring diagram. As such, one of the wires that completes the circuit (from battery+, to the relay, to the panel to battery ground) will be the one that burns up if a short develops.

Another option (and maybe the one everyone should do now) is to remove the Main and Accessory relays from the metal panel and put something over them (currently I'm using a sock ... don't laugh) that isolates them from the panel.

No good, and many possible bad things can happen with unfused 12Vdc powering the relay panel.

[nkemp]

If you want to test to see if your Main and Accessory relays can or have developed internal shorts the process is pretty simple:

If your panel is grounded, remove the relay from the panel before proceeding or you could get some nasty results. The panel did not come from the factory grounded. You or the PO may have intentionally grounded the panel. Or an internal wire may have worn and may be touching the metal panel. For safety sake, I recommend removing the relay from the metal panel before testing.


Clamp the negative lead on a volt meter to ground (not the metal panel)
Clamp the positive lead to the relay tab or tab screw
While holding the relay, move the relay wires and watch the voltage reading. It should read zero or may fluctuate in the mV range (that is not the concern we are looking for). In my case there are positions where the relay housing stays electrified at battery level (~13Vdc) on one relay. The other simply "bumps" up and down very quickly as if there is a temporary internal short connection.

[Bitsyncmaster]

You can open that "AUX" relay can. I did that with mine and made them solid state. But you can look as to where it's shorting to the can.

[nkemp]

You can open that "AUX" relay can. I did that with mine and made them solid state. But you can look as to where it's shorting to the can.

But not before DCS or I might not make it :)

[Patrick C]

Well, I suppose I will un-ground the bracket and instead install a ground bus similar to what I did for my 12v bus. I'd rather be safe than sorry if this is a known issue with the stock relays shorting out to their metal housings.

Has anyone else had problems with the relay bracket becoming electrified?

[dmc6960]

I've had my panel grounded for 10 years. Though my main and accessory relays have been detached from it for also 10 years. Didn't someone post a link recently to an awesomly inexpensive 70a or 80a relay which could be used as a replacement?

[nkemp]

Well, I suppose I will un-ground the bracket and instead install a ground bus similar to what I did for my 12v bus. I'd rather be safe than sorry if this is a known issue with the stock relays shorting out to their metal housings.

Has anyone else had problems with the relay bracket becoming electrified?

It would be easier to unscrew the two relays, wrap them in something that insulates them and tuck them back in (but don't screw them to the relay panel bracket. The idea being that if they do go hot on the relay can that they don't make the bracket hot.

[Patrick C]

It would be easier to unscrew the two relays, wrap them in something that insulates them and tuck them back in (but don't screw them to the relay panel bracket. The idea being that if they do go hot on the relay can that they don't make the bracket hot.

I think wrapping them would trap the heat inside the relay and rapidly destroy the internal components, no?

[nkemp]

I think wrapping them would trap the heat inside the relay and rapidly destroy the internal components, no?

Depends on what you wrap them with. A layer of loom tape would not thermally insulate them that much. I wouldn't suggest something like pipe insulation as it would provide too much thermal insulation. Besides, relays shouldn't generate that much heat if properly designed and applied... which are both good questions.

[Patrick C]

Well, I installed a ground buss today next to my 12V buss. The relay bracket itself is no longer grounded.

[Bitsyncmaster]

I just installed a new AUX relay I bought from DEL CITY part number 73998. This relay fits like the OEM one, is rated at 70 amps and coil resistance is 92 ohms. It is a black plastic case so would not have a problem with electrifying the metal bracket. Cost for each was about $7. I could not open it up since it is sealed shut.

So I could not verify it's 70 amp rating or construction inside.

Marking on the relay:

Song Chuan
897-1AH-D1
Coil:12VDC
70A 14VDC

[nkemp]

Good find. Same connectors so that no rewiring required?

[Bitsyncmaster]

Good find. Same connectors so that no rewiring required?

Yes, plug and play. The two power terminals are the large copper ones same as OEM. They do send you some crimp terminals but just keep them for your other needs.