Metal halide lights trip GFCI all of a sudden

In article ,
says...
You have one of 2 things going on. Either your GCFI is getting more
sensitive which is causing it to trip. Or you have a cap going out on
one of your ballasts. What you should do to check is try only running
one halide at a time and see if it still trips. that way you can
isolate it down to either one ballast or the GFCI.

One suggestion I would have for you is redo your tank wiring, having
everything on the same GFCI is very dangerous for your tank. If
anything goes wrong you loose all power to your tank, which normally
means you loose all life in your tank after a while. Move atleast a
couple of filters/water movement devices to another circuit so that you
will not loose everything.

Kim

Hi Kim-

Thanks for some good advice. When I set up the tank years ago, I had to
run a circuit to a relatively remote location, and just ran one. After
some time, it occurred to me that splitting it over two circuits would
have been better for the reasons you brought up. But, with a 90 gallon
tank in the way, it was harder to do at that time than I was willing to
handle, and since it worked well for years I eventually just let it be.

The tank is highly redundant, by design. In addition to the main filter
(a big old Eheim, which presumably died the second time the power was
out for~16 hours), there's an internal power head with sponge filter,
which doesn't die when the power goes out. Lots of plants, not too many
fish, etc.. The main filter probably represents 50% of the
biofiltration capability in the system.

So, the real threat is the power coming back on after a prolonged outage
dumping nasty dead-filter water back into the tank. When the filter
went dead, I manually unhooked it and let a few gallons run through to
waste.

Meanwhile, I'll see about testing the MH systems individually if
possible. They may be connected inside the metal box; not sure if they
are intended to be operated separately. And, it's an intermittent
problem, so it will take a while to sort out.

If that doesn't do it I'll swap the GFCI.

Thanks

Marc

"MAG" wrote in message
.net...
In article ,
says...
You have one of 2 things going on. Either your GCFI is getting more
sensitive which is causing it to trip. Or you have a cap going out on
one of your ballasts. What you should do to check is try only running
one halide at a time and see if it still trips. that way you can
isolate it down to either one ballast or the GFCI.

One suggestion I would have for you is redo your tank wiring, having
everything on the same GFCI is very dangerous for your tank. If
anything goes wrong you loose all power to your tank, which normally
means you loose all life in your tank after a while. Move atleast a
couple of filters/water movement devices to another circuit so that you
will not loose everything.

Kim

Hi Kim-

Thanks for some good advice. When I set up the tank years ago, I had to
run a circuit to a relatively remote location, and just ran one. After
some time, it occurred to me that splitting it over two circuits would
have been better for the reasons you brought up. But, with a 90 gallon
tank in the way, it was harder to do at that time than I was willing to
handle, and since it worked well for years I eventually just let it be.

The tank is highly redundant, by design. In addition to the main filter
(a big old Eheim, which presumably died the second time the power was
out for~16 hours), there's an internal power head with sponge filter,
which doesn't die when the power goes out. Lots of plants, not too many
fish, etc.. The main filter probably represents 50% of the
biofiltration capability in the system.

So, the real threat is the power coming back on after a prolonged outage
dumping nasty dead-filter water back into the tank. When the filter
went dead, I manually unhooked it and let a few gallons run through to
waste.

Meanwhile, I'll see about testing the MH systems individually if
possible. They may be connected inside the metal box; not sure if they
are intended to be operated separately. And, it's an intermittent
problem, so it will take a while to sort out.

If that doesn't do it I'll swap the GFCI.

Thanks

Marc

I don't know about your tank but my electrical connections tended to corrode
over a few year's time. You say you have had your tank up for 5 years.
Check ALL of the electrical connections for corrosion. Corrosion can cause
"leakage current" which is what will set your GFCI into FREAK out mode and
trip the circuit breaker. It may be working PERFECTLY.

ALSO, make sure there is no moisture buildup anywhere in the electrical
system. Afterall we are talking about a high humidity system that is rather
close to water. Moisture buildup could cause a current LEAK somewhere and
BAM! The GFCI is again working perfectly.

You may find that the GFCI is indeed week, however I'd check the other above
mentioned things as well.

YMMV,

DJay

In article 7EaPe.9270$Vg7.2812@trnddc06, says...
I don't know about your tank but my electrical connections tended to corrode
over a few year's time. You say you have had your tank up for 5 years.
Check ALL of the electrical connections for corrosion. Corrosion can cause
"leakage current" which is what will set your GFCI into FREAK out mode and
trip the circuit breaker. It may be working PERFECTLY.

ALSO, make sure there is no moisture buildup anywhere in the electrical
system. Afterall we are talking about a high humidity system that is rather
close to water. Moisture buildup could cause a current LEAK somewhere and
BAM! The GFCI is again working perfectly.

You may find that the GFCI is indeed week, however I'd check the other above
mentioned things as well.

YMMV,

DJay


Thanks DJay-

So far, I've dissassembled all the electrical connetions, made sure they
are clean, and reassembled. Inside the transformer box, the two MH
systems are connected through hardwired connections so unless I want to
unsolder some connections I can't test them individually.

Meanwhile I've also replaced the GFCI outlet itself (because this was an
easy thing to do, relatively), and removed from the circuit a UPS system
supplying backup power to the filter. The UPS lead-acid battery is
shot, and doesn't hold a charge anymore, so it was just taking up room.
I had a concern that the poor state of the battery might have been
introducing some sort of ground leak to the circuit, making it more
sensitive to issues when the MH bulbs fire.

I also changed the older of the two MH bulbs.

Since doing these things, I've had good operation for two days, but
since the problem was intermittent, I'm not sure I've sorted it out yet.
Time will tell! The next thing to do would be replace the caps in the
MH transformer box, but that would cost ~$60 so I'd rather hold off
until I know it's necessary.

Marc

"MAG" wrote in message
.net...
In article 7EaPe.9270$Vg7.2812@trnddc06, says...
I don't know about your tank but my electrical connections tended to
corrode
over a few year's time. You say you have had your tank up for 5 years.
Check ALL of the electrical connections for corrosion. Corrosion can
cause
"leakage current" which is what will set your GFCI into FREAK out mode
and
trip the circuit breaker. It may be working PERFECTLY.

ALSO, make sure there is no moisture buildup anywhere in the electrical
system. Afterall we are talking about a high humidity system that is
rather
close to water. Moisture buildup could cause a current LEAK somewhere
and
BAM! The GFCI is again working perfectly.

You may find that the GFCI is indeed week, however I'd check the other
above
mentioned things as well.

YMMV,

DJay


Thanks DJay-

So far, I've dissassembled all the electrical connetions, made sure they
are clean, and reassembled. Inside the transformer box, the two MH
systems are connected through hardwired connections so unless I want to
unsolder some connections I can't test them individually.

Meanwhile I've also replaced the GFCI outlet itself (because this was an
easy thing to do, relatively), and removed from the circuit a UPS system
supplying backup power to the filter. The UPS lead-acid battery is
shot, and doesn't hold a charge anymore, so it was just taking up room.
I had a concern that the poor state of the battery might have been
introducing some sort of ground leak to the circuit, making it more
sensitive to issues when the MH bulbs fire.

I also changed the older of the two MH bulbs.

Since doing these things, I've had good operation for two days, but
since the problem was intermittent, I'm not sure I've sorted it out yet.
Time will tell! The next thing to do would be replace the caps in the
MH transformer box, but that would cost ~$60 so I'd rather hold off
until I know it's necessary.

Marc

Marc,

Well if you "found" the problem then don't go looking further. You can't
help but monitor the situation as the days go by. If the breaker doesn't
flip then you found the problem.
I would get another AC outlet from a different breaker string over to your
tank though. I think another poster mentioned this as well.

Good luck,

Djay

In article dCvPe.27992$xp.23311@trnddc02, says...
Marc,

Well if you "found" the problem then don't go looking further. You can't
help but monitor the situation as the days go by. If the breaker doesn't
flip then you found the problem.
I would get another AC outlet from a different breaker string over to your
tank though. I think another poster mentioned this as well.

Good luck,

Djay


Hi again Djay!

Sadly, running new electrical connections isn't an option at the moment.
We're selling the house (it goes on the market Monday), so I'm
frantically trying to finish all the "make house look good" projects.
Opening a new electrical project is out of the question, and in a couple
months we'll be moving anyway.

The odd thing (to me, not being an electrical engineer) is that the
problem manifested so intermittently. Four times over a two-week period,
though admittedly the last two times were on successive days. And, it
would never do it repeatedly if I flipped on/off the timer switch. It
was as if it took time for the problem to "build up" overnight. This
lead me to suspect bad caps in the transformer box.

As days go by I'll see if the problem recurs. In 10 minutes the timer
should flip for today, and I'll see if I've got another day of success.

Marc