Electricians or ballast experts help needed!!!

[zanemoseley]

I made a DIY dual 250 MH retro about 2 years ago. It consists of 2 probe start ballasts in what they call a transformer can which is just a big electrical box.

About 2 months ago I lost a GFCI and since it was a fairly new one I just thought it was cheasy and replaced it. I also put my lights/heater on a separate GFCI then the pumps so if they tripped I'd still have water flow. After first GFCI fried I got a new one and put the pumps on it and put the lights on a old yellow single outlet GFCI that had been proven good over 1.5 years.

Now today I was in the fish room after the lights have been on about 6 hours and the GFCI kicked. I tried to reset it and it didn't wanna do it. So I unplugged it and plugged it back in and it reset, if I remember it may have tripped again and I switched it back on but can't remember. The MH's were having a hard time refiring since they had just been turned off recently. As I was flipping the power switches for each ballast to give them a rest one fired on. At this point one switched on and one was off, I figured since they one was on I'd try the other again and flipped the switch, immediately the GFCI flipped and wouldn't reset again. I tried unplugging it and no luck. I unplugged the power strip from it and plugged in in another outlet and it makes these faint but rapid clicking noises like the other broken one did.

Am I screwed and gonna have to buy new ballasts? They do run hot since they were raw ballast in a metal container with just some vent holes. I did have the ballast box elevated a bit for air to get underneath it. They were hot but have been fine for 2 years like that. Previously it didn't trip the GFCI enough to really troubleshoot the problem.

I'm really kinda nervous plugging the strip straight into the wall in fear it may cause a wiring problem. I'm wondering why I never had a breaker trip though?

 

[AllenFord_SC]

Sounds like an accident waiting to happen to me. I would go for the new ballast as opposed to a new home due to electrical fire.

 

[AllenFord_SC]

BTW I am in no way a ballast expert or an electrician. Unless all the times I have been zapped counts.

 

[cee]

You've got something shorting to ground somewhere. Bad insulation in the ballast wires due to overheating, I'd suspect, or water somewhere.

 

[zanemoseley]

Here's a picture of both grounds.

The first one is the one that tripped the GFCI the final time when I flipped its power switch on. BTW both ballasts have independent power switches.

Here's the whole box.

Hey Dave one qustion, why would it have blown the GFCI if it was just a bad ground, shouldn't it have just tripped it?

 

[zanemoseley]

BTW if you see burn looking marks on the edge of the ballast its just weld marks where they weld the plates together.

 

[Tammy3770]

the transformer on the right has a bright spot, is that the picture?
ok wait did you say you tried the ballast in two different oultets??
and it did not work in either outlet??

do other things trip your GCFI?? or just that ballast?

 

[zanemoseley]

Thats a shinny spot in the picture.

And I tried the GFCI in two outlets not the ballast, the ballasts work they just randomly seem to pick a time to short out or something and killl GFCI's.

I'm seriously thinking of rewiring the entire ballast box tommorow.

Any other ideas???

 

[zanemoseley]

Ohh and no other things that I know of trip my GFCI's however before I split the lights/heater and pumps onto different GFCI's they were all on one so it was hard to tell what tripped it. Tonight was the best situation for me to get an idea of whats going on with my GFCI scenario.

 

[cee]

The case of the capacitors are grounded (although not as well as they should be), as is the transformer core. You couldn't see ground fault evidence if it occured inside either of those components. Also, the first ground lug looks like it's gotten a little hot in the past (black charring, evidence of green insulation melting). I would suspect that transformer.

GFCIs trip faster than breakers. Breakers can actually let an incredible amount of current through for several 60 Hz cycles, more than enough to kill you many times over. Hence the GFCI.

Dave

 

[Tammy3770]

ok the wireing is hard to follow in that pic so I cant tell if its wired right.
the wires look good to me.
the insulation on the grounds show signs of what you said about heat building up due to no exhaust holes. however they still look good.

check those resistors,
But my initial guess would be that one of the transformers are going bad. the insulation between the laminates have a weak spot that worsens with the heat.

you are lucky to have a heads up
sometimes they go out slowly and sometimes they just go up in flames.

you can check the transformers but I do not suggest you do that because you are not an electrician. you could megger them but I doubt you have a megger.

 

[Tammy3770]

p.s.
one eays thing you could do is take the wire nuts off one by one and make sure there is no corrosion and that the wires are making good contact inside the wire nuts

 

[zanemoseley]

Dave, the capacitors aren't grounded, they're just velcro'ed to the sides of the case. There is nothing to ground them with unless you manually solder a wire from them to the case.

Tammy: There is "some" ventilation, if you look close on the left side you can see some ventilation slits. Also there is no transformer, the run at 110V. The large things are the ballasts and the small deals are the capacitors for the ballasts to fire.

I have a feeling you may be right about the laminate between the layers of copper for the ballast core. These things are made in china and probably don't like the heat much either.

 

[Tammy3770]

um.. I wont argue with you
good luck and I mean that sincerely

 

[zanemoseley]

I'm not arguing with you I was just letting you know what they consist of. They don't run off 12v via a transformer they run off 110v. I also just letting you know that there is "some" ventilation and its not a sealed enclosure. I really appreciate your help and don't know why you thought I was arguing with you.

 

[amoredelanotte]

troubleshooting problems the best thing to do is process of elimination

1 check all wiring from sockets to ballast look at sockets
closely to see if any saltcreeps are around it .. this will heat up
transformers

2 using a voltage tester check the voltage on both
ballast aka transformers if they are the same ( plus/minus ) a few volts
try swopping out the capacitors from one transformer to the
other . if lights go on then its the cap that needs replacing
if not then its the transformer .... this works when u dont have
a voltage tester

3 assumming that there are 2 grounds ? i take it u have
2 power cords ? remove both grounds from the transformers
and mount it on the case/box thus creating a chasis ground
remember GFCI are case ground sensitive to voltage or frequency
transformers create magnetic fields which all differ from one another fields/eddy current / frequencys which will cause
the GFCI to trip if u have to cords use the same GFCI recepticle

use ground probe in your tank ?

2 options for GFCI
1 use GFCI circuit breakers <~~ this is great but also depends if u are using dedicated cicuits
2 use above grade GFCI recepticles . ( hospital grade )

ps nice constructed case although i would use a fan with a thermal switch to ( cool down temps ) home or building type
metal halide fixures use one tranformer per fixture enclosure
we reefers use 2 in a small enclosure with fins or with out i still feel its best to have a fan to cool them down.

hope this helps ? any question pm me Maurizio

 

[amoredelanotte]

try using band iron . to support the caps to the case at the same time grounding them down . band iron is a thin metal strip with
1/4 holes every 1/2 u can get this at any plumbing or hardware store
u will need 4 bolts/8 washers 4 nuts and have to predrill holes in the box

 

[cee]

Zane:

Ground the cap cases as stated above.
The transformer tied to the charred lug is experiencing ground faults, which is heating up the lug to wire interface and causing the charring. Do you have an ohmmeter? Measure resistance between the 120 V windings and the transformer core. Do the same on the "lamp" side of the transformer. Both must be done with the unit unplugged and no lamp connected to the transformer (the cap won't matter is it's M57/M58/M59 ballast). Repeat for the 2nd transformer. It may not show it, but I suspect you'll see a difference in resistance due to an internal fault, which Tammy explained better than I. A megger or hipotter is a better way to go, but I doubt you have access to one and you don't want to use one unless you know what you're doing. Most importantly, if you measure it LIVE, remember these things put out about 350 V which can and will KILL you. Be safe!

Dave

 

[merlock116]

take your ballasts and a gfi to another part of the house on a different circuit (which probably has a much smaller amp draw). try them there and see if the gfi still kicks off.

 

[zanemoseley]

The good news is that the transformer is already case grounded in a sense since those bolts in the corners go straight to the case.

I'm gonna rewire the whole thing today by just cutting off the last 1 in of each wire and restripping it. I'll get some strapping soon but I'm really busy with school till Wed. I don't have an ohm meter nor do I really know what are the 120v windings or the "core". Maybe I could borrow one from school.