Maker Pro
Maker Pro

Troubleshooting: LED lamp

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
Hello, I work on a ship with a floating ground, 480vac wye, no neutral.
I have installed LED illuminated push button and indication lamps on a switchboard, the source for the lamps is thru a transformer 480:120 Vac X1 is bonded to the switchboard.
The problem; the LED lamps are dimly lit when they should be off, I am trying to learn why the LED is lit, and how to properly solve the problem.
Richard
 

Tha fios agaibh

Aug 11, 2014
2,239
Joined
Aug 11, 2014
Messages
2,239
Whats the specs on the led push button?
So your control circuit is supplied with 120v? Is there a small transformer integrated into the pushbutton to drop the 120v to a smaller voltage for the lamp?
What type internal lamp?
 

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
CR104PBL11L3L2 its A GE extented gaurd momentary pushbutton, Full voltage lamp holder. 120v
BA9S120L LED Lamp bayonet style single contact ,120v ac/dc.
 

KJ6EAD

Aug 13, 2011
1,114
Joined
Aug 13, 2011
Messages
1,114
There may be an arc suppression capacitor in the switch that's providing a sneak path for AC.
 
Last edited:

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
If there is an arc surppression Cap. What is the best option in defeating it? Removal or insert a diode.
 

cjdelphi

Oct 26, 2011
1,166
Joined
Oct 26, 2011
Messages
1,166
If it's dimly lit, you might get away with throwing in a diode, that should stop the dim...
 

Tha fios agaibh

Aug 11, 2014
2,239
Joined
Aug 11, 2014
Messages
2,239
Sound like leakage current from the controller is partially turning it on. A bleeder resistor between the output signal and common would remedy this but could be a pain if there are a lot of lights to do. You might consider just relamping with a conventional incandescent pilot lamp. (Should fit in that light)
Suppression usually isn't added to these types of pilot lights. The light and contact blocks are designed as one component so it'd surprise me if that's the case.
The diode sounds like a bad idea because it would cut half of the wave.
So it might slightly cut down the glow but it'd also cut the incoming 120v in half when fully on.
What voltage do you read across the light when it off and slightly glowing?
How much current does this light draw?
 

KJ6EAD

Aug 13, 2011
1,114
Joined
Aug 13, 2011
Messages
1,114
I suspect that your panel is designed with a dual input lamp test feature and had neon bulbs installed originally. If so, you should consult the panel designer before upgrading to LED.
 

Tha fios agaibh

Aug 11, 2014
2,239
Joined
Aug 11, 2014
Messages
2,239
I suspect that your panel is designed with a dual input lamp test feature and had neon bulbs installed originally. If so, you should consult the panel designer before upgrading to LED.
The test feature is just how you wire the contacts. The part number CR104PBL11L3L2 indicates a full voltage 120v Led lamp. I think originally it had incandescent lamps.
Neon lamps don't need much current to start glowing, similar to Leds.
 

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
The panel was originally set up with Incandesant lamps, I changed the lamp fixtures to extented gaurd push button LED for the simple reason that LED lamps last 10 times longer than incandesant, and if you think the engineers that monitor that panel can let you know when they run low on lamps or change them themselfs you would be mistaken, of course the dim glow is a nuance. I think I will try a the resistor once I return to the ship. I greatly appreciate your input gentelmen. I'll think of some more question for you all.
 

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
the specs list it as 120V ac/dc 1.4watts
the LED is white so that is 3.6V @ 20mA to activate
120 - 3.6 = 116.4 / .020 = 5820
If I figure it correctly the LED internal resistor is 5.8Mohm
116.4 * .020 = 2.3 watt

.025 / 120 = .0002
.050 / 120 = .0004
1/4 watt resistor can handle 2uA @ 120v
1/2 watt resistor can handle 4uA @ 120v

at 1/4 watt every 5k resistance is 1V drop
at 1/2 watt every 5k resistance is 2V drop

So I am thinking a 1/2 watt 5k resistor
 

Tha fios agaibh

Aug 11, 2014
2,239
Joined
Aug 11, 2014
Messages
2,239
If it is rated 120v and 1.4w then that is what I would base it on.
1.4/120=.012A is pilot light current.
Better yet measure it with an ammeter.

I was proposing a pull down resistor to mitigate leakage current form your controller. That would be wired in parallel with the light. You are calculation is for a dropping resistor in series with a 3.6v light to limit current. Two different things.
I would measure the off state voltage at the light. Thats what you want to drop.
Eg; Output signal to neutral.

Anyway your 1/2watt rating is too small.
120/5000=.024X120=2.88 or almost 3watts.

I don't know what the leakage current is or the voltage you want to drop but you probably want to start with a higher resistance like 20k at least 2watts.
I await the input of intellect from the other esteemed forum members.
 

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
I would measure the off state voltage at the light. Thats what you want to drop.
I did not think about that, it's so obvious.

That would be wired in parallel with the light.
Do you mean to Shunt the lamp with a high value resistor? Or parallel to ground by taping off the Neutral Pole and bonding to ground?
I hope this drawing shows up. Is this your proposal? adding in the ground and resistor off the neutral?
upload_2016-1-28_7-53-14.png
 

Tha fios agaibh

Aug 11, 2014
2,239
Joined
Aug 11, 2014
Messages
2,239
Yes, I'd shunt the two contacts of the light from the 120v (hot) side to the common (white wire?) don't go to ground. I'd put the high value resistor right across the back of the light.
You probably want to make sure whomever has authority over the ship is ok with these modifications.
It's always wise to cover your arse.

The important thing here is sizing the resistor correctly. Also make sure the resistor has lots of clearance since it may get very warm. You don't want to risk melting anything around it.
 

RBeersJr

Feb 4, 2014
10
Joined
Feb 4, 2014
Messages
10
I see what your saying, I'll build a mock up this circuit in the shop and monitor the current and temperatures before installing this. It is the switchboard after all.
Thank you for your help, I'll let you know the out come of this in about 4 weeks.
Richard
 
Top