Slightly O.T Interesting Light bulb controls

I had a new one for me today. I got called to an X-Ray machine today because it wouldn't turn on. The X-Ray is complicated but the actual controls to turn it on are quite simple. A start - stop station run in series with customer interlocks (ours) and then to a contactor and two lights to indicate power on or power off.

Not having a drawing I started probing for voltages on the pushbuttons to find that the PB's, lights, and contactor all terminated on an unlabled PCB.

Looking at the PCB all I could find was a transformer, a couple of relays and some resistors. I was getting frustrated as to why those "(insert your own explicitive here) Europeans" would complicate something that should be simple with a mysterious circuit board.

Turns out the purpose of the board was to sense when one of the on or off light bulbs was burnt out and to prevent energizing the contactor if that was the case.

In as safety sensitive application like an X-Ray I think that is a great idea. I've just never seen it before. It solves the whole "what if the light bulb burns out" scenario.

Maybe I have been living in a hole the last 10 years. Is this common in the controls industry?

I am looking for either yes I see this all the time or no that's a new one on me responses.

If I am not mistaken, this is quite common in safety muting applications - i.e. in case when a section of a light curtain beam array has to be temporarily disabled, or "muted", to allow product entering the machine, for example. The muted state must be indicated by a lamp and it is important to assure that the lamp is good.

So the muting controller will be checking for the burnt lamp and will keep the guards in safe state when the filament is burnt.

I've been trying to figure out how I would implement this with a PLC and I haven't come up with a simple DI/DO solution, any ideas?

I reckon you need to have some kind of current sensor for that.

I have seen this in muting applicaitons. The Siemens safty relay I use for light curtain muting has an output labeled "Bulb". I had to read the manual before it hit me what it was for. I know saftey is saftey but I would think we could put in LED bulbs and eliminate such a circuit. I wonder if LED's would work? The electronics to do the check might not work with LED's. Anyone have a circuit to check for bad lights?

A mad time of the morning to come up with dippy ideas

allscott said:

I've been trying to figure out how I would implement this with a PLC and I haven't come up with a simple DI/DO solution, any ideas?

Click to expand...

How about a 12vdc lamp in series with a 12vdc relay coil from a 24vdc source then use the relay to switch your DI?

If either the lamp or coil fail = no op

silva.foxx said:

How about a 12vdc lamp in series with a 12vdc relay coil from a 24vdc source then use the relay to switch your DI?

Click to expand...

Well, it might be a bit hard to find a relay with coil resistance equal to the resistance of the lamp filament. It might work but I would stay away from such circuit.

LadderLogic said:

If I am not mistaken, this is quite common in safety muting applications - i.e. in case when a section of a light curtain beam array has to be temporarily disabled, or "muted", to allow product entering the machine, for example. The muted state must be indicated by a lamp and it is important to assure that the lamp is good.

So the muting controller will be checking for the burnt lamp and will keep the guards in safe state when the filament is burnt.

Click to expand...

We have that exact setup on our light curtains for our ABB robotic palletizer. It made things kinda fun when the bulb died after a couple of years, and no one had ever looked at the light curtain before.

Personnally, I'm all for a piece of equipment that works reliably enough that no one remembers how to work on it.

LadderLogic said:

Well, it might be a bit hard to find a relay with coil resistance equal to the resistance of the lamp filament. It might work but I would stay away from such circuit.

Click to expand...

How about a normal 24 VDC lamp, in series with a 5.1V zener (on the common side). Then take the signal across the Zener into a TTL input?

I don't use any such safety circuits, but then again, in 6 years of only using LED Pilot lamps, I haven't had a single one burn out.

I have several xrays and that is exactly how they work. If the light does not come on, then the tube can not be energized. This is a safety warning so you have a visual cue that there are xrays present. Unfortunately, I don't have the board level schematics to see exactly what they do, but this is extremely common in the xray field.

As a side note, anyone who deals with high voltage switchgear can show you a similar application. There is a RED light on the front of the switchgear that is ON when the breaker is closed. This light does more than just tell you the breaker is closed though. The circuit is routed through the trip coil of the breaker proving the circuit. That gives you an indication that the breaker CAN trip electrically. If you walk into a switchgear room and see no red lights on the cubicle and the breaker is closed, you should check immediately to make sure that it is a burnt out bulb. Otherwise, if there is an emergency and the relay calls for a trip, the breaker will NOT trip and the circuit will open when it burns through. I have seen several instances of this. One where the 125 VDC was off and several large transformers exploded.

a lightbulb when it burns out usually creates an open circuit so you should be able to wire it the same as a standard safety. The xray machine we have also has the lightbulbs in our case we use the AB stacklamps although in our case if the yellow or red isn't working we can use the green bulb if we happen to be out of stock.

Drive the lamp with a Sinking (NPN) output. The lamp has +24 on it all the time. You can sense the return side of the lamp when it is off. It should have 24 volts if the filament is intact.
This is easy on the PLC side. Tie the sensing input to the output (one input and one output jumpered together).
This is how your car computer knows a cable or sensor is disconnected under the hood (Bonnet).

Sorry Keith, but there's no need to check a light bulb when it's supposed to be off!

Too often light bulbs die the very moment they're energized. So the detection would not have noticed the dead light bulb until the next time it was turned off, and that's too late. Damage (or injury) may already have occured!

Checking current when energized is the correct way...

allscott said:

I had a new one for me today. I got called to an X-Ray machine today because it wouldn't turn on. The X-Ray is complicated but the actual controls to turn it on are quite simple. A start - stop station run in series with customer interlocks (ours) and then to a contactor and two lights to indicate power on or power off.

Not having a drawing I started probing for voltages on the pushbuttons to find that the PB's, lights, and contactor all terminated on an unlabled PCB.

Looking at the PCB all I could find was a transformer, a couple of relays and some resistors. I was getting frustrated as to why those "(insert your own explicitive here) Europeans" would complicate something that should be simple with a mysterious circuit board.

Turns out the purpose of the board was to sense when one of the on or off light bulbs was burnt out and to prevent energizing the contactor if that was the case.

In as safety sensitive application like an X-Ray I think that is a great idea. I've just never seen it before. It solves the whole "what if the light bulb burns out" scenario.

Maybe I have been living in a hole the last 10 years. Is this common in the controls industry?

I am looking for either yes I see this all the time or no that's a new one on me responses.

Click to expand...

Without being contraversial - european and american standards of safety are quite different !