7W CFL lamps on PLC relay out

[theripley]

Our client have a centralized machine status indicator, with the use of CFL enclosed in a colored diffuser-acrylic plastic. Each CFL is connected directly to the relay output PLC (Mitsubishi FX2N-128MR-ES/UL). This machine status indicator was running good for the past 4 years, until some PLC outputs were damaged a few months ago. Initially, I thought, it was because of its blinking function as some CFL were programmed to function as blinkers. Since the existing PLC has no spare out, we recommend to install another PLC.

Unfortunately, after we connect it to the new PLC, it still damaged its output. I measured the current, it reads only 50 mA, & voltage is 220 Vac. With this, I tried using interposing relay, surprisingly it did not damage the relay.

Any idea what was the trouble?

 

[OkiePC]

Our client have a centralized machine status indicator, with the use of CFL enclosed in a colored diffuser-acrylic plastic. Each CFL is connected directly to the relay output PLC (Mitsubishi FX2N-128MR-ES/UL). This machine status indicator was running good for the past 4 years, until some PLC outputs were damaged a few months ago. Initially, I thought, it was because of its blinking function as some CFL were programmed to function as blinkers. Since the existing PLC has no spare out, we recommend to install another PLC.

Unfortunately, after we connect it to the new PLC, it still damaged its output. I measured the current, it reads only 50 mA, & voltage is 220 Vac. With this, I tried using interposing relay, surprisingly it did not damage the relay.

Any idea what was the trouble?

First of all, install circuit protection or relay isolation. Fuse each output with a quick blow fuse or use a GFCI to supply them. Use a $0.50 fuse or a $5.00 relay as your sacrificial part when the load goes ipeshat.

 

[Lancie1]

Any idea what was the trouble?

I am using CFLs in my household lighting, and when one fails, it can really make a big stink! I cut open the last one that failed and examined the little circuit board. Compact Fluorescent Lamps are not just a resistive load, unlike an incandescent lamp. CFLs have a small electronic circuit board inside the base, containing capacitors and inductors. That circuit board contains a switch-mode power supply with a driver transformer, AC rectifier, EMI filter, and half-bridge output inverter (making up an electronic ballast). It is possible for the CFL circuit board to fail as a short circuit, or simply to produce inductive currents that get fed back into the PLC output.
From a Phillips Electronics description of a 25 watt CFL circuit:

After start, L2 and C3 form a series resonance circuit which is able to generate a large voltage across C3. The worst case ignition voltage is about 900 Vpk for low temperatures. The combination of ballast coil L2 and igniter capacitor C3 has been chosen in such a way that the voltage across the lamp can exceed this high level while the current through the BUJ101AU is smaller than 1.5A. The circuit is able to re-ignite for mains voltages down to 150 Vrms.

If the PLC ouput is an AC transistor, it could be damaged after some time by the CFL power supply. If the PLC output is a relay, then the relay contacts can be damaged by arcing. As Paul said, use fast fuses. Another option might be to switch the customer lamps to some other type, neon or LEDs maybe. LED lights would require even less power.

 

[KB1GNI]

As a side note, I wouldn't think that blinking CFLs would do much for their longevity.

Another vote for LEDs.

 

[Lancie1]

Good point. Any fluorescent lamp, compact or otherwise, will last longer if the switching is minimized. During start-up, a high voltage level is needed to get a fluorescent lamp burning. CFLs use a high-frequency high-voltage output startup pulse, apparently in the range of 900 volts and 25 to 45 kilohertz. With that high frequency, there has got to be some ripple feeding back into the PLC output, even with the filtering.

Erika, I bet it was not your idea to use the CFLs for indicators!

 

[theripley]

Erika, I bet it was not your idea to use the CFLs for indicators!

It was a work by my former colleague, who happened to have resigned long ago. The client requests for a repeat order & since they, my former colleagues, had that design working for 4 years, I followed that design. Initially, I was hesitant about connecting the CFLs directly to the PLC output. But I was assuming that CFLs are resistive devices so I thought it will not be able to damaged the PLC out. Turned out to be a wrong move after all.

What bothers me is that when I transferred some CFL to the new PLC, it damaged the new outputs in seconds. While the other CFLs, which are still connected to the old PLC, still works good even now. We checked for a possible line-to-ground, findings: negative. I checked the ground-to-PLC output (when OFF), reading: 100 Vac. I checked, again, the ground-to-PLC output (when ON), reading: 100 Vac. Certainly, I have no idea where the trouble is coming from.

 

[theripley]

Another observation is some CFLs flickers when it is OFF. I tried researching about it & found out that CFLs are very sensitive to leakage current coming from a programmable switch. However, most programmable switch which leaks current are triac, not relay output. Please help me understand this situation.

 

[KB1GNI]

Is this a 220V or 110V system? In your first post you mention 220 but in post #6 you refer to 100.

If it is a relay output, and the output to ground is the same off or on, that sounds like the relay contacts are welded shut. However, if you have a floating neutral then all bets are off.

What do you measure from neutral to ground?

 

[OkiePC]

If it is a relay output, and the output to ground is the same off or on, that sounds like the relay contacts are welded shut. However, if you have a floating neutral then all bets are off.

What do you measure from neutral to ground?

I concur with KB1GNI.

I also concur with the recommendation to go back to [gasp] incandescent, "purely" resistive loads or LED's.

 

[theripley]

Is this a 220V or 110V system? In your first post you mention 220 but in post #6 you refer to 100.

Our supply voltage is 220 Vac (line-to-line). However, when I tried reading voltage from any PLC output to ground (line-to-ground), I read 100 Vac. This is true even when PLC output is on or off.

What do you measure from neutral to ground?

I am sorry if I got you confused, but I did not mean neutral-to-ground but line-to-ground.

If it is a relay output, and the output to ground is the same off or on, that sounds like the relay contacts are welded shut. However, if you have a floating neutral then all bets are off.

I conducted the following test when the PLC output is off:

1.) Continuity testing between output-to-common, result: no continuity
2.) Voltage reading from output-to-common, result: approx. 200 Vac
3.) Voltage reading from output-to-ground, result: approx. 100 Vac


I am confuse with the significance of the PLC output-to-ground voltage reading of 100 Vac. Please help.

 

[Lancie1]

What do you measure from neutral to ground?

KB1GNI, From previous posts and discussions with Ripley, the power system is 220 VAC, but no neutral - one leg of the 220 is grounded. Thus a PLC output should read some voltage to "ground" (the other 220 leg) whether it is on or off (it should be much more when ON). How much that line-to-ground voltage will be is dependent on the quality of the power line ground, and whether the PLC has its "common" grounded locally, or only at the power line service entrance point. With a perfect ground, and wet dense soil, the ideal voltage-to-ground (in this system) should be 220 volts.

2.) Voltage reading from output-to-common, result: approx. 200 Vac
3.) Voltage reading from output-to-ground, result: approx. 100 Vac

This would seem to indicate that the common is only grounded at the service entrance point from the utility company. It seems that you line voltage is also low, only 200 volts instead of 220. Is this reading with all the plant equipment running? What is the voltage at the PLC power supply input terminals? (In other words, is the plant voltage 220 at any other point?) If it is 220 volts, then something is pulling down the power to the PLC output terminals. You many have some shorted outputs on some of those CFL indicators. You could temporarily remove all PLC outputs one at a time, checking the voltage each time to see if it goes back to 220.

I tried researching about it & found out that CFLs are very sensitive to leakage current coming from a programmable switch. However, most programmable switch which leaks current are triac, not relay output. Please help me understand this situation.

Are there several CF lamps very close to each other, with some being on and the flickering ones (supposedly) being off? Remember that at high frequency and voltage (as on the CF lamp), the inductive voltage is much larger - the bulbs are radiating, so ones very near may get enough transmitted (through the air) power to flicker. A big problem with fluorescent lamps has always been the electrical "noise" they create in the air and on the power line. I have worked in buildings with the old-style fluorescent ballasts. Lots of us tried to listen to radios while we worked, but the electrical noise from the fluorescent lights blocked out most radio stations. To stop the flickering might require RL filters on the PLC outputs to the lights. The filters would need to be tuned to filter 35,000 Hertz.

The best cheapest long-term solution is to change the type of lamps (in order of preference for longivity and durability) to LED, neon, or plain old incandescent. With indandescent, you might have a problem with heat because to get the same light your lamp watts would need to be much higher. You might have to use a transformer to step down your PLC voltage voltage to match the new lamps.

 

[Lancie1]

Transmitting Power through the Air!

..the bulbs are radiating, so ones very near may get enough transmitted (through the air) power to flicker.

Before a bunch of you jump on me about transmitting power through the air, I want to explain. It is possible for power to be transmitted through the void, if the frequency is high enough. This is what radio transmitters do all the time. Although the received power at each receiver miles from the transmitter is very low, the total power being transmitted at the tower is very large, as shown by the size of the cables feeding the signal to the top of the transmitting tower.

I remember those old "crystal" raido kits that we kids ordered though the mail and put together. They had no battery, but pulled operating power from the air (radio signals contain power). I spent many nights listening with an earphone attached to a board with an antenna connection, a inductor/tuner, and a diode. The avialable power was not enough to drive a speaker.

There also have been many schemes and future plans to transmit household utility power through the air. All it requires is the proper frequency converters on both ends. Spillover noise would stil be a big problem.

 

[theripley]

Lancie1,

Thank you for your reply, as always.

Here are our other findings: (Please see attached diagram)
1.) We read approx. 130 Vac from L1 to PLC output (where flickering happens). When I took its connection from PLC (detaching P2 from P1), the 130 Vac reading is coming from that detached wire. Please take note that the bulb still flickers. Therefore, we took out the other end of this wire (detaching P3 from P4) and got the same reading with both ends of the wire detached. However, the flickering stops.


2.) We are reading approx. 130 Vac between L1 & PLC output (P4) when the bulb is taken out, but when it is in its socket we read approx. 60 Vac.

Please take note that this reading was taken when PLC output is off.

PS:
On the issue of flickering lights, I tried switching the flickering lamps with another brand, the flickering disappears. The previous lamp was a not known one (opple), I switched it with a Phillips.

 

[Lancie1]

Hmmm....I am not sure how your PLC and lights are wired. For example, normally your "L1" common for the lights (we use L2 here) would be connected to the grounded leg of your supply voltage. Where is that connection? I do not see it indicated on your drawing. Your ungrounded leg of the 220 volts should feed your PLC output common, getting switched to each light as that output goes on.

It seems that you might have your 220 volt grounded leg hooked to the PLC common output terminal. That is not usually a good idea.

Are these 220 volt CF lamps as I assumed?

 

[Lancie1]

From the (very poorly written) Mitsubishi FX2N Hardware Manual, page 6-5, I think your wiring should look like the attached picture (modified for clarity). The PLC "COM" terminals do not mean that this should be the grounded leg of your power supply, but instead is a internal connected common bus inside the PLC. If you have your grounded leg connected to the PLC common terminals, that means you are trying to switch the grounded-leg of your power supply (instead of the ungrounded leg).

In the U.S. National Electrical Code, that practice is a no-no. Trying to switch the grounded leg can lead to problems, and also will cause you to measure erratic and unpredictable voltages-to-ground at the PLC output terminals.