The Lamp Room

What causes a fluorescent tube to end up bright pink and mercury starved when it has been run on emergency lighting control gear in particular?

I have changed countless Bell branded 8w tubes over the years from magnetic ballast fittings and it's unusual for them to end up bright pink. This is from road sign lights, where the tubes often run 24/7 and are left to rectify until one or both cathodes breaks. Still, even when EOL, they can usually be resurrected and usually have the correct colour.

However, when it comes to emergency lights, it was common for them to end up bright pink which coincided with the cathode/s being blackened. What specifically is happening in the tube to cause the this? The tube in the picture hadn't just been turned on either, it had no break in operation from new as these ran 24/7.

Fluorescent lamps can glow like this on emergency lighting, if there is a current leakage from the battery.
The specific one in your picture, runs at cold cathodes and hence both ends are blackened.

The emergency light in the photo doesn't have an internal battery. It is powered from a central transformer and the fitting can run on 50/110v AC/DC. During a power cut, the central batteries power the fittings.

During mains operation, the fitting preheats both ends of the tube. If one cathode breaks, the power goes to the remaining cathode like this example.
These fittings always had high rates of mercury starved lamps once they reached EOL. Many would be left for a while at EOL as well.

So it is strange that both ends are heavily blackened and the lamp looks like this.

Possibly, the gear isn't driving the tube very well - cold cathode starting, or under or over driving it when lit.

Dor what was it about the current leakage from the battery ? Please explain



The way how i assume it goes :

The tube runs normally until normal EOL (depletion of emitter)

Once it EOLs, the ballast (EM ballast) output is still sufficient to keep driving the tube and additional bad cathode voltage drop at the normal current

This will happen more commonly with 8W tubes as the discharge voltage drop in them is very low, so it leaves plenty of headroom for additional voltage drop before it reaches the ballast capability limits (which would make the ballast drive at a lower current, if no EOL protections are in place to cut it off)

The tube will keep going till the other cathode EOLs too (independently)

And will keep going after that

Something related to the cathode sputtering affects the discharge. There may be materials (probably not the tungsten, but something else that was there ?) released from the overheating cathodes or overheating phosphor near them that contaminates the discharge, or that reacts with the mercury and causes the tubes to slowly go mercury starved

The contamination might make the discharge voltage rise, and as long as the ballast can keep up with it (due to the high headroom) this will make the power supplied to the tube rise, so keep the brightness higher vs. borderline mercury starving tubes on normal gear

Thanks for the explanation! Yeah, those emergency light didn't have any EOL protection. Tubes used to get roasted in these at EOL and they would be left glowing pink for a long time. I saw some interesting wear marks on tubes such as this Radium from a few years ago. It lasted incredibly well and was left to run until it wouldn't produce light anymore. It ended up with these dotted marks.

I seem to recall reading that fluorescent tubes aren’t particularly happy to operate on DC supplies (either by battery or dynamo).

On the older fluorescent lights as used on ships, the switch was a special rotary type which would cycle ‘on / off / on / off’ (In one direction only). Each time the switch was turned on, the effect basically was to reverse the polarity to the tube electrodes. This prevented the damage to the tube encountered on DC supplied fluorescent lamps.

In the case of mains powered emergency fluorescent lights, I’ve noticed that even new tubes on ‘non maintained’ circuits can blacken fairly quickly, even though they don’t operate except in the event if a power failure.

I was wondering whether it has anything to do with the presence of the little label which would say something like; ‘the pins of this lamp may be live, even if no lamp(s) are operating’.

Does the charging supply for the battery go through the lamp electrode(s) to indicate the lamp is healthy?

All present day EM inverters supply HF AC to the tube, which is the best for tube life

The problem is with the power - common EM inverters supply about 4W or so into a 8W tube without any cathode heating, so it runs cold but at a high enough power to sputter the cathodes fast

In fittings where the same tube is switched to line gear when the line voltage is restored, the gear will finish off a tube that got initial damage from the EM gear

In the case of mains powered emergency fluorescent lights, I’ve noticed that even new tubes on ‘non maintained’ circuits can blacken fairly quickly, even though they don’t operate except in the event if a power failure.

I was wondering whether it has anything to do with the presence of the little label which would say something like; ‘the pins of this lamp may be live, even if no lamp(s) are operating’.

Does the charging supply for the battery go through the lamp electrode(s) to indicate the lamp is healthy?

Yeah, many emergency lights drive the tubes quite poorly on the battery. I have seen exceptions though. My 1995 Wirefield bulkhead drives the tubes fine on the battery and so does my modern Virgo branded one with a HF ballast and relatively small battery. Yeah, I've seen the stickers which say something about the luminaire being live even when no lamps are lit. As you have noticed, some fittings preheat the cathodes constantly when running. I believe this is done to prevent the tube from cold starting if the power goes out. Maintained fittings do this as well when the tube is running on the mains. The pre-heating is only just enough to warm the cathodes slightly.

Some emergency lights even drive the tubes poorly on the mains. I've seen some with 1 wire to each end for the maintained running. I used to have a Robus fitting as well which suffered from the transformer failing. Eventually, tubes were only lasting a few days.