- Reaction score
- 13,478
The lighting on my boat includes about thirty T5 8W 12" tubes in conventional 12V fittings of good make (mainly Labcraft) with HF inverters, taking one, two or three tubes as required. For over 25 years these have worked more or less flawlessly; the only outright failure was when water got into a fitting and blew the inverter transistors, which were easily replaced. A year or so ago there was a spate of tube failures, often with their ends blackened and overheating and melting the plastic housing. I put this down to a bad batch of cheap Crompton-branded tubes that had been installed over the previous few years and replaced them all with Sylvania 6000-hour 400lm white 535 tubes. Now they are failing again, not overheating this time but one or both ends blackening / low output / flicker / poor starting.
This time, I suspected the fittings, so I took one down that had been working perfectly before developing one badly blackened tube end, and tested the output capacitors of the inverter. These should prevent any assymmetry in the drive waveform creating a DC component in the tube current that would cause mercury and eroded cathode material to migrate. Disappointingly, the capacitors (and inverter) seem perfectly OK. It might be happening more on the fittings that rely on emission-heating of the cathodes; there are no filament windings on the inverter transformer, the two pins at each end are paralleled and a high open-circuit voltage is used to strike the tube cold. However, the other type of fitting does not get as much use so it's difficult to compare their behaviour.
What is happening here? Are current production tubes shorter-lived than they used to be and the 6000 hours is now rather optimistic? Is some change in manufacture (e.g. reduced mercury content) making them unsuitable for emission heating? Have Sylvania gone downhill and should I put in GE or Philips or what? I could convert the lot to LED but I don't like the LED drop-in replacements for these fittings very much, nor the T5 LED tubes which tend to be either expensive or give poor light quality. Ultimately I'll probably fit my own variable colour-temp (dual-white) LED tape inside these fittings but I'm kind-of busy at the moment and would prefer to spend 20 minutes swapping the tubes one last time, getting a couple of years out of them and then doing the lot properly.
This time, I suspected the fittings, so I took one down that had been working perfectly before developing one badly blackened tube end, and tested the output capacitors of the inverter. These should prevent any assymmetry in the drive waveform creating a DC component in the tube current that would cause mercury and eroded cathode material to migrate. Disappointingly, the capacitors (and inverter) seem perfectly OK. It might be happening more on the fittings that rely on emission-heating of the cathodes; there are no filament windings on the inverter transformer, the two pins at each end are paralleled and a high open-circuit voltage is used to strike the tube cold. However, the other type of fitting does not get as much use so it's difficult to compare their behaviour.
What is happening here? Are current production tubes shorter-lived than they used to be and the 6000 hours is now rather optimistic? Is some change in manufacture (e.g. reduced mercury content) making them unsuitable for emission heating? Have Sylvania gone downhill and should I put in GE or Philips or what? I could convert the lot to LED but I don't like the LED drop-in replacements for these fittings very much, nor the T5 LED tubes which tend to be either expensive or give poor light quality. Ultimately I'll probably fit my own variable colour-temp (dual-white) LED tape inside these fittings but I'm kind-of busy at the moment and would prefer to spend 20 minutes swapping the tubes one last time, getting a couple of years out of them and then doing the lot properly.
Last edited: