From my knowledge. Most if not all fluorescent lanterns run the lamps in cold cathode mode. Isn't this less efficient in terms of lumen output than hot cathode mode?

Also why is running lamps like this in cold cathode mode so destructive?


Has anybody here designed a circuit to run these guys on battery power in hot cathode mode? If so please post diagrams here I'd like to rework a couple lanterns of mine.

I'll let a more knowledgeable member such as James complete/correct my mumbo-jumbo, but basically cold-cathode ruins the electrodes. The exact physics behind it go beyond my knowledge, but it likely has something to do with the cathode's emissive substance being expelled from the electrode's structure or something like that.

Designing a proper battery-operated ballast for F4T5s is far from impossible. But it was often preferred to go in full cheap mode and make the tube light with as few components as possible. Plus the power drawn from the batteries was often lower as a result, extending their lifespan.

Why it is destructive? To get discharge, you need free electrons. To get them, you need to liberate them from the cathode material. One way is to provide high electric field, which practically pulls them out by force. That is how cold cathodes work. The side effect is, that high electric field also accelerate the ions from the plasma, which then hit the cathode material with high momentum (because they are rather heavy) and practically sandblast it off, while that released material land somewhere and darkens that region out, aka sputtering. There are various designs to both reduce tyat abrasion and to make the material to practically settle back. The other drawback of this type of electrode is tge high voltage drop across this region, which consumes power but does not generate light, but heat up the cathode instead (by the ion bombardment).

The hot cathodes use different approach: By heating the cathodes, the heat is what provides the electrons with the energy to get released readily, so immediately contribute to the light generation, eliminating the area with high field, so there is nothing accelerating the ions anymore. But for that the cathode has to be kept warm at the emission temperature.
If not, the heat won't be sufficient to liberate the electrons, so some of the energy has to be supplied from elevating the field. Normally cathodes are designed at a balance, when the field is low enough to make the ion speed low enough to limit sputtering, but not too high so the emission coating does not evaporate that easily. If you operate the lamp at too low current without providing auxiliary cathode heat, the temperature will drop and the sputtering will become faster.
The problem is, keeping the cathode warm also cost a power that does not generate light. And this tends to be lower percentage with longer lamps.


Why fluorescents are underdriven in battery lanterns?
There is a problem when designing rather low power battery lantern, lets say 130lm:
If you use 4W lamp and drive it full, you get full lifetime, but the efficacy would just be of the 4W lamp, so need to feed it by 4W from very expensive batteries (assume single use ones). So such lantern may be cheap on lamp cost (because lamp will last 5..8k hours), but rather expensive on battery cost (needs full 4W).
But if you use 8W rated lamp but drive it at 2..3W, the lamp cathode will be underdriven, so result in shorter lamp life, but consume just 2..3W from the batteries. So what you get is 30..50% savings on battery cost, at the expense of tge lamp lasting few 100's hours. That means you pay more for the lamp wear, but save many times more on the battery cost (don't forget 30% lower drain uses to yield 30% higher usable capacity with most batteries, so double the battery runtime). By thbe way thhe same is the reason why the lamps are driven at DC (the lamp itself is what rectifies the current): At DC only one side is the cathode, so heating only one side is enough. This by itself means about factor of 2 of total power reduction without any detrimental effect on the lamp life, so the 2..3W the F8T5 uses to be operated at means in fact operation at 50..66% of the rated load, so still fairly close to the rated hot cathode operation. It just takes longer to warm the electrodes up, so the life limitation comes mainly from starting (at battery power the turn ON time per start is a few minutes, not the 3 hours like the general lighting is rated for).

So for a lantern designed to deliver what F4T5 is rated to deliver, about 8W lamps are used.
The F4 and F6 are then used for even lower light outputs, so power levels around 1..1.5 and 1.5..2W. Bellow that the fluorescents do not make any efficacy sense anymore, you will become better with incandescents (all above is written from the perspective of 70's till 90's, before LEDs were available).
 

So say if I wanted to run an F4T5 at full wattage could i use the circuit board from an F8T5 lantern then rig in some sort of preheat circuitry? Or would I have to fully design a circuit from scratch to acheive my goal?

The way I see it. While obviously LED is the more practical choice. I would still like to utilize the mini fluorescent lantern (and I currently do) but the power to weight ratio of these lanterns is a little out of date. Nowadays we have large capacity lithium ion packs. And USB fast charging. Theoretically I could pump these lights at full brightness and still maintain the same lighting duration of the old alkaline lantern batteries if not even longer.. But personally I am no electrical engineer so I'm not entirely sure how to design the circuit. I know. How an AC preheat circuit works but would that method combined with the electronic DC ballast even function as I would want?

No, you would need to redesign the ballast. Just connecting a 4W tube onto a 8W inverter would fry it.
Easier option would be to add a cathode heating winding. That way you can maintain the cathode at operating temperature even with lower power. The winding should yield 3..4V, so should have corresponding turn ratio towards the primary (collector) winding and the battery voltage. Best way to determine is to thread one turn of wire through the transformer and using an oscilloscope to determine how much voltage correspond to that single turn (you have to look on the section when the transistor is ON...). And then recalculate it to get the desired 3..4V, I would expect something around 5..10 turns.
Remember, there is just one cathode, so you would have to identify which lamp end is it. Usually there uses to be some room to fit the extra 5..10 turns.
The cathode is the side, which goes negative when the transistor is switched off, so when there is an exponential like waveform section.

Otherwise there is a possibility to design a completely new full power ballast this lines, but you have to design tge exact voltsges according to what lamp you want to operate. But it needs at least 3-cell LiIon, otherwise the IR2153 won't start up.