Does anyone know the current specs of F13T5 and F14T5 tubes, thanks in advance.

From Osram catalogue:
L 13W (F13T5): Rated at 50/60Hz preheat, current 165mA, arc voltage 95V, preheat current 220mA
FH 14W HE (F14T5HE): Rated at high frequency (>20kHz), current 165mA, 85V, preheat current (programmed start) 210mA


The rating numbers (for the arc and preheat currents) may look the same, but the important difference is the frequency at which it is rated.
At 50 or 60Hz the ionization level is able to follow the current, so the voltage stays constant, so becomes rectangular. Because of the shape mismatch, the power factor of the arc is below 1, so the real power is less than Volts*Amps. For a sinewave current the power factor is then 0.9, if the current shape becomes more spikey (higher crest factor), the power factor becomes even lower.
At 20+kHz the ionization can not follow the current and so stays constant, so the voltage shape follows the current shape, yielding unity power factor.

So when a tube is rated for sinewave current 165mA/95V at low frequency, the corresponding high frequency rating would be no more than 148mA for the same power, counting for the higher efficacy at high frequency the rating becomes 135mA for the same output.

So F13T5 rating for high frequency (>20kHz) would become 135..150mA/95V

The F14T5HE will exhibit shortened life at 50Hz, because the electrodes are not designed for sufficient thermal inertia to average out the slow mains cycle, plus the lower frequency would ask for higher rms current in general, leading to filament overheating.

Thank you for the answer, I was wondering if F14T5s could be used on preheat F13T5 ballasts, but I guess I'd have to figure out how to make a ballast that is similar to the ones in my Craftsman PL-S floodlight (so something like this one), unless those wouldn't work either.

Those ballasts seem to have still good printed circuit boards (no heat damage,...), so you may definitely build a new ballast on these. But I bet you should be even able to fix those. To me the solder joints there look suspicious: Like crack had formed, mainly around the transistors (they are "flapping in the breeze") there. Not good, mainly when exposed to mechanical shocks or vibrations (what a worklight definitely is), the transistor weight poses quite large load on these joints and they tend to crack over time.

What I really hope to do is to reverse engineer the design, so I could possibly build bigger versions for higher wattage tubes.

You could run a F14T5 HE fluorescent tube on a F13T5 or 10/13w G24D-1 PL-C preheat HX autotransformer ballast at the expense of shorter tube life, but there are some conversion adapters that allow you to use F14T5 HE fluorescent tubes in standard F20T12 preheat fixtures at least on ballasts with an OCV of at least 220v, but will need to give some a test on a 120v F20T12 choke and I will report the results in this thread when I get the chance.

Update: the F20T12 to F14T5 HE adapters that I received today actually work pretty nicely on 120v F20T12 preheat choke ballasts, but they underdrive the F14T5 tubes quite a bit and they seem to cause the tubes to exhibit some bit of flickering. The reason that they managed to work on 120v F20T12 preheat choke ballasts was because they were designed with 2 lamp series circuits in mind that would run off European 220v-240v F40T12 preheat choke ballasts. They do contain an integral ballast to convert the mains frequency oscillations to high frequency oscillations. Here are the results of using my European 2 foot T8 to T5 HE adapters on 120v F20T12 preheat choke ballasts:

https://www.lighting-gallery.net/gallery/displayimage.php?album=lastup&cat=0&pos=32&pid=208402

In this photo, they are running in my 2 lamp F20T12 preheat halfpiper adapter.

Aren't these designed to operate directly on the 120V? Or even 230V? Both could explain the underpowering when on 120V with the "20W" series choke...

If the series conection on 230V is really possible, I doubt they would behave in any way different than on the 120V with the choke.

Other issue could be the 120V choke: Many are rated for wide plethora of lamps (often F10T8, F15T8, F20T12, FC8T9,... all on a single choke) and actually are quite severely underdriving the F20. Their rather high impedance then means the F14T5 ballast getits input current restricted so lower than rated voltage on its input, leading to lower than rated power output...

In theory, a lamp that is operated on a single lamp preheat circuit on a certain OCV, it also “sees” the same OCV when it is operated on a 2 lamp series preheat circuit with an OCV that is twice as high compared to the OCV of the single lamp preheat circuit. Because of this, I believe that the lamps will also operate at a similar brightness even when these lamp-adapter systems operate in a series pair off a 220v-240v choke ballast or an autotransformer ballast with a 220v-240v OCV.

Aren't these designed to operate directly on the 120V? Or even 230V? Both could explain the underpowering when on 120V with the "20W" series choke...

If the series conection on 230V is really possible, I doubt they would behave in any way different than on the 120V with the choke.

Other issue could be the 120V choke: Many are rated for wide plethora of lamps (often F10T8, F15T8, F20T12, FC8T9,... all on a single choke) and actually are quite severely underdriving the F20. Their rather high impedance then means the F14T5 ballast getits input current restricted so lower than rated voltage on its input, leading to lower than rated power output...

In my case, I was using these adapters on 120v F20T12 preheat chokes that drive F20T12 tubes at full power and they were only rated for F20T12 fluorescent tubes in my case.