I have an actinic f40t12 24 inch bulb and I'm wondering what makes it 40 watts. This bulb is identical to the other f20t12s physically, but not internally (excluding phosphors). Is there a difference in gas pressure, Hg amount, gas fill, electrodes, etc that set these two types of 2 foot lamps apart?

It could be the same gas as the F20, just higher current electrodes.
You may figure it out (but take care of covering, due to the UV...):
Connect it to some F20T12 ballast and measure the arc voltage. If it is about 60..65V, it is indeed just heavier electrodes for 0.75A.

 

It could be the same gas as the F20, just higher current electrodes.
You may figure it out (but take care of covering, due to the UV...):
Connect it to some F20T12 ballast and measure the arc voltage. If it is about 60..65V, it is indeed just heavier electrodes for 0.75A.

I'll have to go by the harbor freight store since I have a coupon for a free multimeter.

And one extra note for multimeter measurements: Unless the multimeter is of the "true rms type", it may read the lamp arc voltage quite off, so if you get something in the +/-20% range, it probably is correct. The reason for the error is, the multimeters are calibrated to read values equal to rms value on a pure, ideal sinewave signal.But the voltage across the lamp is more of a rectangular shape. That shape difference means the rectifiers in the meters will convert to a different DC value.
 Internally each AC meter contains some form of rectifier converting the measured voltage (after a range select divider) to DC, as only the DC value could be then converted to the displayed number.

And the rectifier design influences, how is the AC converted to the DC. The "true-rms" meters do really what the name says: First make a square of the input signal (using a multiplier, feeding the signal to both inputs for the multiplication), then filter it (so the "mean" function) and then do the square root conversion (another multiplier in an opamp feedback). All this is usually within one IC, but that IC at first cost more money, but as well needs more power (so drains the battery faster) than a simple opamp with two diodes (doing just the "rectified average") or even just a diode with some resistor (does something between average and peak; present in the simplest multimeters, where the only AC ranges are the 200V one and one higher, usually labeled as something between 400..1000V, all other ranges are just DC).

2x 20 watt choke will do fine in perallel

A cheap multimeter can indeed tell lies, but it will tell lies with a F20T12 as well, and in the same "direction". So just compare the 40W to a 20W using the same multimeter

Let lamps warm up. The arc voltage goes up as the lamp warms, in the 36W T8 i tested it went from under 90V just after striking to over 110V after proper warm up

And i would not connect the multimeter before the lamp struck, i can imagine a multimeter "blasting off again" from starting pulses on Preheat, especially a cheap one

Thanks, Ash. I've tried the bulb in an f20 fixture. It lights, but only if the cathodes warm and then I remove the starter.

I tried a GE and an osram starter and the osram worked the best and I actually didn't need to remove it, It flashed for about ten seconds and then it stabilized.

Interesting, I didn't know you could get 2-foot F40T12 lamps here.

Based on the specs of a couple of European 2-foot 40W ballasts I've seen on here, the lamp is approximately 52V by 0.9A. So three 14/15/20 chokes in parallel would do the trick.  A 150 watt incandescent bulb would work great too. So would an F24T12/HO ballast. And I think a 35 watt HPS choke would work well, too.

It is an aquarium actinic lamp. It is basically a HO version of its 20w counterpart.