@RRK
Thanks so much, I think there is a decimal point between the 6 and 5 but the quality is so horribly low that it isn't very visible.

A dangling 1st Generation (Flat Bottom) Westinghouse OV-25 in Morton, TX. The refractor, reflector, and lens are gone, exposing the socket and ballast from the upper housing.

https://www.google.com/maps/@33.7270287,-102.7592095,3a,15y,57.33h,95.63t/data=!3m8!1e1!3m6!1s6l7obgTR3eWSWABx548PzA!2e0!5s20130301T000000!6shttps:%2F%2Fstreetviewpixels-pa.googleapis.com%2Fv1%2Fthumbnail%3Fcb_client%3Dmaps_sv.tactile%26w%3D900%26h%3D600%26pitch%3D-5.632944148939899%26panoid%3D6l7obgTR3eWSWABx548PzA%26yaw%3D57.333883606559866!7i13312!8i6656?entry=ttu&g_ep=EgoyMDI1MTIwOS4wIKXMDSoASAFQAw%3D%3D

in 2007 my family installed CFL bulbs on vanity light fixtures in the master bathroom to replace old incandescent globe lights. One night in 2018 I was taking a shower in the bathroom when one of the CFLs started flickering, and it even made popping noises. I got out to see the EOL show and the bulb self extinguished after 2 minutes after heavy flickering. Toward the end there is an orange glow in the base before the bulb went dark, but I was not expecting a burnt electrode smell afterwards. I took the bulb out and it was a dollar store off-brand CFL bulb.

BTW there is a small error in your circuit. You are missing a comma for 884 cathode bias voltage value. It is 6.5V, not 65V as on your schematics, though 20K / 650 Ohms values for bias voltage divider are correct.

The circuit is generally nothing more than a classic NE-2 neon lamp flasher, except the fact that 884 allows ignition voltage threshold control by a grid voltage (go google a 884 datashhet) and thus having X-sweep to sync.

2M resistor simply adjusts capacitor charging current (from 450V bus) for sawtooth.

This is a very crude X-sweep sawtooth oscillator made of 884 thyratron tube and adjustable RC network. R is 300K + 2M potentiometer, C is selected with a rotary switch.

So, 2M resistor adjusts X-sweep rate within selected timing capacitor range.

Yes, I did notice that the aluminum windings on the one I disassembled looked pretty thick. I don't think iot did much though because the top of the light fixture when running for a while was still too hot to touch.

So essentially both ballasts run lamps at similar power, but since the commercial version is CWA (more constant-current than HX) it will overdrive energy saving lamps. Very cool.

The commercial ballasts generally have copper/thicker aluminum windings so they don't heat up as much.
Most of them are high power factor, because commercial/industrial customers get billed for power factor as well.

The reason a F25T12 cannot be run on a commercial ballast has something to do with the capacitor. It is the same reason
you should not run an F34T12 on a commercial ballast. It will make it overheat (or at least run hotter). If you run
an F25T12 on a commercial ballast it will also run at full 40W power.

Here's a discussion about the F25T12 lamps: https://www.lighting-gallery.net/gallery/displayimage.php?album=7829&pos=14&pid=253074

Posted this on Reddit, thought I would put it here too cause why not.
https://www.reddit.com/r/AskElectronics/comments/1pzrlwb/vacuum_tube_oscilloscope_project_what_does_this/
Anyone have good knowledge on vacuum tube scopes? There is a 2 meg pot on the plate of the 884 that I don't really understand.

We have a real tree, but only decided to go this route after 15+ years of service from the Sears Trim-A-Tree had nearly all of its light garlands die, and unfortunately the light sets used unique/hard-to-find bases, so replacements were costly.

This is the third year of cutting an ~8ft live tree, and the whole family loves it, including the furry feline. (Him especially  )