What's the view on the single phase plate transformer choice for this power level? How much is the 240 going to dip at key down? (did he connect to the 250?)
The new plate xfmr is fine. The main issue with this power level is the V drop, from the pole pig in the street...to the primary of the plate xfmr. It's one big loop. It's typ 3-000 CU wire from main 200 amp panel to the meter base, then more 3-000
CU up to the pothead, where it's crimp spliced into just 2-00 aluminum drop line. The other end of the aluminum drop line connects to more aluminum..that runs between utility poles. In my case, my drop line terminates across the street, dead center between 2 x poles.....(mid span drop) then down 1.5 poles to the rinky dink 50 kva xfmr..... that feeds 9 x homes...and we all have...'200 amp' service.
On buddy's HF 3x6 amp, his dropped 8 vac under a 90 amp load. 6 vac from street xfmr to his main 200 amp panel, then another 2 vac, from main 200 amp panel...over 20' to the HV supply. He used 2 ga from main panel to HV supply. In his case, he had 240 vac coming in, unloaded..... so we used the 230 vac tap on the xfmr. That's normal to use a tap that's a full 10 vac lower than the no load incoming vac.
Mine used to be 247.2 vac in spring + summer.... and 239.9 vac in fall / winter. I could set a clock by it. These days, it's all over the map, varying by the time of day, hr, week, never know what's it's gonna be from one day to the next. I check once or twice a week, and lately it's sitting around 234 vac... and once in a blue moon, as high as 236 vac.
My plate xfmr primary has taps for 198-208-218-230-240-250. Only options are the 230 tap..or the 218 tap. I don't like the idea of using the 218 tap...in case the incoming vac shoots back up to 240 +.
On SSB, it's really a non issue. When Rich Measures pulse tuned his 4x10, with 30 wpm dots... the old bitty next door left
her porch light on all night long. Her porch light would pulsate like a machine gun. You can see lights blink on CW with just a 3 vac drop. On SSB, you never get that effect. Rich used 150' of just 4 ga wire from his HV supply, to his main panel. His plate xfmr was puny at just 68 lbs (later switched to a slightly bigger 86 lb xfmr, still way too small). His no load B+ was 9900..which dropped down to 9 kv...sucking 500 ma of idle current ( which then maxed out the CCS rating of his 4.5 kv xfmr).... then while talking on SSB, it dropped a bunch more. With a cxr it dropped even more. This is what happens when stuff is not sized correctly.
On high C filters, you get this massive current spike every 8.3 msecs, when on 60 hz single phase power. W8JI measured 60 amps on peaks, every 8.3 msecs, just on an AL-1200, running 1.5 kw out. Scale that up to 10-15 kw out, and you can see that the peak currents on high C filters is just wicked. Issue then is, you also get wicked V drops everywhere, from Cu wire, AL wire, all connections and splices, breakers / fuses / cross connects, relay contacts, contactor contacts, etc, etc.
Then add in the 'normal drop' from the plate xfmr, + glitch resistor...and you further compound the issue. This is why you don't want a plate xfmr with a high secondary DC resistance, cuz it will kill V regulation every time.
In my case, I opted at the last minute, to locate the HV supply (HV supply is just the xfmr + FWB / contactors etc, no filter caps) very close to the main panel.... then ran Teflon 393 coax, to carry the B+ over 20', then down 5', to the HV filter box, then to both RF decks. Then used 1-0 cu wire from main panel to plate xfmr box. I can't do anything abt the drop from the street, except use more C filtering, which works good.... esp on dynamic loads, like SSB. On data modes, you get max static load drop, but nothing u can do about it...and for static loads, everything is in steady state anyway.