sounds like the IBM Token-Ring coax of a couple decades ago.  it was two wires TP inside the coax shield.  don't remember if it was air core or not, like the old ugly Motorola car antenna coax stuff.  iirc from some studies in a college 500 class, that was 125 ohm impedance and I would not want to run more than QRP down that stuff.

so it's been done, kinda looked like a two-pin N connector to join it.  you would be much happier with LMR and a balun.

now, IBM found the limitations early, including serviceability, and they moved to twin RG-58s welded together at the outer insulation, and used BNCs for connectors.  you get the same effect tying all your lines together going up a tower so the same woodpecker gets them all at the same time

I would say that a VSD motor (variable speed drive) controller, or any other electronically switched motor, is going to show a low impedance, as there is no capacitive coupling to the power line from the switching devices.

whatever, some basic techniques are going to work a significant part of the time in EFI control.

(1) shield everything that can be... PC boards, inter-system wiring, etc.  return shields to frame ground, and make sure the bare or green ground wire (brown in British countries, varies in others) is firmly bound to the metal frame from incoming power.  insulate the inside.  if there are heatsinks, use perforated metal for the shield to let some heat out.

(2) the stuff you can't shield, filter.  a low-frequency ferrite toroid core is a good start, generate common-mode noise filtering by wrapping as much of all the wiring off a specific section of the system that you can around the toroid, and lightly twist-tie it in place.  space as evenly around the core as you can.

(3) the stuff you can't filter, bypass.  for 120/240 volt service and below, "Y" series safety capacitors across power lines, and "X" series from most control-type wiring to ground.  it may be necessary to get some contractor splice boxes (common sizes are 4x4x4 or 6x6x6 inches, one or both ends removeable, the rest have knockouts) and mount them outside the system to house a Corcom type filter.  in the case of a furnace, use a close nipple and locknuts to hold the box to the furnace where the power used to come in, run the furnace leads to the LOAD of the corcom, hook the power line to the LINE side.  had ground to the box with locknuts, and wires from there to the furnace frame and the corcom, as well as the power input.  all the connections to the same screw with locknuts for a low-impedance common ground.

(4) install any manufacturer noise kits if availiable, but don't wait for them, try what you can without hacking the system up and voiding the warranty.

new production is nowhere near the tubes you have in those RCAs.  run 'em and be glad.

this has been asked before.  consensus is that with the FCC spectrum purity regs required of manufacturers, nobody would touch this with a ten-meter pole (heh, heh.)

you might build it well.  but Splatterin' Sam Slipshod down the block, well, a propane torch and hammer are not the tools of choice to maintain good RF and ground paths inside.

or wind a coil around the tuning fork, hook it to the freq counter, strike the fork and hold it to a magnet.  if there is enough wire, you can get enough voltage to feed the counter's chips.

the reverse was used in the 30s for master clock timing in radio.  General Radio had an oven-mounted tuning fork driven with big-boned triodes that output 115 volts at exactly 60 cycles to power clocks.  the triodes ran magnets that vibrated the fork.

this was before Hammond (the organ guy) slyly made the electric grid stable by giving away his then-signature product, the syncronous-motor clock, to power company executives.  they didn't keep good time until the plant operators started hanging on their controls and keeping the speed constant, and the power execs made it happen so they could keep the pretty clocks on their polished desks.

807 was the 6L6 with the plate moved out of the stem and up to a cap on top of the bulb.  that got the frequency band of the tube past 20 meters.

james, why spend $15 more when the home inspector can kill your house sale?  that's the way to sell it.

baseline is this... and it even applies here to wormwarmer grounds and low-hanging hatsnatchers... it is plain dumb to hack something three times when doing it right the first time costs half that and doesn't need to be done again.  do some basic research on the wacky wacky webbiepoo and find out what the present standards are.  then if you can't figure out how to do that, learn.

DRN, that's a Cinch or Beau block by design (Beauchane licensed the design as a second source,) not a Euro screw-against-leaf-against-wire block.  strictly LV stuff or under-chassis stuff in this country.

I am constantly amazed at what folks try to get away with, and how little they look at what's around them.  when we gutted the basement to the blocks and rebuilt it, out of need, I found a blind connection in a closet wall.  oh, it had a box.  but no cover.  no mechanical backup to the wire nuts.  all to run another two feet to an outlet.

the curious thing occurred when we tore up that closet (headers weren't even nailed to the floor joists or the concrete, of course.)  there were 12 feet of Romex coiled and jammed and twisted in the ceiling, and THEN it came down to where the blind connection was.

never underestimate the power of "hold my beer, and watch this!"

here's how they do it at the dollar-a-holler commercial AMers

http://www.thebdr.net/support/xmtr/maint-ammod.pdf

and a little video from w2aew which shows attenuations needed, etc

http://www.youtube.com/watch?v=y4Zt_LJX1Tc

if you used to be 57 Kb and are now under 28 Kb after a move, the telco equipment is likely a major part of your issues.  full V.42bis type speeds come with a normally loaded line direct to the voice switch.

half that or less likely means you have a DLC (digital loop carrier) or "pairgain" multiplexed system between you and the voice switch.  DLCs typically cram twice the expected 24 voice-band channels onto the same T1 carrier channels back to the voice switch.  that means half the bandwidth is gone with the engineering.

if you have creepy wire in the ground, you could sink as badly as 5 to 9 Kb.  same if your pairs have been "unloaded" of the compensating 88 mH load coils, which is now done to make more locations suitable for DSL service.  coils are nominally placed at 6000 foot intervals, typically at 3 Kfeet, 9 Kfeet, 15 Kfeet, etc.  unloaded, at 9 Kfeet you have the voice quality and volume noticeably deteriorating already.  the parallel load coils correct for the signal skew of the wire capacitance over that distance and fix response issues as well as signal distortion in the voice band.  coils kill faster services like a nail in coax.

if dialup, slow enough beyond today's net already, goes that subpar, check for DSL service.  in some locations with some carriers, you might be able to keep your ISP.  typically not on the high-speed stuff 5 meg and up because it's usually on fiber backbones, falling under different FCC regulations where a telco can't be forced to share the service with competive resellers.

the chances that you can be cut off pairgain to direct voice pairs is slightly less than "freaking heck no, no way, son."

there are one or two guys who build specialized computerized audiophool tube testers for the common preamp and power tubes in that service.

the guy who decides to use 833As for output is on his own.  there are a few. (sigh)

read the article mentioned above.  if you only care about 10 tubes and want to test a hundred of them to find the two best and move the rest on eBay, then you want the Cardamatic 123A.  throughout my childhood, I had access to one in master control of the TV station.  had access to lots of cool stuff in those days. now I have to build it, or rebuild it

that was a poor radiator, not a dummy load

back in the old days, when tubes had one number and everybody was lucky to have that, I'm told they tuned for deep, sharp dip and not wide, broad, barely visible dip that indicated you had one harmonic in tune, but not much else.  and then you adjusted your tuner until the amps stopped coming back to you.  then you advanced your loading and retuned, and checked your wavemeter or your pencil-arc instrument to be sure there was power headed out.

I think it wasn't until the 40s or 50s when they had the Monimatch circuit to read SWR at affordable cost, at a glance.

tuners don't really add any amps to the antenna... what they do is warm the shack instead of passing reflected power to the transmitter when adjusted so the rig "sees" a 50 ohm 1:1 at the power amp.

you may well get a tubload of QSLs from a random wire and tuner, but that's because things were copacetic for low-power contacts.

which is fine.  solve for X,Y,Z all you want, if you are making contacts, the math is optional.

that's a plain old Telechron clock, right?  a little oil on both motor bearings (Oilite) and on all of the shafts of any gearing  may do you wonders.  since the clocks were apparently custom-made for Hammarlund and have been Unobtainium for 30+ years, worth a shot.

you might find looking in the back of a vintage aka thriftshop clock radio of a US brand what looks like a similar clock.  there were several sizes of Telechron mechanisms, and the dial face was customizeable.

just a little mutter... with the engine running, when this camera would do the most good, you have charging voltage on the power bus, which would be between 13.2 and 13.8 volts.  somebody somewhere, someday, might be trying to turn an Explorer into a Dreamliner and need to know the voltage swings.