On 6/25/24 19:53, Chris Albertson wrote:
Gene,
I’m one of those nuts who still builds electronics that use vacuum tubes.
While incandescent light bulbs are nearly gone, tubes are not. The heaters in
the tubes are just like those old lights and don’t like the higher AC voltage.
The solution is to place an (about) 1 Ohm power resistor in series with the
heater circuit. The exact resistance depends on details.
I didn’t know about the 13th power but it seems about right after some
calculator button punching. Tubes can cost from $10 to $30 each so liftime
kind of matters. I usally set the heaters to run 5% under nominal voltage
and they will last basically “forever”, or “thousands of hours”. I’ve been
applying you observation in the other direction.
A standard light dimmer would make an old hot-filament light bulb run forever,
It makes an ugly waveform but the bulbs don’t care, they just see the lower
peak voltage.
Actually they see the rms voltage. Some of the "tubes" I am fam with as
a bc engineer weigh hundreds of lbs, and if made today would cost
$250,000 or more. But haven't been made or used in 30 some years as
better, more power efficient designs have been invented and the sheer
force of power bill expense has forced transmitter replacements to use
the newer tech as it runs on less power input for the same output. The
station I finished out my working time had an older GE transmitter, that
used a big water cooled triod as its finals, Econco could put a new
heater in it for $7000 1985 dollars. That transmitter was built in the
mid 1950's and all its filament power trannys were wound for the 110
volt line voltage when it was new but on the 127 volts of 1985, tube
life was understandably short. Most obvious was the pair or 4cx250B's
used for visual modulation, had a life of about a month. Its not a 6.3
volt tube, but a 6 volt tube, by 1985 those tubes were being lit with
around 7.15 volts. I installed a powerstat to turn the transformers
supply down to run those 2 tubes at 5.5 volts. 3 pair of those kept it
on the air till it was turned off & replaced with a digital tx at
midnight June 30, 2008. That was so successful I did all the tubes in it
so I could turn them down to 80% if they were good. Cut my tube budget
to about 5% of the previous amount.
It was originally equipt with 2 of the big water cooled triods because
the FCC thought the sound needed as much power as the pix did in the
50's. But by 1952, all the receivers got their sound from the
inter-carrier difference of 4.5 mhz. One characteristic of that is that
in the mix, the strength of the resultant 4.5 mhz signal is determined
by the weakest of the two signals fed to the mixer. By that reasoning, a
much weaker sound signal was recovered without any AM in the result if
the sound was a lot weaker, so I took the big tube down and fed the
diplexer straight from an ailing pair of 4-1000's making about 1500
watts, thereby reducing the chroma buzz to the vanishing point by
improving the aural signal to noise from 50 db to over 75 db. They were
not, even running their heaters at 85 to 90%, a particulay long life
tube. When they (the 4-1000's) were discontinued by the makers in the
mid 90's I bought up everything I could find including some Russian
versions. And used them up. Then reused the best ones again.
I figured when I bypassed the big tube, I could put it back in service
if the phone ever rang from a viewer complaining about the sound. It was
down to less than 200 watts of output by the time it was turned off
forever, but the phone never rang. Still making 26.7kw synch tip peak
from the visual.
Another bit of interesting history, since I am familiar with the
4KM100LA klystron, was that the FCC had no clue why, but when doing an
audio proof of performance, we were allowed to remove the video input
while measuring the aural as there was about a -55 db buzz in the audio.
Caused by the video creating a bad case of ICPM, (Inter Carrier Phase
Modulation), created by relativity's effects on the electron beams
speed. So the 4KM100LA effectively got longer at synch tip power levels.
About 120 nanoseconds longer. The electrons gain mass as they speed
up, and lose mass when slowed dow. Said another way, it takes more
energy to speed them up 1% than it takes to slow them down 1%. Throw
that into the mixer, and the video puts a buzz in the sound.
I had a breaker feeding a 20 horse water pump, single phase the pump,
stalling it in about a second, my hand was moving the 3 feet to the off
button as I recognized the pumps complaints but my arm was about 3/4
second too late, blew a hole in the visual tubes collector bucket
filling the tube with steam, which crowbared the buildings 1200 amp
entrance breaker. Went down with a 10 gauge bang.
While thinking about ways to get back on the air, I rolled the tired
aural klystron into the visual cubical and retuned it a little wider so
I could feed a small amount of the aural carrier into the visual with an
N tee. Bypassed the diplexer, and got back on the air at about 1/2
power. The first thing I noticed was the missing buzz, both carriers
were now being subjected to this same ICPM time distortion by being fed
to the same tube, so it canceled in the mixer. Ran that way for about 3
months while Varian was making us a new tube. For that sort of money the
gov had to call in a quorum of the legislature twice, once to let the
contract, and once to write the check.
Yeah, just one of my BTDT war stories as a broadcast engineer.
Modern lights all run off voltage-regulated power supplies. Any time you can
use a regulator, the higher AC volts only improve power distribution
efficiency. Less i-squared-r loss.
On Jun 25, 2024, at 4:22 PM, gene heskett <ghesk...@shentel.net> wrote:
On 6/25/24 16:24, Viesturs Lācis wrote:
otrd., 2024. g. 25. jūn., plkst. 17:17 — lietotājs Todd Zuercher via
Emc-users (<emc-users@lists.sourceforge.net>) rakstīja:
Hang on here, back up a bit. Start at the beginning. What is the input
voltage on the name plate for the entire machine? What voltage are you
feeding to the machine? For example if the machine's name plate calls for
400v, but you are feeding it 460v, that would explain having about 30v too high
voltage at the servo's transformer. Does the machine have a large transformer
in it similar to this that the main power goes through?
This got me curious and I checked what my multimeter thinks of mains
voltage. It was showing 405VAC where it is supposed to be 380VAC. And
I was thinking that this is multimeter error rather than entire
powergrid being 25V higher, but little search on web revealed that our
national standards require voltage tolerance of +/- 10%.
The latter, over time from 1946 till now in my case.
Within my lifetime since I wired the house my stepfather was building for us at the tender age of
12, when the REA wired the country after WW-II, the "std wall socket" voltage delivered
from the pole pigs that feed our houses, has risen from 110 volts, to 127 volts, each time paid for
by the shortening of the life of our incandescent light bulbs. So the "220" grounded
center-tap drop feeding our services now reads 254 if your meter is accurate, not the 220 in the
vernacular. In case anyone wonders, the life of an incandescent light bulb is determined by the
13th power of the applied voltage. Your trivia factoid for the day.
Hope for the best, plan for the worst - Peter, what do you think about
rectifying 232VAC 3-phase for 8i20?
That, not allowing for rectifier losses, would be 327 volts. Peter possibly
would have some other considerations.
Viesturs
Cheers, Gene Heskett, CET.
--
"There are four boxes to be used in defense of liberty:
soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author, 1940)
If we desire respect for the law, we must first make the law respectable.
- Louis D. Brandeis
_______________________________________________
Emc-users mailing list
Emc-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/emc-users
