Illegal high power HF amplifiers

Discussion in 'Amateur Radio Amplifiers' started by N8FVJ, May 20, 2019.

ad: L-HROutlet
ad: l-rl
ad: L-MFJ
ad: abrind-2
ad: Left-3
ad: Left-2
ad: Subscribe
  1. WA7PRC

    WA7PRC Ham Member QRZ Page

    The 3CX3000A7 datasheet shows it can produce up to about 7.2 kW. My SWAG is, it might be able to supply enough drive to six of the twelve 4CW250000B (datasheet) tubes used in the PA at US Naval Radio Station, Oso, WA (aka "Jim Creek"). :p

    It would loaf along at the ham power limit but, you'll need 7.5V @ 51.5A (386.25W) just to power the filament. Oy vey!
    AF7XT likes this.
  2. AF7XT

    AF7XT Ham Member QRZ Page

    it's the 20kv plate voltage of the 4CW250000B that is a bit difficult to wrap my melon around.
    ...and I've seen US Naval Radio Station, Oso, WA "JimCreek"
    WA7PRC likes this.
  3. WA7PRC

    WA7PRC Ham Member QRZ Page

    The filaments run 12V @ 660A (each). Before you hit it w/ HV (from the transformers in the basement), the filaments are each pulling 12V x 660A = 7.92 kW, or 95 kW for twelve. :D

    When I visited Jim Creek, I arrived thinking the 16 kW RFOUT transmitter that I worked on for KEZX-FM was high power. After the tour, not so much. EVERYTHING about Jim Creek is yuuuge! :D
  4. SM0AOM

    SM0AOM Ham Member QRZ Page

    20 kV is about the upper limit for plate voltages used in fielded transmitters.
    Above this, insulation requirements become difficult and the vacuum in the final tubes may also be difficult to maintain.

    Even the previously pictured 500 kW AM Thomson-CSF transmitters and the 100 kW ISB Telefunken used plate voltages in the 14 to 15 kV regions.

  5. VE7RF

    VE7RF Ham Member QRZ Page

    You will electrocute yourself just as fast with 2 kv vs 20 it's a moot point really. You would also have to defeat all the various interlocks. To discharge the filter caps quicker, I 1st shut off the B+...... then hit the ptt/footswitch (no drive applied, vox shut OFF)...then go for idle current. Idle current of the tube will suck down any filter cap real quick almost asap. Idle current on a 3x3 is 150 ma...( 5.7 kv). A 3x6 idles at 200 ma...(7.7 kv). Once down to zero, a temp short is placed between B+ and B- ( alligator clips I make up with 10 kv rated belden test prod wire, and HV rubber boots at each end ). Breaker in both main panel + also breaker in HV supply shut OFF as well. Then it's safe to work on....( RF deck + HV supply) .

    I only do the above procedure if I'm gonna work on either the rf deck or B+ supply. Normal deal when done for the night is to put amp into stby...then shut off the B+ supply...then shut off the fil V....then the blower runs for a few minutes ( blower over run). Then it takes quite a while to drain the B+ down to zero. I only have a 100K @ 3 watt eq resistor across each of the 25 x 10,000 uf (450 vdc) lytics. 7700 vdc / 2.5 megohms = 3 ma...= 23 watts. Takes a loooong time to bleed down to zero. Who cares, I'm then off to bed anyway.

    Since my B+ supply is split into 2 x separate racks....and RF deck is in a 3rd rack..... I'm real careful with the interconnecting B+. I don't use any HV connectors... the 213-U is hardwired inside the RF deck /HV supply with compression lugs + silver soldered.... + HD strain reliefs. I don't trust typ hv wire either, like belden's silicone covered HV wire, way too easy to cut into it, if it got caught on something sharp. I use plane RG-213..... but with the braid + outer jacket stripped back 3". Braid cut off...and 3/8" wide copper strap clamped and soldered 360 deg.. to 3/8" tiny bit of exposed braid. Other end of copper strap is bonded to chassis..with a 10-32 SS machine screw /nut etc. Same deal at both ends of the well marked 213. ( marked with wraps of red tape every few inches). The belden silicone HV wire ( or GTO HV neon sign wire) is only used inside the RF deck..and also throughout the inside of the HV supply. Plane RG-59 will show no leakage current at all..when hi-pot tester cranked up to it's max of 30 kv. 213-u is through the stratosphere. ( provided braid cut back 3+ inches in both cases). I shoulda borrowed the 100 kv hi-pot tester from work... b4 I retired. You can set a user defined threshold.... like 5-10 UA... then it kicks offline asap. Not much in the way of 'trade secrets' today.

    Well I did manage to finally get my 127 lb hypersil Dahl 4H @ 3A choke into the bottom of the rack late yesterday. Had to use a cum-along..... and it barely cleared. It sits on a pair of inverted Al channels...with 4 x glastic insulators holding the channels up off the 1.5" plywood. Choke goes into the B+ line, with filter caps on either side of the choke. (C-L-C filter). A spst-nc vac relay is wired across the choke so choke can be shunted. An adjustable HB spark gap is also wired across the choke. Spark gap is just a pair of silver plated 1/4-20 brass bolts, with lock nuts on either side of the pair of 1 x 1 x .125 angle als. Mating Silver plated brass acorn nuts on inboard ends of brass bolts. Acorn brass nuts are dome shaped..with the domes facing each other. Gap is set for aprx 150-200 vdc. The acorn nuts are easily replaced. And no, you can't put a choke in the B- leg on a GG triode linear amp. The v drop across the 9.5 ohm choke is now yo-yoing bias..... which is then in addition to normal regulated bias. It wouldn't work anyway,since I install RVS connected safety diodes between B-....and chassis..... in both the HV supply...and also the RF deck. The diodes on each side of the choke...would shunt the choke. Bias is adjustable...on the fly. Just a string of 50 x 6A10 diodes in series. 1st 10 diodes are always in the circuit... the remaining 40 diodes are tapped every 2 or 3rd diode...with a 20 position rotary switch. A spst toggle wired across last diode (fine adjust). A 43,000 uf lytic (3"x 5") is wired across the entire string. Bias Vdc does not budge an inch, rock solid. Then the idle current can be set anywhere u want..regardless of B+ value. Reduced idle current for CW/FM/DATA just 20 ma. (u can't reduce idle current to zero on any GG triode.... or u will get key clix on cw). 11 m ops will parallel 10A diodes (10A10 /6A07) for each bias diode...and often 3 or 4. 6A10 diodes run more than warm..with 2A CCS flowing. 1N5408's run more than warm with 1A CCS. Ok, ya'll got ur trade secrets for today.
    KM1H likes this.
  6. WA7PRC

    WA7PRC Ham Member QRZ Page

    The HT value is one of many things VERY large at US Navy Radio Oso. Among them...
    • The wire used in the antenna system is copper-clad 750 MCM wire rope that criss-crosses several times between two small mountains (measured in miles or km). Each end of each span is terminated w/ four 4 foot long porcelain "dog bone" strain insulators. The antenna system Q is so high, they can't run wide (and fast) FSK. 67 wpm is their speed.
    • The antenna tuning room is 80 x 80 x 80 feet, entirely covered w/ copper sheet. The tuner uses two 12' diameter variometers. The tuning loops use geared-down 5 hp 3-phase motors. The "line" between the tuning room and antenna is 12" diameter.
    • The PA tubes each use about 50 gl/min cooling water. They dump the heat into Jim Creek. With the transmitter cranked up to 11 (2 MW RFOUT), Jim Creek is 1° warmer downstream.
    • The antenna presents a Z = 0.456Ω load to the transmitter. When I toured, the antenna current monitor indicated 1296A (only 766 kW).
  7. W2VW

    W2VW XML Subscriber QRZ Page

    Maybe this should be qualified to include tubes without an ion pump.

    Plenty of tubes run fine right up near 40 kv.

  8. VE7RF

    VE7RF Ham Member QRZ Page

    miles away from me !
    Jim Creek is at the top of the list to be 1st hit by a Russian hypersonic nuke (within 5 mins) per Russian state TV in late 2019. Just looked..and the facility is only
    60 miles from me.
    WA7PRC likes this.
  9. KM1H

    KM1H Ham Member QRZ Page

    Im glad to see you back on QRZ Jim and giving a few of our less informed members a dose of reality and others a serious look into what QRO really entails.

    Last edited: Jan 19, 2020
  10. KM1H

    KM1H Ham Member QRZ Page

    Then you can say you went out with a glow:rolleyes:

    OTOH Im only ~ 20 miles from the Seabrook nuke plant which is also a target along with a huge amount of defense related business such as BAE at about 5-6 miles.

Share This Page