Discussion in 'Amateur Radio Amplifiers' started by N8FVJ, May 20, 2019.
As the predriver? Some of our Italian ham friends may know...
I'd rather be down the street from a ham running a 3CX3000 than one running a pair of 3-500Zs trying to make 1500 pep output.
This entire thread is a joke, right ? Puritans on crack ? AES was selling Bird 10 kw slugs like pop corn. Coaxial dynamics still sells 5/10/25 kw slugs that fit both CD + Bird 43 meters. Any monoband ham yagi will easily handle 20+ kw. Done all the time. 10-25 kw CM chokes for the balun is old news. Tetrodes like 4x5's,and 4x10's and 4x15's can be run either grid driven...or as high mu triodes..with screen + grids bonded to chassis.
For triodes, typ 3x3, 3x6, 3x10, 3x15, 3x20 GG triodes are used..and also surplus MRI tubes like the socketless YC-156/172. The 3x3 will do 1.5 kw out with 50 watts of drive...and 5.4 kw out with 200 watts of drive...and almost 8 kw out, if u have a 400 watts yaesu. Back in 1969-1970 era, A SB-220 in the CW position was used to drive a 4x5, with screen + grid strapped to chassis...7.5 kw out. The modern version of a Henry 8K Ultra is being sold in Japan...and are currently being shipped world wide....including the USA. Same deal, 5kw+ with 200 watts of drive..they even have a demo on U tube. Model KL-1 $10K
The 3x /10/15/20 triodes require a small IPA ( intermediate power amp). And yes the entire mess will run QSK on CW. Using paralleled + sped up gigavac G2 ceramic vac relays....(2.5 msec op /rls). 4150/4800/7200 vac pole pigs are used, along with custom built dahl / several other xfmr makers....(5.2 to 5.6 kv). 2-3 x paralleled wafers used for the bandswitch. (model 85/88/90/92). Sometimes ceramic vac relays used to change taps on the tubing coils...( next size up from a G2).
3/8" - 1/2" - 3/4" - 1.125" copper tubing is cheap at any home depot...then silver plated. On 10-11-12m, 1-2" wide silver plated strap, wound flat is used for the tank coil. This is all old news. Been going on since 1957. 5kw into a dipole is a 'no-no'.... but 1.5 kw into stacked yagi's @ 200' is..."ok" ? Gimme a break. Those OM-4000's will do an easy 4-5 kw out..with very very little drive..20-30w. They are sold as manual tune and also full automatic. Both versions are sold all over EU, Asia + NA. The..."FCC approved" attenuator is pulled out, or bypassed, so it can be driven to 4-5 kw out. Outside the USA, the pad is NOT even installed.
Typ NA residential service is 240Vac @ 200A. Typ a 50/100/125A breaker is installed into spare slots on main 200A panel. ( they all eat up 4 x slots). Then anything from 2 ga to 3-000 Cu, is used..and routed to HV power supply. 3 pole contactors used ( 2 poles used). Or all 3 x poles are paralleled to make one big SPST contactor. In that case, 2 such assy's are used, one for each hot leg of the 240 vac. A 3rd contactor used for the step start. Sometimes a small varaic is used for the step start, then switched offline.
You don't even want to know what the 11M /CB ops are up to... beyond crazy. Latest I heard just a week ago is a fellow who's running 3 x 4CX-15,000's in parallel...driven by a single 3x3. In those cases, 240 Vac @ 400A service is installed...with a 200A subfeed to HV supply. 7-16 Din 13-30 Din. and EIA flange coax connectors used. The 7-16 din is popular and plenty big enough for < 20 kw ham use. RG-393, Teflon version of LMR-600, and LMR-1200DB / LMR-1700DB used for coax...along with the 7/8" and bigger heliax. A 2-3 turn, like 2-3' in diameter, around the mast is used as a 'rotor loop'...with yagi way up the mast. 11M ops will typ slide a ton of type 43 torroids over the heliax for a CM choke. Either owa (direct fed DE) used...or a hairpin with large diam tubing. Corona balls installed on tips of DE. Phallic shaped solid AL tips also used for corona. (solid AL rod with one end rounded off).
If you want to see what a ...."short wave broadcast" AM linear looks like, W8JI has it depicted in 2 x places on his home page. Uses a single 3x15. 25 kw pep out.
These days, folks buy the new 3x3 tubes ( greenstone brand) from Henry radio. $739.00 new. Econco can rebuild these chinese tubes..and also Svetlana tubes..and also Eimac. Same socket used for the 3x3 +alos the 3x6..... is available from RF parts. RF parts sells a superb chimney for the 3x3. The 3x6 tubes use straight up/down teflon chimney.... 1/4" thick.
Tuned input typ consists of a pair of broadcast air variables + tapped coil..dead simple. (11m ops will use vac caps for the tuned input). Flat input swr from 160-10M... handles a kw of drive easily. Bifilar fil chokes are easy too. 2 x 8 ga magnet wires on a 1/2" x 8" long type 43 rod. 2 x rods used for 80A (3x6). 2 x 10 ga used for 60A (3x3). 4 x rods used for 160 A assy (3x15, 3x20). For a 2 x 3x15 or 3x20 bifilar... ( 6.3V @ 160A PER tube).... go to plan B. Instead of 6.3Vac @ 320A.... fils are wired in series.....and a 12.6 Vac @ 160A fil xfmr is used..slick. They got that idea from the Kenwood TL-922. Sola constant V xfmr used to feed a variac....which feeds the input of the fil xfmr. Then no V fluctuations on the cathode /fil.
Any other trade secrets u want to know about about ?
WOO HOO !!!!!
Years ago, Mr Eimac (will leave his name out of this) phones me up at work.... all excited about the latest 'big tube'. They took my advice and made a
socketless 3CX-20,000A7. Called a 3CX-20,000C7. 2 huge threaded lugs point straight down. (what I suggested was a socketless 3x3). Huge grid flange with 12 x holes. Just whack a hole in the chassis with the hydraulic punch. ( I have a set of green lee punches plus a 10 ton hydraulic punch from harbor freight) . Simply bolt the grid flange to the chassis...done.
No megabuck expensive socket required. The Eimac YC-243 is another popular tube.....socketless 3CX-6000A7 / YU-148...aka '3x6'. Same deal, just bolt the 4.25" grid flange to chassis with 12 x 1/4-20 bolts...done. The biggest GG, hi-mu triode that Eimac makes is a 3x20. 3x20 and 3x15 are the same tube.... except the 3x20 has a bigger diam anode cooler. Eimac built a one off jumbo GG triode for a 11m op.... 3x35.... only 1-2 built....and was being tested by the end user 2 years ago. Biggest plate xfmr I saw was a 1000 lb unit...aprx 3 years ago...pix sent to me. The greenlee punches work good for stuff like meter holes, blower holes. I use a 7 ton, lever operated bench style roper whitney punch for smaller holes... but It only has 3.5" of throat depth.
We are allowed 2250 w pep out on ssb, here in Canada....and that's measured ......'across an impedance matched load'. The load is the ant feedpoint, so factors in feedline loss. With just 1/2 db loss, (aprx 10%) we now require 2.5 kw pep out of the amp to get 2.25 kw pep to the ant. An 8877 is maxed out at 2.5 kw. 8877 can't be rebuilt either...'throw away tube'..and is stupid expensive ($1000.00 for the chinese version....and a whopping $2800.00 for the Eimac... per last years ameritron parts catalog). New Eimac 3x3 ....last week is $2200.00 Back in the day, the 3CX-3000A7 was only $50.00 more than the 8877..... no brainer. No 3 min warm up either. The YC-156 MRI eimac triodes were plentiful a few yrs back.... but they have dried up....since MRI's have gone SS. The Eimac YC-156 is a 3CX-15,000B7..... but with a smaller (4.94" OD anode cooler. 8 min warm up. Superb tube though..... 200 w drive = 8-10 kw pep out.... no IPA required. YC-156 is another ....'throw away tube'. Econco can rebuild the 3x3 till hell freezes over. The greenstone 3x3 is all the rage, cheap price, works good, very rugged, rebuildable.
The secret to long tube life is to initially operate the directly heated cathode at rated voltage for 1st 200 hrs....(emission increases the 1st 200 hrs). Then the fil V is reduced till the PO...just drops off a few watts..then fil V increased by .1 or .2 Vac. Most will install an hour meter in the fil circuit.... typ on input side of the small variac. Buddy has > 8K hrs on his tube.... amp built back in 2007. Fil V is still cranked way down. Broadcast FM will typ get 4 yrs from a 3x3...35 k hrs before sending tube off for a rebuild. 11m ops do the opposite, they increase the fil V, like by 5-10%.... to increase emission even more... to obtain even more plate current...... tube's don't last long.
For my application... the 3CX-3000A7 is bomb proof, with it's 225 watt CCS grid diss.... which increases a bunch more since grid flange on socket is bolted to the cold chassis.
IMD is typ -59 db pep for IMD3....which is outstanding. Just loafs along at 2kw pep out..which is ample. Mr Eimac did mention the.....'world wide secret underground amp builders group'. I just laughed. Apparently in Japan, a pair of 3x3's is called......'the Tokyo Express'. I learn something new everyday. I own 4 x drake L4B's. 1290 w pep out is more than ample.. quiet, instant on. ... 3 are in storage. The hb 3x3 was a good make work project, and learned a lot. I like to test various concepts. Not wanting the nightmare of 9 x tuned inputs, I opted for the simple pair of broadcast variables and tapped 4 uh coil ( both caps padded on 160M). 6:1 jackson bros 4" diam ball drives for each air variable. Silky smooth, and calibrated 0-100 over the 180 deg arc. A cheat sheet made up, for various bands, or upper / lower portions of lower bands). Vac tune and load caps used..plus a roller coil....+ tubing coil for 20-17-15, 17/15 m tap done with a relay. Just...'dial up the numbers' on the 3 x turns counters + both verniers. Buddy got the idea of the air variables and tapped coil from Orr's very last hand book.... (80M monoband amp) . These days, I'm experimenting with different versions of HV supplies. I'm trying a C-L-C filter setup. Looks good on software (Duncan's free PSUD-2). GM3SEK's free pi/pi-l software is used to design both the tuned inputs and hi-power PI / PI-L kw output. Silver plating is done with the ..'cool-amp' glop, in my kitchen. Cool-amp corp is in Ore. Sri for all the drivel.
preaching to the choir
But oh Man! Those voltages, currents and power scares the heck outta me!
Not sure what to do with the 3CX3000A7 amp that I have now.
Similar build with roller inductor but vacuum vari-caps instead of large air caps.
It is indeed a beast of an amp.
The air caps are only used for the adjustable tuned input. Output side is always vac tune+ vac load.
Arco compression caps + fixed coils, work good on tuned inputs, where a separate tuned input is used on each band.
The biggest Arco is used on the lower bands..and padded on 160 + 80M.
Vac load caps are easy to pad, if required, since they are at a 50 ohm point. If a roller is used (no bandswitch obviously), a simple relay
can be used to pad the vac load cap. I would suggest to hi-pot test any vac cap u intend to use...and ditto with any surplus vac relays.
Ditto with surplus tubes. A new / rebuilt metal tube will last a looooong time. A lot /(most) of these ebay tubes are low emission....and if bought,
need to be rebuilt.
A good 3CX-3000A7 will hi-pot test to 20 kv, between anode to grid (temp strap between grid + cathode). With a temp strap placed between
anode to grid, the grid to cathode will hi pot test to 5 kv. The 3x10/15/20 tubes all hi-pot test to 30 kv ( anode to grid). Dunno what a 3x6
hi-pot tests to. A good 3-500Z will hi pot test to 12 kv.
When the tube is driven, the RF voltage formed in the PI tank output.... will backfeed through the plate block caps...and superimpose itself
on the anode. Total V on the anode is typ slightly less than double the loaded B+.
Take the usual precautions and u won't have safety issues. I install HV meters in any box that has B+. One in HV supply..and a 2nd one in RF deck...
(even if RF deck and B+ supply are in the same rack), wired between B+ and B- of filter caps in HV supply. Don't wire between B+ and chassis..that's old arrl fubar design..dangerous. A Buss HVU series HV fuse on output of B+ supply..which feeds the 50 ohm 'glitch resistor'..then off to RF deck. Any B+ to grnd arc on outboard side of glitch resistor results in a blown HV fuse... event over. The 50 ohm glitch resistor LIMITS the fault current to a safe
value....while the HV fuse INTERRUPTS the fault current.
HV fuse will open off in < 2 msecs every time. With the HV fuse open..... the HV meter in RF deck will read zero. Meanwhile the HV meter in the
HV supply ( wired between B+ and B- of filter cap) will still read normal B+. In effect, I have HV meters on both sides of the HV fuse.
We also install a 2nd HV fuse, this time between plate xfmr sec (either leg, but just one leg)..and input to full wave bridge assy (or FWCT, FWD etc).
If any B+ to chassis arc inboard of 1st HV fuse, then the 2nd HV fuse will open off.... protecting the diode board assy/xfmr.filter cap. If this 2nd HV
fuse opens off, the BOTH HV meters (RF deck + HV box) will read zero. A magnetic-hydraulic type of breaker is used on the HV supply's 240 vac input.
The 'fast trip' / 'instant trip' type is used. The instant trip versions have the hydraulic fluid removed. They are much faster than the typ thermal type breaker.
IF the 1st HV fuse (B+ fuse) is omitted, the 50 ohm glitch resistor will be fried..... by the time the 240 vac breaker opens off. Paralleled 6A10/10A10 /10A07 diodes are wired between B-....and chassis ( rvs connected config). This ensures the B- floats above chassis potential, so the normal plate and grid current meters can be used. No multi meters used anywhere. Fast rear panel 3agc type fuse, wired between chassis and negative terminal of grid meter. IF dc grid current exceeds fuse value, grid fuse
opens off.....and no path for DC grid current. Power output of amp drops to zero. Input swr, between xcvr (or IPA) and input to amp rises to infinity, xcvr shuts down, event over...(input high swr protection is also used). The pair of HV fuses + fast grid fuse are for what's called....'hard faults'. Tubes like 8877, 3CX-800A7 / YC-156 also require fast electronic shut down in the event of loss of B+ (grid current will peg, if amp driven), over drive (excess grid and / or plate current)....due to the delicate grid
structure. These are called...'soft faults'. Monoband amps are of course considerably simpler, no band switches, no padding relays etc, etc. Everything can be optimized for that one band. Sometimes 2 x rf decks used, one optimized for lower bands.... and the 2nd optimized for upper bands. Like 160-30m...and 20-10m. Both rf decks can share the same B+ supply / fil supply etc.
Hot air can be ducted to outside in summer...and inside in winter. Blower can be remoted from RF deck. Separate ducts used to sealed RF deck. In buddy's case, the hot air output of the YC-156 80-10m amp is routed to input of the YC-156 160m monoband deck.... then output of 160m amp, routed back to basement.... then either stays in basement or routed to outdoors. Shack is on main floor, above the basement located blower assy (and also HV supply). The 11m ops will duct the cold air output of a HVAC unit into the input of the blower assy. 3M filters used on input side of blower assy, then zero dust build up below chassis, tube fins etc. Ok, more 'trade secrets' revealed.
Like I had said, all that high powered stuff is scary!!! hi!