Best tube for a new, high power rig??

[steve_qix]

If someone were to build a from-scratch transmitter that would run - say, an "AM Killowatt" or close to it, what would be the best tube based on:

1) Availability - the tube has to be easily available, preferably a new design that is in current production - current production being the key even if an old design
2) Cost - the tube can't be unreasonably expensive
3) Relatively low impedance - the tube has to be able to produce a lot of power at voltages around 2000 volts (or lower would be even better).  Parallel tubes are OK in this design
4) Class C operation (for obvious reasons a low grid dissipation "linear amp" type of tube can't be used)

An obvious choice is the 813, but from what I can gather they are new old stock (meaning there is a limited supply) and not being made anymore.  Something similar to this in a new design would be ideal.

I've looked at tubes like that 6c33c-b  Interesting device!

[KB2WIG]

Shuguang Vacuum Tube FU13 813 PAIR Hi-end Tube Integrated Amplifier Replacement

Price:

US $160.00


http://www.ebay.com/itm/Shuguang-Vacuum-Tube-FU13-813-PAIR-Hi-end-Tube-Integrated-Amplifier-Replacement-/290708859371


Get 'em now, there selling like ricecakes.

Back to reality, RF Parts is taking orders for Chinese 572B.  There are a few 811 LinYar schematics floating around; The 572 would make a nice improvement in the design, allowing for higher output levels. But(?) who knows the quality of the new Chinese stuff. The same thoughts for the Chinese 813.
klc

[K1JJ]

One of my favorite dream projects has always been quad 813s plate modulated by quad push-pull parallel triode-connected 813s.  (813s - 4X4)

All glass, low air requirements and relatively low voltage.  Proven design.

There are many NOS 813s stashed away in junk boxes that will never be used. Axe around and you will find all you want.

Or are you thinking more of a commercial production run, thus needing a large supply?  Tube PDM rig?

T

[KL7OF]

One of my favorite dream projects has always been quad 813s plate modulated by quad push-pull parallel triode-connected 813s.  (813s - 4X4)

All glass, low air requirements and relatively low voltage.  Proven design.

There are many NOS 813s stashed away in junk boxes that will never be used.

T

Yes....4X4 813.....I have the filament transformer...

[WBear2GCR]

Many Chinese 211s being made... probably will work fine on 160 and 75m??

3-500s?
They work fine on 2,000vdc.

The 6C33C are indeed interesting.
The sense I get is that the NOS are almost gone now.
Unsure if any of the ruskie plants are making them any longer or not.
The ones coming in nowadays seem to be a bit prone to being gassy...
...not sure they are really suitable as an RF output tube or not.
They'd work ok as a modulator, and of course are super low Z plates.

                         _-_-

[steve_qix]

If I go forward with this project, it's going to be a pulse width modulated class C RF amplifier.

Good low-z triodes in the modulator are ideal.  If the RF amplifier is operating at 2000V at 400mA, the modulator needs to be able to handle 1000mA of current properly (this will be the peak current at 150% positive modulation).  All tubes need to handle a peak voltage of 5000V which should not be a big problem.

The reason I like triodes in the pulse width modulator is they have a lower internal voltage drop when turned on hard, and usually can handle higher current for a given power level.

If there were a good tube that could run at, say, 1200V at 400mA, THAT would be almost ideal.  A pair of those would work out just fine and the modulator would be easy to build as well.  The power supply for such a rig (with a pair running at 1200V @ 800mA)  would be 3000V @ about 340mA, which is quite practical to construct.

[kb3ouk]

It appears that 833As are still in production, there's also the Russian version, the GU-48. That tube would probably be one of the best suited for the voltage and current handling capabilities you want. I seem to recall that 805s were also still in production, but that might be a little low on the voltage handling capabilities.

[AB2EZ]

Steve

Consider the Russian GS35b triodes. A pair of them worked great for me in my legal limit plate modulated grounded grid transmitter. At carrier, I ran them with 1700V on the plates at 300mA of total plate current. I had them biased at around 50mA of total plate current (essentially class B). This allowed me to use my Ranger to drive them.

They seem to be readily available at around $160.00 new ... But I don't know if they are currently in production.

The downside is that you have to fabricate suitable tube sockets.

I employed some modulation of the Ranger driver in concert with the plate modulation of the GS35b's in order to get good modulation linearity.

Stu

[N2DTS]

Do they still make tubes?
I thought most were no longer made.
I know you can get some from China and maybe Russia, but quality is low.
I have a pair of Russian made 811a's both have loose plate caps and they are not the standard size.

Four 4x150a's will do a KW out easy, input power of 1800 volts and 800 ma, 1440 watts input, likely over a KW out, say 500 watts plate dissipation out of 1000 watts.

You can get them new in the can for $40.00 to $60.00.

[w1vtp]

Regarding the GS35: Mark, KA2QFX is the consummate builder.  He did a fabulous job making the GS35 tube socket.  No.  I don't think he's taking orders.  I just want to give credit to Mark's great talent.  Check this out.  Note the detail for the tube socket:

http://home.comcast.net/~w1ia/new_page_7.htm

Regarding the 813: I like the idea of the quad 813 being a fan of that great tube.  This would probably be my first choice since I have a bunch of them new old stock and tested.

Regarding the 3-500:  I also think the 3-500 is an overlooked tube.  A pair or even three would make a great RF class C phone deck

Regarding the Russian GU-48 - it's like the 833 except it's 1/2 the price - very much overlooked as a transmitting tube.  I've attached some views of the GU-48 FYI

[w1vtp]

GU-48 data sheet

[N2DTS]

Four 813's will do 2000 volts and 800 ma all day long.
That is 1600 watts in.
Only 500 watts of plate dissipation in the tubes though.
My 4x4x150 rf deck used to be four 813's.
I put 3000 volts on them and tried it as an amp on ssb once and I do not have a watt meter that reads high enough to tell what they put out.

If you do not abuse them, I think the 813 will last a lifetime of ham operation.

[kb3ouk]

In addition to the 813, there's the GU-13 which is the Russian version of that tube, and an interesting tube called the GK-71, which is very similar (physically) to the GU-13/813, except it is a true pentode like the 803. The datasheet I seen looks like electrically it is somewhat between an 813 and 803.

[W3GMS]

The Continental 314R-1 uses a pair of 3-500Z in the final PDM modulated with a single 3-500Z as the switch tube.  Bob - 2ZM has one and it sounds great on the air.  3-500's are very plentiful, still made and relatively inexpensive for what you get.  

Sounds like a fun project!

Joe GMS    

[w1vtp]

The Continental 314R-1 uses a pair of 3-500Z in the final PDM modulated with a single 3-500Z as the switch tube.  Bob - 2ZM has one and it sounds great on the air.  3-500's are very plentiful, still made and relatively inexpensive for what you get.  

Sounds like a fun project!

Joe GMS    

What a pretty & powerful looking rig.  Is that Steve QIX salivating?

[w4bfs]

.......If you do not abuse them, I think the 813 will last a lifetime of ham operation.

do any of you remember 'poverty radio' ?    he was a 75 mtr ssb op that drove a pair of 813's with a swan 350 full bore with 4kV on the plates ... he stated 1500W out and that they would last about a year with a spectacular end-of-life light show display .... his signal sounded clean and loud .... said he put -9V on the screens .... "What hath God wrought ?"

[W2PFY]

I see 3-500's as a suggested type but I often see NOS 3-400'S  that are also NOS for a fraction of the cost of 3-500's on ePay. Some of them are sealed from the factory....

Another route would be a pair of 304TH tubes that at 1500 volts will make 1000 watts of audio. I know that your not going to use them in class B but they will be able to meet your current demand. Prices have fallen on this type of tube so a pair might be had  for $120.00 and there is plenty of those out there.

[W2VW]

Same idea as an 833 but lots in use in RF heating.

http://www.rell.com/filebase/en/src/Datasheets/5867a_tb3_750.pdf

http://www.ebay.com/itm/Factory-New-5867A-TB3-750-Industrial-RF-Tube-ON-SALE-/291053678046

[W2PFY]

The 5868 is more like the 833A and its plate dissipation is 450 watts..

[W2VW]

The 5868 is more like the 833A and its plate dissipation is 450 watts..

I think that's might be the next step up. Some spec sheets call 5868 a TB4-1250. That numbering system is sure a departure from Eimac's.

[Opcom]

There have been many good comments so far. I'd like to add that the 1980 Eimac datasheet (revised 8-5-80) for the 3-500Z gives a pulse width modulated application data. The tubes are still being made and are only $150 each.

First thing might be to see if the Chinese ones will take the peak voltage, but why shouldn't they? 9K on the modulator, 3K on the class C amp (and I assume it means 3KV on the modulator's cathode, 6KV across the modulator itself).

It calls for the Tyler circuit, but that looks like an efficiency issue that may not matter at the somewhat lower ham PEP levels. If it ain't enough double up! It's interesting but I have no schematic for this circuit. Does anyone?

edited I guess I missed the 314 R1 PDM post.

[John K5PRO]

I have a 5868 in my LaRose VHF dielectric preheater that I got from the Black Hole years ago in Los Alamos. It kicks out 90 MHz at ~ 500 watts in a self oscillating grounded plate circuit. Very wierd beast. Low mu triodes are preferred in dielectric heating as they change in grid current isn't so drastic when the loading on the output shifts (as the plastic is cooking, foam is foaming, moisture is drying out, etc). Think of it as feedback, in a self-oscillating circuit. I wouldn't recommend using the 5868 unless a surplus of NOS tubes is found, as most from industry would be totally spent and tossed out. No one rebuilds glass tubes this size.

810 was and still is a classic triode for modulator service in this power class, as is the 813 for RF service. See datasheet attached.
I personally like the 3-400Z and 500Z triodes, as they are great tubes if you can find quality anymore. So many have been incorporated in older ham linear amplifiers too. A decade ago the Chinese versions were not fit to test, but I think they have improved. I did a short comparison back then, attached here. Amperex, another source for 3-400Z and 3-500Z, has dropped out of the market last year, being from the now closed Covimag plant in France that was partially owned by Richardson. CPI/Eimac hasn't made one in a decade. This leaves only Chinese tubes available new. As finicky as a Continental 314R-1/Collins 828C-1 is for the switch tube, I am hesitant to try one in mine. I asked the broadcast forum yesterday if anyone has done this, as there are at least 4-5 of the Power Pebble's (314R1) in service by hams now or being converted. Mine is the one discussed on the Rescue of BC Transmitters article on the sidebar here. A friend is about to embark on work with o 314R1 and so I have gotten interested again. I just sold off my BC1H1 so I now have some room in my garage to focus on one big iron. 

[John K5PRO]

Tyler high efficiency operation is described in the attached files. It is also class F, where harmonic modification is done in the plate, drain or collector circuit, to effect a more squared voltage or current waveform and raise efficiency towards the ideal condition of switching.
Sorry for the poor quality of the RCA BTA5T product sheet, I think I found it online years back.
73

[John K5PRO]

These days, this scheme is the rage among wireless designers, who work far above 500 MHz. As is Doherty, Outphasing, other great designs that were developed for AM with tubes many decades ago. Attached are some more papers on class F high efficiency amplification.

[John K5PRO]

Here are my last two papers for now, on class F high efficiency, for those interested. I have been testing a version of this, sort of empirically, in my 2 Megawatt tetrode VHF amplifiers at work, using a quarter wave stub on the output. Problem is, with 9 inch diameter coax, its not easy to place the stub in the correct phasing from the plate of the tube! Cut and try is VERY expensive, so I picked a location, and ran with it. Seems to have moderate effect, but the second harmonic is totally gone, which is nice.