To build or not to build
 

If everything works out this weekend I should have about 10w or so to
drive into a linear. My friend here in town thinks I should buy used
Dentron Clipperton or Heathkit. I don't think 10watts would do either
justice. If 10w would give me a KW out with a commercial unit I would
consider buying a good used unit.

Plan bravo is to build. I would have to do tubes in keeping with the
theme of that side of the shack. CW, AM, & SSB are all modes that I
would operate. I only need to operate 160m - 40m. I am experienced
transistor engineer and my early schooling was in tubes , so not beyond
my reach for this type of a project. I have plenty of other homebrew
projects to keep me busy if this one gets scratched.

I need help picking a topology and a tube set. Price and likelyhood of
getting the thing working are two top runners. From the 1965 ARRL
Handbook I see a few different tubes that might qualify. A Pair of 813
in parallel, 3-400Z, 3-1000Z. I am soliciting some opinions on the
topology. Again, all I got is about 10w am going into. I would like
to do a kw or so pep. I am not looking to buy any parts at this time,
so not looking for any 'sales calls'. I need comments like - look at
this tube with this plate voltage kinda thing. Or try driving this toob
with a such and such to get a kw out. I would like to keep the design
required to a minimum and would consider some 'reverse engineering' a
plus at this point. If I can get most of the design out of an old ARRL
handbook then that would please me.

73
Bob
N9NEO

To build or not to build
 

If everything works out this weekend I should have about 10w or so to
drive into a linear. My friend here in town thinks I should buy used
Dentron Clipperton or Heathkit. I don't think 10watts would do either
justice. If 10w would give me a KW out with a commercial unit I would
consider buying a good used unit.
=================
That's a 20dB amplification ...a bit much for a single stage amp.
You need a 10 dB intermediate stage ,resulting in 100 W output.
The final 10 dB to 1 kW can then be achieved with a (standard) Poweramp.

Frank KN6WH / GM0CSZ

To build or not to build
 

Highland Ham wrote:

If everything works out this weekend I should have about 10w or so to
drive into a linear. My friend here in town thinks I should buy used
Dentron Clipperton or Heathkit. I don't think 10watts would do either
justice. If 10w would give me a KW out with a commercial unit I would
consider buying a good used unit.

=================
That's a 20dB amplification ...a bit much for a single stage amp.
You need a 10 dB intermediate stage ,resulting in 100 W output.
The final 10 dB to 1 kW can then be achieved with a (standard) Poweramp.

Frank KN6WH / GM0CSZ

The usual 6146 final stage doesn't take in more than a couple of watts
for 100 watts out.

I don't play with heavy iron, but AFAIK most linear amps are limited to
10dB of amplification because most hams have 100W output rigs and the
FCC kinda frowns on non-home-built amps that can do more than that
(something about CB'ers).

I've seen tetrode amplifier designs that look like they could easily do
this. You'd have to actually _pay attention_ to isolating the input
from the output, and you'd have to do all the care-and-feeding stuff
that screen grids require, but it should be possible to do it in one stage.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

To build or not to build
 

To build or not to build
 

Frankly I don't think 10W will drive any linear to 1kW.
You will get the most gain from a Tetrode but I still
don't think it will go 20db. I suggest you search for
richard measures web page and read some of his articles
about the merits of GG vs grid driven amplifiers.

I think you will need an intermediate stage to get
enough drive. Then you might as well pick up an old
Heath SB200 or SB220 for the final stage.
The SB200 will give you 500-600 watts out with a
pair of 572s and you can buy them for 300 bucks.
Plus the tubes are not too expensive if you need new ones.

chuck

To build or not to build
 

On Thu, 15 Dec 2005 04:42:33 +0000, Highland Ham
wrote:

That's a 20dB amplification ...a bit much for a single stage amp.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

That's true in grounded grid mode, but quite doable for grid driven.
According to the Amperex data sheet, a pair of 3-1000Zs in class B
grid driven mode will give 3560 watts out for 28 watts of drive. For a
single tube, that scales down to 127 watts out per watt in. 10 watts
in will give 1270 watts out, a power gain of 21 dB.

That's the only tube I've looked at so there may be others that will
give a full 1500 out for 10 in. On the other hand, 1270 is close
enough to 1500 that you'll probably never notice the difference.

You will have to neutralize it of course, lest you end up with a high
power VFO. :-)

73, Bill W6WRT

To build or not to build
 

Chuck Sherwood wrote:
Frankly I don't think 10W will drive any linear to 1kW.
You will get the most gain from a Tetrode but I still
don't think it will go 20db.

Most modern tetrodes will easily provide 20dB of gain, and with good
stability.

The favored technique is to use an untuned grid circuit which is
resistively loaded (often called 'passive grid').

In class AB1, grid-driven, a tetrode requires very little RF input power
(in simple theory, none at all). All the grid requires is a voltage
swing. For good linearity, the tube should never be driven into grid
current - in other words, even at peak RF input voltage the grid should
never become positive. That means the peak RF voltage should only be
equal to the grid's negative DC bias. These basic requirements are
totally independent of the output power capability of the tube.

For tube data, see:
http://www.g8wrb.org/tetrodes.shtml

The driver stage will need to see a 50 ohm load impedance, so the most
common input circuit in commercial tetrode amplifiers is simply a 50 ohm
resistor. However, 10W into 50 ohms is only 31V peak, and a typical
value of grid bias is probably about -50V (depending on the specific
type of tube and its recommended DC bias conditions). It is quite true
that 10W could not drive such an amp to full output.

But you're not forced to use a 50 ohm input resistor. Think about using
a broadband step-up transformer to get a larger voltage swing.

For example, a unun transformer with a 4:1 impedance step-up, terminated
in a 200 ohm resistor. This will provide a 50 ohm load to the driver
stage, and 10W will deliver about 62V peak to the grid of the tube. So
any tube that will operate in class AB1 with a grid bias of 62V or less
can be driven to full output by 10W.

10W into an input circuit consisting of a 9:1 unun terminated in 450
ohms will deliver about 93V peak. This is more than enough to drive any
modern ceramic tetrode to full output, so the driver stage can actually
be operated at significantly less than its 10W maximum, for improved
intermodulation distortion.

It should be quite easy to find or develop a suitable broadband
transformer design, especially since the original poster only wanted to
operate 160-40m.



--
73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

To build or not to build
 

In article ,
Ian White GM3SEK wrote:
73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

Damn you guys are good. I need to hang out here more often!
chuck

To build or not to build
 

Ok, lots to think about here. I figured 20db would be a lot to ask
out of production unit. Tim's made a couple good points. A Tetrode
may be a good way to go, but I wonder if a 4146 driving into something
else might be a bit cheeper. Maybe a pair of 813s? As I said the 1965
ARRL handbook claims a kw out with a pair of 813s,but I have my doubts.
They are cheep enough - $55 ea at tubesandmore.com The Taylor 572B
can be had for about 200 bucks for 4 of em. Would this be a viable
option?

I wonder if the 4CX1000 or 3-1000Z solutions can be done for a good
cost relative to other implementations. Also, are they easy to get
working? I'm afraid I might end up chasing parasitic ghosts. I am
aware that layout can be critical and would look to clone somebody
elses design that's up and running if that's the way I go.

Thanks for the comments. I thought it was narrowed down to a large
tetrode with a lot of gain or two stages with 4146 maybe driving 813s
or the 572Bs that get run in a lot of commercial rigs. Then I realized
I would have to buy a plate transformer and a chassis and a box and a
whole bunch of other stuff. Right about now some kinda intermediate
predriver to get between the 10w out and the 100w in that I need might
not be a bad deal. Then again I really need to buy from somebody I
trust cause I hear a lot of people whining about used linears they
bought. So nothing is decided.

To build or not to build
 

On 15 Dec 2005 20:20:15 -0800, wrote:


A Tetrode
may be a good way to go, but I wonder if a 4146 driving into something
else might be a bit cheeper. Maybe a pair of 813s? As I said the 1965
ARRL handbook claims a kw out with a pair of 813s,but I have my doubts.
They are cheep enough - $55 ea at tubesandmore.com The Taylor 572B
can be had for about 200 bucks for 4 of em. Would this be a viable
option?

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

(I assume you meant 6146 in the above)

Remember, adding a driver stage complicates the design because you
will have to add bandswitching to the driver too, unless you can
figure out a way to broadband a tube amp, something not easily done.
If you grid-drive the 6146 you will have to add neutralizing as well.

A pair of 813s have only 250 watts plate dissipation. Getting 1kw out
of them is pushing it and their life will be short. For just a bit
more you can get a single 3-500Z and have twice the plate dissipation.
This is still not the best way to go, IMO. If you are really stuck at
the ten watt drive level, the class AB1 tetrode passive grid solution
would be the best approach.

Big, strong amplifiers are not cheap. Don't scrimp or you'll be sorry.
Are you aware a 1500 watt rated bandswitch will cost about as much as
a PAIR of 3-500Zs? And that's only the beginning. Plate transformer,
power diodes, filter capacitors, tune and load capacitors, chassis and
cabinet, blower, on and on.

Homebrewing a high power amp is a prestigious accomplishment, but in
terms of overall cost, buying commercial is cheaper in the long run.
The money you spend on a homebrew amp will likely never be recovered.
A commercial amp can retain a large percentage of it's price when the
time comes to sell. In fact, if you buy a quality used one you might
well be able to use it for years and sell it for more than you paid.
Something to keep in mind.

My wallet hurts just thinking about it. :-)

73, Bill W6WRT

To build or not to build
 

Hey Bill you got some really good points there. (Yes, 6146)

I figured a homebrew Amplifier project could be a costly undertaking.
I got over $500 into the HX-50 with the cost of the unit and parts to
get it running. I had about 80 hours into a Knight T-150 to get it
working correctly. Was my first rig and I called it a learning
experience to ease the pain. The reason we play with these radios is
because we like to rather than to make any money at it.

I've got a few big mosfet projects in front of the amplifier, so not in
any hurry. (I can do a full PEP setup with a half dozen fets for about
200 bucks.) I'll do some more poking around and see what I come up
with. Project may in fact become a reduced output amplifier.

Thanks everybody for comments. I will mull over and figure this all
out as other projects get finished.

73
Bob
N9NEO

To build or not to build
 

To build or not to build
 

Ken Scharf wrote:
A single 4cx800 will get you close to the legal limit.

That's being over-ambitious. The manufacturer's data (Svetlana) says
that a single 4CX800A "will conservatively produce 750 watts PEP SSB,
and 750 watts Key Down CW in any of the three modes: grid-driven,
grid-driven passive input, and cathode-driven."
http://www.g8wrb.org/data/Svetlana/pdf/4CX800A.pdf

You can certainly screw more power from the tube, but intermodulation
levels will rise rapidly. When selecting a tube, a good rule of thumb is
to expect an RF output equal to the tube's anode dissipation.

If the original aim was for something "in the region of" a kilowatt, it
would be well worth considering a single 4CX800 at slightly reduced
output.



--
73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

To build or not to build
 

Ok, thanks Ken & Ian & others. My post has spawned just the type of
comments that I had hoped for.

Ian, I've been over to Tom's tubes and I see you have boards over
there. If I do the tetrode I will probably pick up a kit to help out.
I don't like to re-invent unless there is a clear-cut advantage to
doing another way. I'm up in the air over 2 stages or the Tetrode. I
think the tetrode is probably going to win. I've posed the question to
some AM types on AM Fone to see what they say.

To build or not to build
 

Ok, thanks Ken & Ian & others. My post has spawned just the type of
comments that I had hoped for.

Ian, I've been over to Tom's tubes and I see you have boards over
there. If I do the tetrode I will probably pick up a kit to help out.
I don't like to re-invent unless there is a clear-cut advantage to
doing another way.

Ah, I didn't realise you were a neighbor of Tom's. (As I hope you
understand, I was writing with the 'In Practice' hat on, and definitely
wasn't trying to sell you anything.)

I'm up in the air over 2 stages or the Tetrode. I
think the tetrode is probably going to win. I've posed the question to
some AM types on AM Fone to see what they say.


--
73 from Ian GM3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

To build or not to build
 

Ian White GM3SEK wrote:
Ken Scharf wrote:

A single 4cx800 will get you close to the legal limit.


That's being over-ambitious. The manufacturer's data (Svetlana) says
that a single 4CX800A "will conservatively produce 750 watts PEP SSB,
and 750 watts Key Down CW in any of the three modes: grid-driven,
grid-driven passive input, and cathode-driven."
http://www.g8wrb.org/data/Svetlana/pdf/4CX800A.pdf

You can certainly screw more power from the tube, but intermodulation
levels will rise rapidly. When selecting a tube, a good rule of thumb is
to expect an RF output equal to the tube's anode dissipation.

If the original aim was for something "in the region of" a kilowatt, it
would be well worth considering a single 4CX800 at slightly reduced output.



I didn't have the data sheet in front of me, but I just assumed that
with a plate rating of 800w it would max out near 1200w out.
The 3cx800 tube will put out this much and there are single tube
amps using this bottle that will do that. Guess the 4cx800 is a
bit short on the plate voltage or max cathode current rating to
make it to 1200w.

Well you could use a pair of 4cx800's to get to 1500w out if that
is your goal. A pair of these are STILL cheaper than a single
4cx1000a or 8877 (especially with the cost of the sockets
figured in).

To build or not to build
 

Hi Ken,

1500w was the target, but I see it's not so easy a target to hit on a
budget. I just did an inventory of shack and realized that I spent a
whole bunch of cash over three years just getting back into radio. A
couple of boatanchors, high performance receivers and some decent test
gear have kept the accounts payable department busy here in MA. (The
N9 call followed me east after college) Wife and I have other interests
that also run on cash, so I can tone down the amplifier performance a
bit no problem. I can be happy with a smaller amplifier and maybe some
time down the road build a bigger beast. 100w-200w of carrier on
160-40 would suit me fine right now. I can build giant multi-kilowatt
transmitters using mosfets if I want to, but the tube thing has a
certain appeal to me now. I got my first class commercial ticket over
thirty years ago but never did much with the hollow state. I'm having
a blast now with the BAs.

Thanks for the help (to all). I'll be asking lots of questions as the
design goes forward.

73
Bob
N9NEO