[AMPS] How to build the ultimate HF amplifier ?

[Einar Persson]

 Hi John et all

--- On Fri, 23 Oct 1998 17:05:37 +0200  John Fielding <johnf@futurenet.co.za> wrote:

>Hi all - let's start a new discussion topic.  I am generally considered a
>reliable source of info but I have to admit that my experience on hf is
>limited.  Well to the question and I would like all to consider this one.
>
>I will only have the time and money to build one hf amplifier and would
>like constructive inputs for the initial thoughts I have.
>
>Power supply I can happily tackle but the rf deck is the trickiest.
>
>I have the following parts already. 
>
> Tube - Mullard / Philips QY5-3000A (tetrode - don't get confused by the 5
>in the part number) plate dissipation 3000W.  Fmax = 220 MHz and rated for
>SSB operation.  (Was used for tv ultra-linear duty on band 3   170 - 220
>MHz in the UK).  Vmax = 5kV and g2max = 1kV.  Mullard data gives an output
>of 1,5kW at 30 MHz ssb duty with IMD typically 33 to 42 dB.  Plus, I have
>the base and g2 decoupling gear from an old tv broadcast transmitter. 
>Other tube specs I can post if needed.

That is a very nice tube! Excused for the lack of handles!
Quite commonly used among 2m moonbounce operators in Europe, mainly because they 
have been availible as pulls from BC stations. Usually 3dB down but still alot
of power to give.
If it's new it will last a lifetime at the kW level AND it is almost instant on. 
No 3min warm up!

>
>Tank inductor - big (about 4" dia and 12" long)  probably suitable for the
>low bands.  Others I can wind from 1/4" copper tube which I have on hand.

Is the big one a roller inductor?

>
>Various variable capacitors - none really suitable for the anode tuning but
>probably OK for load capacitor duty.  Best one about 250pF with approx. 6kV
>rating - tested at 8kV dc.

V rating makes it sound like a good C1-tune capacitor to me, *IF* you put the blocking cap 
between the anode and C1. But together with "padder caps" it can also be used as C2.

>
>Several high power vacuum relays - about 10kV is the highest rating.
>
>Lots of high voltage / current rf fixed capacitors.
>
>Suitable blower - blows the QSL cards off the wall at 10 feet and frightens
>the dog!
>
>Now  -  why can't I build the tank circuit using the vacuum relays to
>switch in fixed capacitors for each band.  I have enough relays for this if
>one uses 160m, 80m, 40m, 20m, 15m and 10m.  (My transceiver doesn't have
>the new WARC bands so I don't need to include them initially).  Using a
>relatively small vacuum variable for the anode tuning and eliminating the
>normal tank circuit band-switch seems to me to be the way to go, in view of
>the problems some amplifiers suffer from.

I can't see why not. HENRY 8K (similar size tube = 3CX3000) uses fixed capacitor 
switching for C1 and quite a large vaccum variable as C2. (8K = PI-L circuit)
But with big fixed capacitors and small variable ones for both C1 and C2, 
you'll get limited tune possibilities or a lot of switching at least on the 
lower bands.

Why not go for 2 PI-tanks? One for 6-10-15 and maybe 20m and the other for the
lower bands. Would make suitable PI tank values easier to obtain.
A side benefit would be an amp tuned for two fq's.
See added topic below!

>
>Overall physical size doesn't really matter.  (The power supply will be a
>separate item for versatility).  I can have the required metal-work
>fabricated locally by a craftsman I know.  (Rather have enough room to make
>the thing work efficiently than try and jam it all into a box which happens
>to be available).  I plan to pressurise the anode compartment and use a
>scheme similar to the K2RIW design for 432 MHz with a chimney for the
>exhaust.

A good way to cool the tube. I use the same here, but be sure to also cool the 
glass/sealings and don't forget the heater/cathode pins. I have seen several tubes 
with cracks in the glass around the heater pins. I use quite big connectors for the
heater pins, to transport heat away.

>
>Basic spec.  At least 1kW out on ssb or cw - with no strain.

Piece of cake! My 2m amp will do half a dozen more dB's (CW), into the dummy load 
of course, if asked for it. 
The tube is almost on vacation at the kW level.

>
>Yes - We should shortly get a 1kW licence class here.  (I have just
>finished writing the new regulations for the R.A. and part of the regs
>covers IMD and not running an amplifier at it's limit and removing the
>silly existing regulation which makes it an offence to own an amplifier
>capable of more than 400W! -  so big amplifiers will become the order of
>the day).
>
>Finally - If possible, the amplifier should also operate on 6m with similar
>output.

ADDED TOPIC
Why not use two PI-networks? One for, let's say 6-10-15-20m and the other one
for the 40-80-160m bands? Switched with a vaccum relay.
Of course more parts and more space, but alot easier to handle the few C1-C2 pF's 
needed for 6 and 10m, and the many pF's needed for 80 and 160m.

I think the really ultimate home brew amp would have different PI-tanks for each 
band arranged around the tube. (If size isn't a BIG matter)
One tube with 7 PI-tanks, each one tuned to your favourite QRG on the band, switched
in and out with vaccum relays or solenoids.
And with that blower you can place the PA away from both the dog and the shack! :-)
Silent and calm in the shack and easy QSY without messing around with dials.

>
>Anyway, is there a  flaw I haven't spotted.  Come on guys what have I
>missed.  All ideas considered.

John, take a deep look at SM3BDZ's G2DAF-circuit. He sounds as good as the ALPHA,
HENRY etc owners.
Simplicity with no screen or grid supply. 10w of drive will give you about a kW.

>
>John	ZS5JF
>


Einar SM3MXR,  sm3mxr@swipnet.se


--
FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm