开云体育

-----Original Message-----
From: Hans Summers
Sent: 22 July 2005 15:40
To: 'Allison'
Subject: RE: [BITX20] Re: IRF510 on higher voltage?



Hello Allison

I love Huff & Puff. These days DDS is so easy, but H&P is still simpler and
suitable for many people with no microcontroller etc skills. Though I built
a microcontroller-less DDS:

Thanks for the advice. I actually have been planning to build a valve linear
for quite a while. I have accumulated more than sufficient number of PL519 /
PL509 TV sweep valves - there are several designs on the internet providing
for 400+ W using four PL509/519's paralleled. Very cheap and reliable. What
I am lacking is the HT transformer. I quite liked Arv's idea of a solid
state linear because the PSU is there, all ready and waiting and cabinetted!

Thanks for the links, I had a good look around. You're right, it doesn't
look very cheap. The very real possibility of frying the transistors makes
the idea even more expensive. From what I hear, those PL5xx sweep valves are
virtually indestructible, even if you make them practically red hot.

All of my ham stuff to date has been at 10W or below. See my TX and RX at:




I had my first ever QSO in March 2002 (a mere 8 years after getting my
license!). I've had over 500 QSO's now, all using that 1-valve crystal
controlled CW transmitter. Mostly on 80m, but also on 40 and 30 once I'd
converted the transmitter for those bands.

I'm not a committed QRP'er, I like to operate QRP but I'm not averse to QRO
either, just so long as it's homebrew. I've never touched a commercial rig
and gonna keep it that way!

I did make a 2m AM handheld see
- it's not finished yet,
some trouble with the crystal oscillator multipliers. Needs a bit more work.

Tnx again for the advice and links.

73 de QTH nr London, UK
Hans G0UPL

-----Original Message-----
From: Allison [mailto:ajp166@...]
Sent: 22 July 2005 13:30
To: Hans Summers
Subject: Re: [BITX20] Re: IRF510 on higher voltage?


Hans Summers wrote:

Hi Allison

I am moderator of the BITX20 and have been following your postings
with interest. I have a partially built BITX20 awaiting a little
debugging and perhaps some of the mods. Unfortunately very little time
to work on it at the moment. The pictures are in the G0UPL folder on
the group page.

Hello Hans,

I have built and experimented with you milimal H&P circuits. Good ideas!

I was thinking of using this PSU with switching, as a variable
medium-high voltage supply for whatever I needed it for - but Arv
K7HKL (who you'll know from the BITX group) suggested, why not turn it
into a solid state linear? 50V should suit MOSFET's quite well for
400W and the current capability more than adequate. I have attached a
couple of (poor quality) photos of this monster PSU. You'll see that
the it's divided into two compartments, the top contains just the
choke and three capacitors. It should be possible for me to move those
caps elsewhere which would leave quite a lot of space, I think enough
for the linear. I could bolt on four aluminium panels, one of which to
function as front panel for metering, low pass filter switching, etc.

Opinion, doing that would represent the hard way with limited success.

To get to 400W is a significant endevor and trying with unknown RF
performance power fets is likely failure prone. All of the 100W or more
designs for RF power use known RF transistors (bipolar or fet) that are
expensive (.5-2$/watt US). Most of the high voltage switching FETs are fast
enough to operate at RF but Miller capcitance and input capacitance
will make themm very hard to drive at HF especially 10M.

I may add that the IMD performace for IRF510 and friends is terrible and
that translates to RF polution. They were not developed for linear
service, they were intended as power switches. The other thing is the
TO220 tab cases have poor thermal dissapation. That makes getting
thermal stability hard.

IF you were into AM phone, it's possible to use switching fets in a
class E design (Not FOR SSB). There are several here doing that with
power to 1.5KW. There are websites that discuss this.

Now, if you really still up for the idea.. Heres a possible path to follow.

Expect to use parallel devices (seperate bias for each device) in push
pull configuration. You will need to accomodate the very high
(800-2000pF) input capacitance and low impedences involved. Expect
elaborate groundplanes and input/output transformers with quality
ferrite. A design of four IRF510s in pushpull (pairs of parallel
devices in pushpull) might get you to 80-100W at 24V or about 1Sunit down
from 400W. Generally parallel devices are touchy and each has to
be biased independent of others.

So, I wondered what you thought of this crazy idea, and can you
suggest any web references on solid state linears at this power? I did
some googling but didn't manage to find anything particularly
applicable. I have heard that Motorola might have some good App notes
about this.

Yes they do but, not for common HEXfets or other low cost power fets.
DEVICES like MRF140 and the like are used and they are not inexpensive.
Typical Vdd for these parts is in the 50V region (at up to 20A!).

Start with this site:

Comm Concepts sells kits for power amps based on Mototrola app notes.
Browse the site for examples, schematics and some of the apnotes. Also you
can find some of the appnotes on the net. However, none
specify low cost fets and likely would not support low cost power fets.

Also, what is your opinion about valve vs solid state? It seems
strange that such a large number of linears are still valve based?
Even the newest models.

Simple, valve designs using obtainable valves are fairly straight foward and
well documented. A 3-500 or 4-400 in grounded grid with 2000V on
the plate will easily hit 400W for 80-10m. A pair of 811/572B are in
the same league as well as a long list of other Valves that maybe had
used or surplus.

Why Valves(tubes)? They are known, durable within experience of most
hams and often can be had inexpensively. When you consider a 3-500Z can
often be had new for around 150$us and MRF150(x4) will cost 200$us and the
FETs may destruct on power up it's not too hard to visualize.

This outfit sells RF power transistors and other items like tubes(valves).



While I've done commercial work at high power I choose to do most of
the ham activity at under 20W. My unusual twist on that is my station
while low powered is supported off of solar pannels and batteries. I may add
that I prefer VHF and up for my experiements.


Regards,
Allison
KB1GMX