Hi Luis,
Am 25.05.2015 09:33, schrieb VIGILANT Luis Fernández:
Yes. Party tickets sales are rising and all the "noise" in the bands and in the
URE forum is making
hams here to try MF.
That sounds excellent! So don't stop transmitting, even if they still
can't receive you. If there is a signal to detect, then there is a
motivation to try it.
Most of the time only to discover that their pricy Icom is almost deaf on MF
and/or the local noise is over S9++
Local QRM. But often this is caused by a non-optimal RX-antenna
configuration! If you put some wire into the RX input and use the
house-earth as the counterpoise you will likely get that result. But if
they have a small garden and can run a large earth rod into the ground,
resonate and match their dipole for 40m against it, feed the signal
galvanically decoupled to their RX, things may look completely
different! Or if they try an active antenna like the PA0RDT-antenna from
their balcony..
Worth to create a website or sharing some photos to that forum...
But there are some few which takes this as a technical challenge and improve
their Rx to the point
of receiving the first signals. And then they can't stop there ;-)
This is exactly why i say that there must be QRO stations on the
band!!!!!!!!! ...to generate a first detectable signal for a (still)
sub-optimal RX-system!! :-)
The PA is a nice and simple design by GW3UEP
http://www.gw3uep.ukfsn.org/25W_QTX/472_25W_QTX/472_kHz_25W_PA.htm
It is a nice construction but class-E PAs need an exact tuning, i.e. a
very good SWR. 1:1.5 can destroy the PA! A high-Q antenna can quickly
destroy the PA.
I replaces the BS170s on U3 with an ICL7667 as you proposed for the RasPi. This
povides more than
enought signal for the PA Mosfet, actually an IRF510. Powering it at 14V got a
6 times increase in antenna
current, which provided the spots out of EA last saturday night. This links to
some pictures of the PA
Nice construction.
First i would recommend to replace the IRF510 by a IRF540, IRF540N,
IRFZ48N or so. These are better for 14V and will give you more power.
The 7667 can handle them, i.e. their input C is still low enough. But, a
class-E has a defined output power at a certain voltage. Tuning is
critical. If your antenna is slightly detuned, or a C becomes warm, the
working point can quickly drift away. So you cannot really squeeze out
the maximum theoretical power unless you optimise things endlessly.
I would recommend a class-D PA running at 14V, using the 7667 as a
driver.... I have a running design, easily generating 100W at 12V. But
you need an opposite phase signal on the input and i don't know what the
output of the U3 is like. Is it a sine wave at 50 Ohm output impdance or
is it a rectangular high-Z output like the Raspberry provides? A small
ferrite transformer with a trifilar winding can provide that opposite
phase signal. It's all no miracle and sounds more complex than it is :-)
Yesterday I tried it at 24V. Little difference. Got 8 times antenna current.
Antenna current or PA DC input current? I assume the PA-efficiency
dramatically decreased in that test :-)
But tried to fine tune the loop
to squize some more power. It detuned for a moment and that was the end of the
mosfet :-(
No mercy at 24V. I have learned to tune at low power and then increase. Should
be interesting to add
a current limiter too. May be you can suggest other ways to protect the PA and
make it fool proof
An MF cirulator perhaps ;-)
All my PAs do not have a protection. I prefer to avoid unwanted working
points instead :-) But from time to time it can happen that a FET
explodes, in my case it alway has been due to a wrong input signal, not
due to overheating or detuning of the antenna. Thus i prefer the cheap
IRFP types instead of expensive and rare IXYS or other FETs. The IRFP460
works well in all my QRO designs. In 12V PAs i only use IRFZ48N types or
maybe sometimes a IRFP064N.
PS: CT1JTQ seems to be an Italian living in south Portugal
In the morning i got a first positive answer from him! Maybe he will
appear on MF soon (only RX-wise so far, they do not have a TX allocation
as far as i know)
73, Stefan
|
