[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[linrad] Info: RX144
- Subject: [linrad] Info: RX144
- From: Leif Åsbrink <leif.asbrink@xxxxxxxxxxxxxxxxxxxx
- Date: Tue, 15 Jun 2004 13:24:56 +0200
The RX144 problems are solved. The most common noise source
was due to inadequate wetting on the component side due to
a too short heating time or too low solder temperature.
The plastic foil trimmer capacitors should not be washed
with water after soldering and therefore all trimmers are
hand soldered in the WSE units. Because of the large number
of trimmers in the RX144 unit our subcontractor decided to
hand solder all the components. I was surprised to find that
the undefined contact point between a wire and the tin causes
modest sideband noise that would go undetected in a "normal"
receiver but that degrades the RX144 by typically 3 dB.
I have resoldered seven units myself and verified that they
all have similar sideband noise levels at a frequency separation
of 100 kHz. Antennspecialisten has preliminary orders for
these seven units and they will be ready for shipment this week.
The remaining 18 units are sent back to the subcontractor and
will probably be ready for shipment in about one month.
The complete WSE/Linrad receiver for 144 MHz has a sideband noise
level below -165 dBc/Hz at 100 kHz frequency separation.
The LO design used in Linrad was inspired by an article in Dubus
by DK4XP who showed that the noise floor would be at -160 dBm for
an optimised oscillator. The oscillators of the WSE units are about
10 dB better than this because they do not have any noise source
(resistor) at the amplifier input. (These oscillators are
amplifiers with positive feedback through a crystal filter)
I have been proud of the LO design of the WSE units until the
Scandinavian VHF meeting last weekend when LA6LCA showed me
how to make local oscillators that are 10 dB better, with a
wideband noise floor at -180 dBc/Hz or below. Replacing the
local oscillators in the WSE units will not give a very large
improvement because there is also sideband noise generated by
the amplifier stages, but that can probably be handeled (at some
cost) so I think a 10 dB improvement will be possible. I will not
make any attempt to make better converters within the forseable
future, but I think it is appropriate to say that the WSE converters
by no means represent the best possible solution. They are good
enough for my own use though.
When using the WSE converters as a 144 MHz receiver, the noise
floor should be 10 to 20 dB above the internal noise floor
of the WSE/Linrad system and then the dynamic range will be
determined by the 1 dB compression point which is at about
+10 dBm at the RX144 input. The noise floor of the entire
WSE/Linrad system is at -164 dBm/Hz (NF=10dB) and 50 % of the
noise originates in the Delta44. When a preamplifier with
25 dB gain and NF=0.5 dB is added, the noise floor referenced
to the RX144 input will be at about -149 dBm/Hz so a signal
that would cause 3 dB loss of S/N due to "reciprocal mixing"
would have to be about +16 dBm which is well above the point
The dynamic range of the Linrad/WSE system with a 25 dB preamp
that places the noise floor 15 dB above the WSE/Linrad noise is
thus about 159 dB(1Hz) which corresponds to 132 dB in 500 Hz
and it is limited by amplifier/mixer saturation for signals outside
the visible window. The system NF at the antenna is then
0.6 dB if the preamp noise figure is 0.5 dB.
For signals within the 93 kHz wide visible window the dynamic
range is about 30 dB lower. Very few VHF transmitters have
sideband noise levels below -130 dBc/Hz so the purity of the
interference source will be the limiting factor. With
10 dB less RF gain, (a switchable 10 dB attenuator) the
WSE/Linrad noise floor can be lowered by nearly 10 dB for
a dynamic range that no commercially available ham transmitter
can meet today. The drawback of using 10 dB less gain is of course
that the noise figure would rise to about 1.5 dB.
Please note that the WSE/Linrad receiver has a dynamic range that
is high above what one usually needs in amateur radio. Other
solutions allow much lower cost and as long as the undesired
signals are below the limits set by the dynamic range there is no
disadvantage of having a much lower dynamic range PROVIDED THAT
THE SYSTEM RECOVERS IMMEDIATELY AFTER SATURATION.
Note that the peak power levels caused by powerline noise may
become extremely high. If an amplifier that goes into saturation
does not recover immediately, all strong signals will become
amplitude modulated by the recovery waveform. Such amplitude
modulation adds more energy to the interference pulse and may
make the Linrad "clever" noise blanker useless.
When used without RF amplifier, the WSE/Linrad system is an excellent
tool for checking transmitters. I hope that some of the units
will be used for this purpose and that results become available
on the Internet from several sources. The commercial transceivers
can easily be made much better at no increase in production cost
but the experience over the last 30 years shows that such improvements
will not become available unless there is a market demand.
Hopefully an increased awareness of what emitted spectra look like
will help a little in this respect.
Leif / SM5BSZ