N4LQ wrote:
>
> I've noticed a big difference in response to qrn when comparing
> direct, single and multiple conversion rigs. With dc receivers,
> qrn is almost non existent, with single conversion i.e.. Omni-C,
> qrn is hardly a factor, with multi conversion rigs, it starts to
> become bothersome ie. Omni V etc. and with quadruple conversion,
> it's unbearable i.e. TS-870 dsp etc. That's why I'm using this Omni-C.
> It's giving my ears a good rest. Band noise is just not a factor.
> Maybe you could elaborate on this Jerry.
The claims to fame for the DC receiver do include low noise. I think
there are two reasons for that low noise perception. First their overall
gain is significantly lower than a superhet. Their RF sensitivity,
generally without a RF stage, runs 20 dB poorer than the most sensitive
of superhets. That alone keeps their output quieter though it often
doesn't hurt copying signals that on the lower frequencies are limited
by atmospheric noise many dB above the receiver threshold. The phasing
type then needs good audio filtering to limit the bandwidth to that of
the audio phasing network, and needs control of phase linearity to make
the phasing successful. So they tend to do all that at very low level so
there's little distortion or over drive of the detector from antenna
supplied noise.
The problem with a superhet is the mixer. Tubes were especially bad for
having low gain and high noise and limited dynamic range. This is still
a problem in the double balance mixer most commonly used in state of the
art receivers today. While the mixer noise isn't quite as bad, the
mixers have significant loss and unless very high level local
oscillators are used they have a limited dynamic range. Part of the
noise problem is that mixer input circuit noise, even if the signal
images are removed by input filters, the mixer converts circuit noise
from both the image and the desired, plus local oscillator harmonics.
Efforts to remove those by tuning at the mixer cause other worse
problems. The noise and low gain (or even loss) of the mixer requires RF
gain before it. Its possible to overcome the mixer noise, at the expense
of high end dynamic range. Noise pulses and static pulses have very
large amplitude and very short times. They easily drive a mixer after an
RF stage into saturation and so cause noise pulse energy at the IF that
is larger than proportional to the bandwidth of the IF because of that
mixer overload. To make things worse, its been shown that unless a
double balanced mixer has all three ports terminated resistively over a
wide range of frequencies, that the mixer loss will increase and the
dynamic range decrease making these problems worse. This is like the
mixers in the Yaesu FT-736 that are coupled directly to a crystal filter
in the IF. Their poor dynamic range and high noise is a real pain to the
user.
More mixers make these problems worse. The mixer noise floor and losses
demand more gain before the filters and then hit the later mixers with
stronger signals and noise pulses which the often can't handle.
Then in the synthesized radio there's another serious problem, the noise
sidebands of the LO. These cause noise from outside the IF bandwidth to
be mixed to the IF bandwidth, as well as adjacent signals, to be noise
modulated and some of the products to fall into the IF lowering the
signal to noise ratio of the received signal in the IF.
According to Ulrich Rohde's latest book on Communications Receiver
design, (Bookpool on the internet had the best price last year) the best
receivers use a multiple feedback RF stage (three winding transformer
and a bipolar transistor) and a passive double balanced mixer with all
mixer ports terminated broad band. These are the circuits of the Corsair
II which also has no synthesizer and hence minimal local oscillator
noise.
While I could use the transverter ports of the Omni V, I don't
appreciate the local oscillator coming from a synthesizer, even one
proclaimed quiet, and I don't appreciate it running LSB for CW on 28 and
14 MHz where I'd like to run transverter IF for VHF and up use. Around
here its important to be able to switch modes rapidly and changing
sidebands between SSB and CW I find more intolerable than the
stupidities from Yeasu, shifting frequency 800 Hz.
I will add transverter connections to my Corsair II. And so far I really
like its performance on HF, it makes the old Kenwood sound very noisy
and the pulse responses of the filters don't turn static into crashes
and so I can copy through fairly bad 75 meter noises.
Turning on the noise blanker to get rid of line noise generally brings
up lots of monkey chatter and there's no solution for that, other than
to prefer to listen to a bit of buzz than to monkey chatter. I have some
ideas that I'll have to work on for taking care of such noises.
Those are my opinions, not everyone will agree, else all radios would
have the same circuits as the Corsair II without umpteen conversions.
Nope, I didn't contribute to the Corsair II's design, but I was
beginning to plan to build my own transceiver until I found it was
already completed in the Corsair II.
73, Jerry, K0CQ
--
FAQ on WWW: http://www.contesting.com/tentecfaq.htm
Submissions: tentec@contesting.com
Administrative requests: tentec-REQUEST@contesting.com
Problems: owner-tentec@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|