[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [linrad] RE: IC706MKIIG
It is my opinion that the K2 by Elecraft still leads the pack if what
you care about is receiver performance. If you need all the bells
and whistles of the Orion, then spend the money.
You may visit:
and click on the rig comparison tab to see a comparison of the Elecraft
K2 with many receivers, including the Orion, based on ARRL lab
measurements. The K2 is currently the best buy per Euro. By the way,
with modifications that will soon be forthcoming, and with an $800
soundcard (Lynx Studios L22), I can equal or exceed the performance
of the Elecraft K2 with a modified SDR-1000. The software defined
radios of the Linrad, SDRConsole added to a good RF interface such
as SDR-1000, R2Pro, etc. type are about to really take over the
For the information of those interested here, there was a design
flaw in the SDR-1000. Gerald put the noise figure setting amplifier
AFTER the mixer. Thus he is amplifying the phase noise and the spurs
of the DDS, thermal noise in the mixer, along with signal of interest.
He is still getting performance that I would not have believed possible
before we discovered the problem. Moving the gain out front of the mixer
along with better filtering, the SDR-1000 improves tremendously. I
believe at one time there was one or more people here attempting to
homebrew the commutating mixer/sample hold circuit no longer referred
to as Tayloe detector (there is much prior art). It is important to
set the noise figure ahead of the mixer (DUH!).
[mailto:owner-linrad@xxxxxxxxxxxxxxxxxxxxxx]On Behalf Of Leif Åsbrink
Sent: Wednesday, December 10, 2003 2:04 PM
Subject: [linrad] RE: IC706MKIIG
The IC706 may seem a bit outside the scope of the Linrad reflector.
If you are not interested, please stop here:)
> Some radios will momentarily put out full power on keydown when
> the power is turned down. In that case your 10 amp supply may
> breifly shut down in response to the peak current demand (some 20
> amps). I have not
> tested the 706 for this problem, but it is common. On 144 with the lower
> power final, your 10 amp supply will never be exceeded so it would operate
> properly there. A scope on your power supply lead would tell you
> if this is the problem
> or not. You could also power the radio from an idling vehicles
> power source
> and repeat the key click test.
It is not something like this. The supply is 13.7 V with a ripple of
10 mV (originating in 50 Hz). There is no visible change in the supply
voltage when the key is pressed (less than 2 mV).
> There is only one radio which has superb keying even on fast QSK
> and that is my Ten Tec Paragon. Amazing!
> you are comparing a cup of tea with a good glass of wine
> Maybe also the new Orion...
If the Orion is not superb I will become VERY disappointed.
Superb keying in "fast" QSK is not difficult at all. The IC706MKIIG
is in fact very close to that technically. On 144 MHz there are no
keying clicks in full break-in mode. When I look at the spectrum
of the carrier and then switch to 50% duty keying, I can see
the level go down at all frequencies except for a narrow region
of about 700 Hz at each side of the signal. Switching to full
break in does not change the pattern at all at modest keying speeds.
At higher speeds the spectrum drops even bit lower at all frequencies
because the turn on delay reduces the duty cycle.
To have QSK with really fast keying one will need pin diodes, but
within what is possible with a relay the IC706 is superb on 144
MHz but it is awful on 14 MHz.
These are the problems with the IC706:
1) Keying is done at two different places (or more). One stage is
softly ramped up until the power is stable at about -60 dB. Then
another stage is responsible for bringing the power up to full power
using a well filtered waveform. The relay switches in the antenna
when the power is at about -70 dB on the first ramp. Just by shifting
the DC voltage a bit so the first ramp is still at zero when the relay
switches will remove the wideband keying clicks that are caused by the
relay. Technically this is trivial. But it will not make much difference
because the wideband keying clicks have a peak power of 75 dB below the
carrier in 2.4 kHz bandwidth. In a narrow bandwidth this is below the
sideband noise of the unmodulated carrier. I mention it because it
is a silly error that is easily corrected. With a better VCO it would
make a difference.
2) Each time the key is released the ALC voltage is reset. This means
that the ALC has to adjust the gain quite a lot at the onset each time
the key is pressed. The ALC is a servo system with a fast attack and
slow release characteristic. The ALC adds about 15% AM modulation
with a frequency of about 5 kHz, a damped oscillation with a logarithmic
decrement of about 3 dB. The keying clicks 5 kHz on the sides therefore
have a peak power only 23 dB below the carrier in 2.4 kHz bandwidth.
This is really terrible to listen to! The 5kHz has overtones, the keying
clicks cover a 25 kHz wide frequency range. To fix the problem one should
simply reduce the drive level. A resistor somewhere to reduce the gain
to a level where the ALC will no longer modulate the signal.
My Linrad project gets diverted all the time. The above is the result
of my activities to try to promote better test methods for tx testing
in product reviews. The terrible keying clicks should not come as a
surprise to the innocent amateur who buys a brand new radio.
Other rigs have excessive splatter in SSB mode, big wideband pulses
when the PTT is pressed and so on.
The WSE hardware becomes useless on the bands if my neighbours do
not have good transmitters....
Leif / SM5BSZ