EMRFD Message Archive 178

Message Date From Subject
178 2006-10-23 16:38:47 Rick SSB-DSB-AM
An excellent question was posted by jr_dakota the other day: Do high-
level-modulated AM rigs suffer the same distortion as SSB and DSB
rigs? I have studied this quite a bit and I don't know the answer--
but I can contribute some thoughts and experience, and perhaps
wisdom.

When I design an ssb exciter for the ham bands, I expect the IM
products to be enough lower than the amplifier IM products that
amplifier distortion will dominate--see figure 9.14 in EMRFD. That
means I usually expect all of the exciter distortion products to be
significantly more than 30 dB below either tone of a two-tone test
signal at the PEP output level I specify for the exciter. The 3rd
order products are often the largest distortion products, but 2nd
order distortion in the audio circuitry will also contribute.

Filter exciters often have significant distortion at audio and at
the IF before the filter. In fact, hard clipping is often
introduced intentionally to increase the average-to-peak ratio of
the signal. The signal from a filter exciter appears clean on a
spectrum analyzer, because a narrow filter knocks everything outside
the desired SSB channel down by 50 dB or more and the in-channel
distortion is buried under the desired signal, though it may be
plainly audible. Phasing exciters need low distortion from mic
input to antenna, because they don't have a filter to clean up the
off-channel distortion products. Because the transmitter has to be
linear to reduce the off-channel distortion products, the in channel
distortion is very low as well. All off the distortion products from
a good phasing exciter may be 30 to 50 dB down from the desired
signal--much lower than the typical 3rd and 5th order IM products of
a typical linear--but still big enough to see on a spectrum
analyzer. A phasing exciter that sounds great on the air will often
look surprisingly messy on a good spectrum analyzer, because of all
those low level distortion products that are not suppressed by a
narrow bandpass filter. In my station, I use phasing exciters at
VHF and UHF, and for HF nets and pleasant conversations. In a
contest situation, where the temptation to turn up the mic gain is
overwhelming, I stick to filter rigs. But processed speech grates
my ears, so I seldom use the filter rigs.

In my opinion, DSB should only be used on the ham bands where the
probability of interference is low...BUT that is often the case. 6m
and 2m are seldom crowded enough that stations are stacked 3 kHz
apart. Here in the northwestern US, there are many times of the day
when the HF bands are apparently open but no one is on. A simple
backpacking DSB exciter that allows you to get on the air from an
interesting location is a most welcome addition to the bands, and at
backpacking power levels, you are unlikely to interfere with a QSO
in progress several hops away. I would much rather work a guy in a
snow cave running a DSB transmitter than sit in the shack on New
Year's eve listening to atmospheric noise because the guy in the
snow cave didn't think his rig was good enough to put on the air.

High level AM modulation is different. Writing out the non-linear
equations for the distortion products is a real chore, and
measurements to confirm the validity of my math have been
inconclusive. AM unquestionably sounds good. Like cholesterol,
there is good distortion and bad distortion, and maybe high level AM
has mostly good distortion. The usual way to demodulate AM, a diode
detector, contributes even more distortion. So my experience is
that, yes, AM has real distortion products, maybe even more than DSB
or SSB by the time you high-level modulate the carrier and then
detect it with a diode. However, in my experience, AM sounds better
than DSB, in the same bandwidth, even when using a coherent
demodulator. The carrier may not contribute anything to the
information, but it provides a precise phase reference for
demodulation, a precise amplitude reference for automatic gain
control, and a strong signal that occupies the channel and
suppresses noise between words and syllables.

There is enough interest in vintage AM rigs that frequencies are set
aside for AM operation in the various HF band plans. Whether or not
one can successfully check into an AM net on 7290 with a high-level
modulated W7ZOI MKII transmitter (see April '06 QST) probably
depends on how hard the guys are listening for you. If they are
running DX-60s, you'd have a good shot. Running AM on frequencies
normally occupied by SSB stations might generate a negative comment
or two.

DSB is a little more compatable with SSB, since there are no carrier
whistles to tune through. If you answer a SSB CQ with a DSB
transmitter, the operator on the other end will never be aware that
you are transmitting DSB. If you call CQ, he may note something
unusual as he tunes in your signal.

So my wisdom on the subject, such as it is, boils down to this: use
a modern SSB appliance when (over)driving a kilowatt linear during a
contest. Use a phasing SSB exciter, backed off to keep the
distortion products low, if you want people to remark on the audio
quality of your signal or if you are chatting with friends and want
to sound like yourself. DSB and AM are both fine where there is
room for 6 kHz wide signals--but AM signals should probably stick
close to frequencies where AM guys hang out.

....and distortion? Well, that's why we do experiments. Feed your
transmitter into a dummy load through a pickoff so you can listen to
it on a good SSB receiver with a SSB bandwidth IF filter. Tune
around the bands to find strong signals that sound good. Now tune
around your own signal. How does it compare?

Finally, regarding the ARRL Challenge.... I like the idea of
starting with basic CW capability and then adding modulation. DSB
and AM would appear to be equally easy to apply to a low power CW
transmitter. AM requires some modulator power, and DSB requires a
little thought to amplifier linearity. With AM you will want to
stick close to the AM calling frequencies, but a crystal for 7290
kHz will guarantee that. For DSB, there are several exciter
examples in EMRFD, and the 0.5 watt linear amplifier in EMRFD figure
2.93 is a gem. My personal preference is high quality phasing SSB,
but keeping the whole station under $50 then becomes a real
challenge.

My experiments described on EMRFD page 9.11 confirm the old wisdom
that copying a DSB or AM signal on a DSB direct conversion receiver
is less than acceptable. So I would argue that an AM transmitter
must be paired with either an SSB or AM receiver. But until there
are are more DSB suppressed carrier signals on the air, a DSB direct
conversion receiver and CW-DSB suppressed carrier transmitter should
be fine. You will be able to work some people and not others, but
that is always the case when running low power.

Have fun with the experiments.

Best Regards,

Rick KK7B