On Sun, 16 Jan 2011 00:38:07 -0800, David Nebenzahl
wrote:
On 1/15/2011 10:14 PM Jeff Liebermann spake thus:
http://en.wikipedia.org/wiki/Direct-conversion_receiver
Before I blunder onward, permit me to say that I've locked horns with
the censors on Wikipedia (actually Wikibooks) and got a first hand
taste of why some of the articles are rather marginal. However, it's
still the best reference around for obtaining general information
about almost any topic and tends to be more understandable than
someone peer reviewed research paper or marginally reviewed book
extract. In topics that I am familiar, I can usually find something
in the article that could use improvement. I usually include a
Wikipedia reference primarily so that those unfamiliar with the topic
can get a general understanding.
OK, so this is why I absolutely *hate* Wikipedia.
Nothing is perfect. Finding errors is not sufficient grounds for
hatred. A lawyer friend has a similar problem with the various laws
and legal decisions. All of them could use improvement and almost any
written law can be misread and misinterpreted. Unlike Wikipedia,
readers are unable to repair the written legal system. Despite
chronic deficiencies, the legal system is still functional, and there
are few attorneys that *hate* the written law simply because there are
mistakes. Some tolerance would be helpful here.
Here's the lead
paragraph in the article:
In telecommunication, a direct-conversion receiver (DCR), also known as
homodyne, synchrodyne, or zero-IF receiver, is a radio receiver design
that demodulates the incoming signal by mixing it with a local
oscillator signal synchronized in frequency to the carrier of the wanted
signal. The wanted modulation signal is obtained immediately by low-pass
filtering the mixer output, without requiring further detection. Thus a
direct-conversion receiver requires only a single stage of detection and
filtering, as opposed to the more common superheterodyne receiver
design, which converts the carrier frequency to an intermediate
frequency first before extracting the modulation, and thus requires two
stages of detection and filtering.
Now, class, how many things are wrong here? (And please correct *me* if
I'm incorrect):
o First of all, superhet receivers have only one stage of detection and
filtering, not two, after the last IF stage, right? (I suppose there may
be some filtering in or around the mixer stage, but I don't think that's
what they're claiming, which I assume is filtering out the carrier.) So
where do they get "two stages of detection and filtering"?
The original paragraph is poorly written and you managed to
misinterpret it. I'll try to do better.
A direct conversion receiver usually uses only a single stage for both
detection and filtering. In any receiver, there is only one stage of
detection. The articles reference to
"...and thus requires two stages of detection and filtering"
should read
"...and thus requires separate stages of detection and filtering". Is
that better?
o Is their explanation of how DCR works even correct?
Yes, it's correct. Mixing with a local oscillator (or reference
signal) on the operating frequency, as in a homodyne receiver, is
considered direct conversion. That includes extracting the carrier
from the receive signal, and subsequently mixing the carrier with the
receive signal to extract the modulation (such as in an I-Q
demodulator). Mixing with a signal that is NOT on the operating
frequency, is heterodyne conversion. Note that it doesn't matter
where the signal is mixed in a (super)heterodyne receiver. One active
stage can do everything as in an autodyne receiver.
http://en.wikipedia.org/wiki/Autodyne
I don't understand
the business of mixing the signal with a LO signal: why would you do
that?
The mixing is to extract the modulation. It's called a product
detector for AM and SSB.
http://en.wikipedia.org/wiki/Product_detector
http://www.google.com/images?q=product+detector
If you extract the carrier signal from the receive signal, mix these
together, and low pass the result, you get demodulated AM or SSB. With
slope detection, you can also demodulate FM. Extracting the carrier
is simply lots of amplification so that the amplitude modulation is
clipped (limited) and thus removed. For SSB, a PLL is often helpful
for weak signals, but not really necessary as just IF noise will
usually suffice to produce a usable carrier component.
They're a little vague: does "synchronized in frequency to the
carrier" mean *exactly* the same frequency as the carrier (???),
Yes. Exactly the same frequency. In order for a product detector or
direct conversion receiver to work, it needs to multiply (mix) the
carrier (with no modulation components present) with the receive
signal. What's left is the modulation.
or some
other frequency to produce a sum or difference frequency? (In which
case, we're back to IF, aren't we, so what's "direct conversion" about this?
No. The distinction is that direct conversion uses a mixing signal
that is exactly the same as the receive signal. If the signal was
offset, it would be consider (super)heterodyne conversion.
If I were in front of a firing squad and had to try to describe DCR
without actually knowing what it is, I'd guess(tm)(R) that it's a bunch
of tuned RF stages followed by a detector.
Sure. The detector is allowed to use the received signal to perform
the demodulation, it's still direct conversion. Think of TRF (tuned
RF) type of receiver as a special case of direct conversion, where the
demodulator is rather simplistic.
Anyhow, I think I've shown that even if I'm way off base, Wikipedia
articles tend to be extremely badly written,
There's plenty of room for improvement. There may be mistakes but
they are NOT badly written.
if not outright full of
doubtful information.
They make an effort to reduce errors. Where it become difficult is
that there are so many areas of technology where there are multiple
points of view with large areas of overlap and controversy. As long
as the source of such "doubtful" information is specified, multiple
points of view are presented, and the author is fairly neutral, I
don't have any problem with presenting controversial information.
What else would one expect of the "encyclopedia"
that any PlayStation-playing, junk-food wolfing pimple-faced
junior-high-school student can edit?
The next time you research a topic, instead of using Google or
Wikipedia, try using Google Scholar instead.
http://scholar.google.com
This should give you a wide selection of papers and articles written
by qualified experts who probably don't own a Playstation. Some of
the papers have been peer reviewed and are thus deemed correct and
often even authoritative.
If that is insufficiently accurate, try searching for terms using
Google patent search.
http://www.google.com/patents
If you think Wikipedia is full of inaccurate information, wait until
you read some of the hogwash found in some patents. Try searching for
"perpetual motion" for a good start.
--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060
http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558