Examining MPEG & FEC Part 2: Signal Processing
ALL ABOARD!
Let’s take a ride on the signal train. Our journey ultimately begins at the source provider, but for out purposes well start at “Mid-Air Station” and see what kind of “reception” we get at “Receiver Station”.
It may surprise some of you that we’ve boarded Analog Railways for our journey. I realize their service in the past was less than desirable, but they’ve revamped the whole thing. They’ve added this new, comfortable, first class upgrade from the old Analog Railroad. They’ve even changed their name to Digital Railroad and they say it’s a whole new railroad. I’m excited! (Analog Railways has gotten a bad reputation in the past, but they have assured us that the new Digital Railroad has really smoothed out the ride…)
We are approaching “Antenna Central” where Analog Railways will go through “Noise Pass” on its way to “Receiver Station” where we find something very interesting. The welcoming committee at Receiver Station call themselves the “ADC”, the “Analog to Digital Converter” crew.
A Little Bit about Modulation
This is a very simplified summary of some primary types of modulation. I am aware that it lacks volumes of information…it is supposed to. (I’m obviously using lossy compression!) :0]
When an incoming signal is received by a meter or your receiver, it arrives as an analog signal. It is an analog signal called a carrier wave. A carrier wave “carries” the information payload waveform which has been modified, or modulated. These two signals are the “analog tracks” of communication.
Prior to the digital debacle, we would modify the characteristics of the analog payload waveform to represent information. We have modified or “modulated” the signal in several different ways. We’ve done it by increasing and decreasing the height or amplitude of the wave and this we called AM radio. We’ve also modified the wave by varying its frequency and we called it FM radio. Changing the phase of the wave gives us Phase Modulation (PM).
Well, now we have modified the payload to (or with) numbers, and instead of continuously sending the information, we send bits of it at a time. It comes in pulses like Morse Code. It is called Pulse Code Modulation (PCM). Where the train used to be one long continuous car, it is now smaller trains of many cars (compressed) that arrive one after the other.
Digital Receiver and Meter Architecture
Whether it is your digital receiver or your digital meter, the first thing signal encounters is the Analog-to-Digital Converter (ADC). There would, of course, be no need for the signal to be converted if it were already digital, but the reverse of what happens here, is exactly what happened back at the broadcast source.
When an ADC is used, it is used in conjunction with two other complimentary components. The ADC does just exactly what it says and converts the incoming signal from analog-to-digital. From the ADC, we now have digital information to process. To process digital information we must send it to the Digital Signal Processor (DSP) for processing. (Here is where digital is found…and nowhere else.)
A DSP is a little computer. It takes the digital information received from the ADC and de-codes the information. The DSP uses codecs (coding/decoding algorithms) to decode all that compressed information. (Codecs are the “spy decoder ring” of digital communications.) At last we’ve got a large amount of decoded digital information!
To present that information on your screen or on your meter, we have one last conversion that needs to be done. This decoded digital information is sent to the Digital-to-Analog Converter (DAC).
To the WHAT!?!?! To the Digital-to-analog converter!
Guess what that does…
The digital-to-analog converter converts the digital information to ANALOG once again. Why? The information needs to be analog because the rest of the system, the Rf waves, TV screen, or meter is all analog – continuous. (I found this to be a little disturbing.)
At the broadcast station, it is the same. They used the ADC, DSP, and DAC to process the information so that it could be transmitted over the analog RF airwaves.
This information is pre-requisite for Part 3: Measuring BER
Jeff, you wrote:
When an incoming signal is received by a meter or your receiver, it arrives as an analog signal. It is an analog signal called a carrier wave. A carrier wave “carries” the information payload waveform which has been modified, or modulated. These two signals are the “analog tracks” of communication.
—————
I’m still chewing on this. Two analog signals? I understand the carrier wave, but what’s the second “analog track”? Would that be the Pulse Code Modulated ‘payload’? I thought being an on/off signal (like Morse Code) it would always be considered a ‘digital’ signal.
I’m probably missing something right in front of my face (perhaps your point about the DAC) so please bear with me and thanks in advance,
Jim
PS OK, (hours later) I have reread this at least a dozen times wondering about my earlier question: I now wonder if your statement about the carrier being ‘analog’ is misleading: isn’t a blank/empty carrier just a carrier (just an empty (rf) transmitted signal) until there is added modulation? Please wake me up.
Jim:
I think he means the two signals are the carrier and the information that modulates the carrier. From your ham days, SSB is a form of analog modulation and the audio that modulates the carrier is analog.
This might help:
What Exactly is 8-VSB Anyway?
In particular, note in above link:
FIGURE 1: BLOCK DIAGRAM, 8-VSB EXCITER
AM MODULATION
ANALOG UPCONVERSION AND THE REST OF THE 8-VSB CHAIN
Also:
http://www.tsreader.com/mpeg/index.html
scroll down to What’s QAM and Vestigal Sideband?
The people that live at the edge of the “cliff” don’t think that the ride is very smooth; they would rather have the little bit of analog snow back.
Jeff:
Please shoot me down if I’ve got it wrong.
Looking forward to part 3, one of my favorites.
Get your chewer ready, Jim.
Here’s a preview of part three:
One track is the analog carrier wave.
The second track is the “ANALOG lock” lock signal. The one that determines signal lock or not. This is analog, too.
The digitally modified signal that contains digital information, rides on the tracks like a train. The train is a message that contains many cars of information and the trains arrive over and over.
Better?
Rabbit is correct that this has to do with sidebands and type of digital modulation. (As I said in the writing, my short summary of modulation doesn’t address the multiple types and complexities of digital signal. Hopefully, it contains just enough information to move us along and not too much to get bogged down.)
Thanks guys! I need questions to make sure things are understood.
HighdefJeff
I did leave out something. When I wrote:
“When an incoming signal is received by a meter or your receiver, it arrives as an analog signal. It is an analog signal called a carrier wave. A carrier wave “carries” the information payload waveform which has been modified, or modulated. These two signals are the “analog tracks” of communication.”
It should read more like this
“When an incoming signal is received by a meter or your receiver, it arrives as an analog signal. It is analog signal that has a carrier wave and a signal lock wave. A carrier wave “carries” the information payload waveform which has been modified, or modulated. The carrier wave has the digitally modified information on its back, but next to it is the analog “signal lock” signal. These two analog signals make up the railroad tracks that come to your antenna.”
Better?
What you refer to as the analog “signal lock” signal…..is that the same as the pilot carrier/signal at the low end of the channel?
http://www.tek.com/Measurement/App_Notes/25_15787/eng/25W_15787_0.pdfsee Reception and Decoding, pdf p10, document p6
or are you including ATSC segment and field syncs (also called clock signal extracted from a data sync segment and framing code segment)?
http://www.8vsb.com/doc/harris-8vsb.pdfSYNC & PILOT INSERTION
http://www.tvtechnology.com/article/11608
I can’t say for sure (haven’t been able to download pdf’s for a while.), but the pilot/carrier signal sounds correct.
I can say that it is not part of the data sync segment. That is decoded/decodable data in the digital signal. The loss of the clock data or a corrupted bit there, is what commonly causes lip sync.
The purpose of this segment was to emphasize the analog nature of the incoming signal.
I am not attempting to describe all of the signals components, but rather to pull out three important parts of the signal, the understanding of which, will make a difference at home or in the field.
Summary:
For simplicity, I narrowed this down to the “signal lock” portion, the carrier wave, and digitally modified message. The first two I liken to the tracks and the third, the train.
Upcoming…Part 3: What good is BER?
Jeff:
Thanks for emphasizing the analog nature of the “digital” signal. It is what has made it possible for me to measure digital signals with my old analog signal level meter, in spite of the fact that so many people have told me that I can’t do it!
It is true that my analog SLM doesn’t the give the precise value as a digital SLM would, but it does give useful readings to compare and aim antennas. I also have a digital SLM and my analog SLM gives readings that are close because it reads a little too high for analog!
I’m looking forward to your thoughts on BER and signal quality. I hope that you will include mention of the shape of the signal and the effect of multipath reflections, especially in urban areas and weak signal locations.
I don’t have a spectrum analyzer but my Sadelco DisplayMax 800 SLM is able to do a single channel scan. The attachments are two screen shots showing before and after moving a VHF-hi folded dipole antenna to a better location. The spike at the low (left) end is the pilot. The second set is in B&W in case you want to print them; they aren’t as pretty, but they print better.
If you want to print them, the B&W ones do better.
Rabbit and Jeff,
Special thank to Rabbit for the link you posted above helped. I spent over an hour last night writing and rewriting an eloquently worded thanks to you with links and quotes from other sources and then the darned Forum or something froze … grrrrr! All was lost.
About the time Jeff (thank you) wrote his first chapter, I decided it might be time for me to go back to the (new?) basics and try to shed misinformation or obsolete things I learned in the past and I found a perfect quote on an online Electronics Dictionary that stated (poorly paraphrased, cuz I lost the link) a digital signal can be expressed as an on versus off pulsewhich properly timed can be part of a data stream … (like Morse Code- on or off) OK, I knew that (duh) and it follows so many comparisons of a HDTV signal to Morse Code that I have read online, in old textbooks or what I have been told. Disclaimer: (I am not referring to the on-off switching of a carrier wave, only the code).
However, Rabbit’s link clearly explains 8-VSB is not simply “on or off” at all!
Far better said (instant to instant) it may not be on, then its’ more on, less on, way more on, way less on … etc – over a dynamic range. Again, not simply on VS off based on a centerline of zero, as might be seen on a scope. On versus off is apparently old school!
Yes, in relation to the past pulse (position?) received the next ‘pulse’ could be said to be off or on in relationshipto that PRIOR moment, but since it does not always ‘step’ in single steps, that looks exactly like amplitude modulation to me and the simplistic analogy to Morse Code is wrong.
Next, Rabbit’s analogy <-(similar word to analog, eh?) to DSB / SSB Ham radio also fits with that what that website shows regarding filling the 6 mHz channel. In fact, that night I dreamed up a way to plant tons of data in a narrow slot and I woke up with single sideband in my head. Thanks, Rabbit! Jeff and Rabbit, I’m far from ‘there’ but getting closer bit by bit (pun). Thanks!
Jim
PS Now it’s time to look at Jeff’s last posts … I hope I don’t sustain too much brain damage. LOL!
The point of all of this is to stop the “all or nothing” mentality and to establish a working understanding useful to the general public, installers and the DIY’er.
All-or-nothing is a broad, over-generalization of the workings of digital systems. It is pretty much useless with the possible exception of presenting the “idea” or “theory” of digital processing.
Not even at the bit level are the bits all-or-nothing. They are “read” and interpreted as a 1(all) or 0(nothing), but the characters don’t actually equal 1 or zero. Their energy is assessed and assigned a 1 or 0. That is a judgment call made while decoding. Closer to 1 = 1. Closer to 0 = 0.
(And if you happen to have a positron fly through and flip your 0 to a 1, you have an error, whether close to 1 or not.)
Rabbit,
My Q exactly, but you beat me to the punch!
Jim