# HF Digital Error correcting? Also, what's up with PSK31?

Discussion in 'General Technical Questions and Answers' started by N0NS, Oct 9, 2008.

Not open for further replies.
1. ### W5DXPHam MemberQRZ Page

Time to put up or shut up, Tim. The circumference of a tree trunk is a fractal length that can be many times greater than a perfect circle depending upon the species of tree. You could never find a tree with a constant diameter so what are you going to divide by? And since pi requires division, exactly how was that division accomplished when rain was falling on the dinosaurs? I'll bet you \$1000 that you cannot find a real-world tree trunk that even satisfies pi to 10 decimal places much less an infinite number of decimal places. Your statements are unreal, absurd, and maybe pathological. You see things in nature that do not exist in nature. I doubt that those raptors in "Jurassic Park" were even smart enough to do long division much less understand irrational numbers.

They're based on your argument, not mine. It was you who said a square wave graphed in a textbook is not a square wave, i.e. the name of the thing is not the thing. I have since been trying to get you to honor your own argument. Argon is the name of an element. So, just as you will not allow me to use the words "square wave" for the graphic of a square wave printed in a textbook, you cannot allow yourself to use the word "Argon" for the element. If I have to say "graph of a square wave" every time, you have to say "the element named Argon". It's your requirement, not mine. My last few postings have simply been based on your argument. If you cannot abide by your own rules, then please release me from them.

I was perfectly happy calling the graph of a square wave by its "square wave" name, as are many distinguished authors of technical books, but you objected.

I would have been perfectly happy calling the element named Argon by its name but, according to your above logic, that should not be allowed. So please, either honor your own requirements or allow the communications on this newsgroup to return to normal.

Last edited: Oct 27, 2008
2. ### G3TXQHam MemberQRZ Page

Tim, sorry you're absolutely wrong on this one!

"Risetime" has the dimension Time - the clue is in the name. By the Wikipedia definition it is the Time taken for the waveform to go from 10% of final value to 90% of its final value. Nowhere does the definition include the amplitude of the waveform. You are confusing it with something akin to "Slew Rate" which would have dimensions of Volts/Time.

Let's take an example - a squarewave between levels of 0v and 5v. Suppose it has a risetime of 1uSec. That means it takes 1uSec to go from 0.5v to 4.5v. A 162KHz sine wave of amplitude 5v peak will do a pretty good job of representing the leading edge of that square wave - it too takes 1uSec to go from 0.5v to 4.5v.

Now let's double the amplitude of the square wave to 10v. To represent the "new leading edge" I don't need to do anything to the sine wave except change its amplitude . I don't need to change its frequency. That's exactly what we would expect from Fourier analysis - doubling the amplitude of the signal doubles the amplitude of each Fourier component - it doesn't change the number of components or their frequency!

No matter what you do to the amplitude of the square wave, the leading edge can be replicated with the same frequency sine wave.

Let's take another example. I pass the square wave through a simple low-pass RC filter. The values of R and C completely determine the rise time irrespective of the amplitude of the input waveform and will also determine the highest frequency present. If your theory applied, every time we changed the amplitude of the squarewave we would get a different set of frequencies from the filter.

Are you saying that the cut-off frequency of a simple low-pass RC filter depends on the input amplitude?

Steve

Last edited: Oct 27, 2008
3. ### G3TXQHam MemberQRZ Page

Absolutely not! I would expect the same high frequency components at 12 times the amplitude. Once again you are confusing risetime and slewrate.

Steve

4. ### G3TXQHam MemberQRZ Page

Of course it doesn't - a 1000 Hz sine wave can reach 2 volts in exactly the same time provided you double its amplitude.

Steve

5. ### W5DXPHam MemberQRZ Page

Objective reality is what matters. I said the "concept of mass" did not exist before humans. To the best of my knowledge, humans are the only species capable of concepts. So how could any concept, including the "concept of mass", exist before any conceptual species existed that was capable of conceiving those concepts?

I'm sorry, Richard, but that is clearly delusional. A ratio, by definition, requires a division. If there is no species around that can accomplish that division, any ratio is impossible. Who sliced the sphere (in your comment above) and made the circumference and diameter measurements?

I certainly would entertain the idea of a conceptual alien species capable of performing those tasks but since we cannot prove they exist, I have chosen to ignore them.

Again, I apologize for my devil's advocate argument pulling you into all of this. That was not my intended purpose.

6. ### KC4RANHam MemberQRZ Page

I told you to leave the damned cat in the box. Going with some sort of "it doesn't exist until viewed/conceived/conceptualized" argument is useless on the macro world, and your attempt simply shows how behind the 8 ball you are.

It takes division, and a whole lot of other operations, to try to determine by what laws and such the planets and other large stellar objects are moving by, but the fact that they appear to have moved millions of years before humans existed shows that your argument is either pure bunk or pure diversion. A star doesn't need division to move in any pattern, but we do if we want to try to determine its pattern (not invent it).

The planets and the celestial objects moved before we got here. How they moved was determined by their mass, before we had the concept of mass. DXP-speak aside, that's the truth as far as science can tell. What next... a meaningless jaunt into perceptions and meta-existence? The planet doesn't really exist, it's all a painted picture on a sphere, we can't believe what we observe?

7. ### G3TXQHam MemberQRZ Page

Maybe not in the "old days". But with the advent of logic circuitry it became very easy. Think about it - Walsh Hadamard transforms only require the input signal to be multiplied by a +1 or a -1 before integration, unlike Fourier transforms which require multiplication by sine and cosine functions. WFTs are typically much faster than FFTs.

Steve

8. ### G3TXQHam MemberQRZ Page

Tim, gotta go to bed so I'll keep it simple:

http://coastworld.blogspot.com/2007/05/bandwidth-vs-risetime.html

Bottom line: Bandwidth = 0.35 * risetime

In other words the risetime completely determines the high frequency components of the square wave. Not a mention of amplitude anywhere

Steve

9. ### W5DXPHam MemberQRZ Page

Please don't try to falsify what I said. I did NOT say the phenomenon named "mass" didn't exist. I said the[size=+2]"concept of mass"[/size] didn't exist before humans. For any concept to exist, there must exist a species capable of conceiving those concepts. As far as I know, humans are the only conceptual species that has ever existed on this planet. As far as I know, zero concepts existed before humans.

What species do you think was capable of inventing concepts before humans existed?

The basic point is that concepts do not magically bring something into existence in the real world, i.e. we cannot think something into existence. A ratio is an second-level abstract concept that simply doesn't exist in the real world. The only place that I know of that the ratio of two numbers can possibly exist is in a human mind and not all human minds are even capable of conceiving that ratio.

Last edited: Oct 28, 2008
10. ### AB0WRHam MemberQRZ Page

No, rise time is (amplitude)/(slew rate)

The slew rate in this case, which defines the high frequency component is (4volt)/(1usec) = 4x10^6 volts/sec.

Steve, if I keep the same frequency components I still have a slew rate of 4x10^6 volts/sec. If the rise time is now 1volt (10%) to 9v (90%) then ff I divide (8 volt)/(4x10^6volt/sec) I get 2usec for the rise time or twice as long. as for the 5volt signal.

The only way I can keep the same rise time is to have higher frequency components so I get a slew rate of (8volt/1usec) = 8x10^6 volts/sec, or double the slew rate.

No, we would get the *same* frequencies and, therefore, the same slew rate.

Using an old nomograph I have, it shows that a 4x10^6 volt/sec slew rate with a 5volt peak requires about 100khz in bandwidth. It shows that an 8x10^6 volt/sec slew rate with a 10v peak requires about a 200khz bandwidth which is just about what one would expect.

Nope. I'm saying the rise time for a bandwidth limited signal is dependent on amplitude.

If you limit the bandwidth to 100khz then a 5volt signal will have about a 1usec rise time while a 10volt peak signal will have a 2usec rise time.

The high frequency components in an unfiltered keying waveform require two pieces of knowledge to be fully specified, the amplitude of the keying waveform *and* the rise time. Those two together determine the highest frequency component in the signal. You have to know both but I just don't remember very many test of amateur CW transmitters ever bothering to actually calculate the bandwith. They just look at the rise time -- period. If it's greater than 5ms it's ok, if it is less then it is suspect. I suppose those are reasonable "benchmarks" but are not very useful in comparing two different transmitter implementations for bandwidth impacts.

tim ab0wr