Allan Songers SA-100 on the scope

[NOSValves]

As you can see from this AC shot that the lost bass response from input to output is identical to the 20hz shot using DC coupled above about 1/2 to 3/4 a line. If anything using AC coupling makes my work look less impressive. I agree that DC does do a better job but I hardly would use something that makes my work look worse to decieve or mislead !

[sfogg]

Craig,

Instead of just trying to prove me wrong take a few minutes to try to consider I may be right.

"I am going done later and showing you that the difference or amount of lost response using AC is indeed the same as DC the scope does not add a filter"

OK, now you really need to take a step back and consider what you are saying as you are wrong on a very fundamental level now...

What is the difference between AC and DC coupling?

Simply DC coupling allows DC into the scope, AC coupling does not. Same as the difference between an AC coupled amp and a DC coupled amp.

How does AC coupling block DC? By applying a high pass *filter* to the input of the scope. Most likely a simple blocking capacitor which is a first order high pass *filter*.

" it just see's the square wave differently."

And why is that?

Because the measuring bandwidth of the scope has changed when you apply a filter to the amp(EDIT: meant scope). If the square waves are high enough in frequency above where the AC coupling filter is there will be no difference between the two. However when you try to measure a square wave at low frequency that is near the filters frequency (and remember square waves have lots of harmonics) you are filtering out some of the square wave itself. THAT IS WHY IT LOOKS DIFFERENTLY , THE FILTER HAS ALTERED THE WAVE.

" I bet its Identical % difference input to output."

Fine, do your tests and report back.

Heck, you can do a very simple test by inputting a 1hz SINE wave into your scope. Don't alter the input level in the scope or the vots/division. Compare the amplitude of the wave AC vs. DC coupled. You will find the AC version is at a lower amplitude then the DC version. Why? Because of the filtering in your scope when AC coupled.

Shawn

[sfogg]

Craig,

" using DC coupled above about 1/2 tp 3/4 a line."

In other words they are different.

You will also note that the distortion of the amp is WORSE the the distortion introduced by the filtering in your scope... that follows what I said above. Try doing it a little higher up in frequency (though that can change quickly since you are both moving further away from the filter in your scope as well as getting higher up in frequency and more into where the amp will perform better).

" If anything using AC coupling makes my work look less impressive."

I disagree when you compare each picture just to itself not between the two. In the first one you see very nice square waves and gross distortion of it.

In your second picture you see waves with a distorted top as the 'reference' and then more waves with a similar shape but a little more slant to them. The second picture by itself makes the amp appear to be much closer to the reference point then the first picture compared only to itself.

Shawn

[sfogg]

Ryan,

"Using AC or DC coupling to measure square wave response results in 0 difference, they are identical in the audio band."

Obviously they are not since the 20hz square waves displayed in Craig's two pictures from his function generator are different. In the pictures I linked to in the other thread the AC coupling was even effecting the square waves at 100hz.

Shawn

[gerbache]

Square waves, even ones that are changing at a rate within the audio band, actually contain frequencies well above and below the audio band. Ever do a Fourier analysis on a square wave? I tend to agree that ac coupling the scope will possibly change the output by acting as a high pass filter with a coupling capacitor. Depending on your model of o-scope, it might not be a big difference at the frequencies being measured, but by its very nature ac coupling will apply filtering.

[mike stehr]

I'm brain-dead on this.

But with Fourier(4EA)analysis, you can break a square wave into sine wave components, with a fundamental frequency with period equal to that of the squarewave?

Something like that?

[gerbache]

The basic gist of Fourier analysis is that you can break any waveform up into a series of sine waves which, when added together, reproduces the original waveform. You can tell a lot about the frequencies in a waveform by simply looking at it. For instance, in a perfect square wave, the rise and drop between high and low points is a perfect vertical line, which means in terms of frequency that it's happening extremely fast. Also, between transitions, a square wave holds a constant, DC value, even if it's only briefly before it drops back down to the low value.

You can see some of this by looking at the responses of the amplifier in the first graphs (which look really good, I might add). When at the low frequencies, notice how the amplifier is dropping the value with a slight curve. This is because a certain portion of the low frequencies has been filtered out. The rapid change between high and low values is preserved, however, because this is a high frequency component of the wave. In other words, this amplifier cannot pass DC waveforms, but it can pass higher frequencies without problems. In contrast, I suspect that if you input a square wave into a solid state amplifier without an output transformer (or any other DC coupled amplifier with no output tranny), the low frequency waves would be reproduced correctly, because they are capable of passing signals all the way to DC.

Now lets move on down through the higher frequency waves. There is a period where the waves appear to be passed nearly perfectly, roughly around the 2kHz range. In reality, there is still some filtering going on, but it is occuring at points in the waveform where there is not much energy present, because the vast majority of the frequencies required to make this wave are able to be passed without distortion through the amp.

As we move on down to the higher frequencies, notice how the flat tops are being held basically flat while the rises and drops are curved (as an exponential curve, I might add, but this is getting a bit too detailed). This is because the extremely high frequencies required to switch between high and low values is above the frequency response of the amplifier, so they are being filtered out. In a sense, this is the exact opposite of the filtration occuring at the lower frequencies, in which the higher frequencies were preserved and the low frequencies were filtered. If you were to continue taking the square waves to higher frequencies (I know, it's above the "audio band," but the actual frequencies contained in these square waves are both above and below the audio band anyway, so it's fair to talk about them), you would continue to see the effects of this filtration, and the edges would continue to become rounder and rounder.

Hope this isn't too technical! I'm trying to make it all make sense, but I'm afraid I'm running on little sleep here, so if I need to try again, I'll give it another shot in the morning.

Oh, one more quick point; the sloped lines we see on the AC coupled scope images are exactly the same as the filtration occuring on the lower frequencies out of the amplifier. The AC coupling blocks all DC signals, so it is cutting out some of the signal. In this sense, the o-scope is distorting the signal somewhat, regardless of whether or not the amp is capable of reproducing it.

[Woodog]

I don't know about y'all, but I'm taking the pic, rotating it 90 deg counter-clockwise, framing it, and putting a title on it "Bearded Man at Dance". Other than that, I've not understood a thing about this thread except that If anything using AC coupling makes your work look *more* impressive.

[NOSValves]

This is like talking to bricks.

[sfogg]

Ryan,

Fine, that is your scope, mine is more or less the same way too with only a slight difference between the two at 20hz. The -3dB point on my scope is 5hz.

Now look above at the 20hz square wave from Craigs function generator in the first pic he posted (top trace) then look at the same thing in the later picture he posted. The difference between the two is AC vs DC coupling. Obviously his scope is distorting the 20hz square wave badly. That is from the filtering in it. Why it is different from yours I have no idea but it is.

This is Craigs scope at 100hz with AC coupling/filtering...

The top trace again is from his function generator and it is still being distorted by his scope.

Shawn

[sfogg]

gerbache,

Thanks for joining in the discussion. In case you hadn't seen it this discussion really started in this thread:

http://forums.klipsch.com/idealbb/view.asp?topicID=46020&forumID=68&catID=19&search=1&searchstring=&sessionID={F4919544-25E0-421B-9835-DAA72FE4D861}

" (I know, it's above the "audio band," but the actual frequencies contained in these square waves are both above and below the audio band anyway, so it's fair to talk about them)"

To expand on this a little for the others here everyone tends to talk about an amp with good square wave response as being 'fast' which is something of a misnomer.

What really is happening is where an amp has good square wave response that means it has very good frequency response to both sides of the frequency of the square wave. A sine wave at 20hz just has its 20hz component. A square wave at 20hz has components that go well below 20hz and also well above 20hz. That is why an amp may do sine waves easily at 20hz but fail miserably on square waves.

This is also why you simply will never find an amp with a -3dB point of 20kHz that does perfect square waves at 20kHz. It simply isn't possible. To pass good waves at 20kHz requires much greater FR/bandwidth in the amp.

So getting back to the point.... when you filter a signal into the scope and input a square wave near the filters frequency you distort the square wave because the filter is filtering out parts of the square wave.

Shawn

[sfogg]

Craig,

Are you capable of being involved in a discussion of things you don't understand without insulting those that are trying to help you learn something new?

Shawn

[BigBusa]

The sa100 is a sweet little amp. I had one for a while but ebayed it.

The only tube amp I ever really kick myself in the *** for selling was a fisher sa300. That amp was just awesome sounding.

[NOSValves]

Shawn,

Why do you keep hammering the same point home here ? I never said that using AC coupling didn't change the square wave. All I said was it changed it at both points the same and the difference presented is a true picture of what the amp does to the square wave. This entire discussion is pointless beyond that. I also agreed above that using DC coupling shows a clearer and easier to comprehend results. My entire point was as long as you show the input and output AC or DC coupled your not misleading anyone. Geeezz give it a rest.

Craig

[sfogg]

Craig,

"My entire point was as long as you show the input and output AC or DC coupled your not misleading anyone."

Your entire point is wrong.....

Shawn

[NOSValves]

gerbache,

Very interesting post ! May I ask what is your background to have such a keen understanding ?

Craig

[NOSValves]

Okay almighty Shawn I'm wrong are you happy now ?? Yes I can learn from people but they have to have some talent as a teacher and from what I can see you don't your ego gets in the way !!

You refuse to look at the proof I showed ! Your to busy thumping your chest. This thread was about the SA-100 not scope use 101 !!

Craig

[gerbache]

I'm a student in electrical engineering, and signal processing is sorta my thing. As it is right now, I know a whole lot more about the way the signals are being manipulated and all than I do about how to construct various circuits in an amplifier (especially with tubes, since we really don't ever talk about them anymore in school).

Nice thing about this is I can use the school's electronics lab equipment anytime I want to, and it includes some pretty top notch scopes and such. I haven't decided yet what they'd think about me dragging an old tube amp into the lab and wiring it up, though..might have to give it a shot sometime soon.

[NOSValves]

I bet they would love it !! This is really cool kind of the future meets the past

Craig

[sfogg]

Craig,

" You refuse to look at the proof I showed ! "

No I didn't, it confirmed what I had previously posted as I already explained to you and I suggest other tests to try.

"your ego gets in the way !!... Your to busy thumping your chest. "

I have posted explanations and corrections trying to help you understand this. You in turn have posted numerous insults to which I have ignored and not responded to in kind. There is someone here having a problem with their ego and thumping their chest but I disagree with you on who that is.

Shawn