Learning to use Oscilloscope

[Fresh_Start]

I'm stumbling along the learning curve with my collection of test equipment.  Got a decent signal generator and a dual trace Tecktronix oscilloscope.  Using a t-connector, I've got the signal generator going to the input jack and to CH 1 of the 'scope.  I'm using the probe to connect to CH 2, starting at the grid of the second triode and checking subsequent grids in the main signal path.

Anybody know offhand what a typical single coil output is in terms of voltage swing?  How about vintage and hot humbuckers?

So here are some things I've observed so far.  First, the Treble and Bass controls seem to induce a certain amount of phase shift for a portion of their range.  Second, for some frequencies, voltage swing maxes out at about "8" on Volume.  Going to "10" just makes the wave tops/bottoms flatter.  Third, the onset of distortion is different in terms of the Volume setting for different frequencies, and the shape of the distorted wave can differ as well.

I was sampling at 80Hz, 300Hz, 1KHz, and 2KHz IIRC.  Is that a reasonable sample set?  

From what I've read, it seems like some guys just look at a single frequency like 1KHz, but if my observations are typical I'd think we should look at a spectrum of frequencies.  It's kind of fun (at least for me) to play with the tone controls too.  

Just for reference, the low E string is 82Hz, high E string is 330Hz, and 21st fret on high E is about 1.1Khz.

Cheers,

Chip

[eleventeen]

For you to see the effect of (the rotation of) a tone control, you'd have to feed the scope a *range* of frequencies...as you are suspecting. You won't see much just diddling the knob of a fixed-freqeuncy audio generator. Doing so (eg; feeding the amp a swept range of frequencies) generates what's known as an "envelope" and it is from that type of scope display you can make some more meaningful/useful observations.

---------------------
                           this would be flat response

---------------------


-----------\
               --------\  this would be treble cut or bass boost
               --------/
-----------/


-----\             /--------
       -----------             this would be a "notch" cut---more typical of  graphic EQ than an amp ctrl

       -----------
-----/             \---------


This is why, to align RF sections of radios & transmitters, a "sweep" generator is used. Because the performance at any one freq is not especially useful. You'd need some type of sweep to see the freq response of a particular section of an amp. Unless, of course, you were willing to make, log, and graph painstaking measurements at different frequencies. Your audio osc got a "VCO input"? Some will do that.

[eleventeen]

By the way, if you google "free audio sweep generator" you'll find a couple of sites that will let you download various tone generators of fairly surprising complexity for free. One I found from NCH software will do an audio sweep! You'll have to diddle with it for a while but it's there!

[jjasilli]

Typical single coils put out about 200mV.  Humbuckers can sometimes hit up to 1V.  Note this is Meter measuremnet; i.e. RMS.  The scope measures peak-to-peak.  You need to know that peak = 1/2 X peak-to-peak; and RMS = the inverse of the square root of two X peak = peak X 1/√2.  peak-to-peak; peak and RMS are three different (but related) things, which you must keep mentally separate and accounted for.  Otherwise your scope and your meter will always seem to be in disagreement.

There are a couple of threads below in this topic where PRR, sluckie & tubesornothing help-out a couple of scope newbies includiung me.

[Fresh_Start]

Typical single coils put out about 200mV.  Humbuckers can sometimes hit up to 1V.  Note this is Meter measuremnet; i.e. RMS.  The scope measures peak-to-peak.  You need to know that peak = 1/2 X peak-to-peak; and RMS = the inverse of the square root of two X peak = peak X 1/√2.  peak-to-peak; peak and RMS are three different (but related) things, which you must keep mentally separate and accounted for.  Otherwise your scope and your meter will always seem to be in disagreement.

There are a couple of threads below in this topic where PRR, sluckie & tubesornothing help-out a couple of scope newbies includiung me.

Thanks JJ.  RMS is good info since I can measure the output from my signal generator with the DMM.  Of course, I really should've measured my guitar's output myself...

Cheers,

Chip

[PRR]

A guitar amp on "10" should makes its full 50 (whatever) Watts with ~~~20mV input. Since you want all the Watts you paid for, and more, even when you pick lightly, this tests sensitivity. (Mondo-fuzz amps may only need 1mV; polite Hawaiian/swing amp might not overload until 50mV.)

OTOH, at lower volume settings it should take 500mV "clean", so you can play with a heavy hand and not overload in the first stage. You need to check both ends of the input range.

All EQ/tone gives phase shift. You can compute phase from frequency slope. Don't get hung-up on that; it always happens. (Discounting some extreme contraptions not found in simple gitar amps.)

Continuous sweep is not too very important. A guitar amp is not complicated enough to have narrow peaks. It can have a narrow "infinite dip" in the tone control, but it is what it is and it isn't important to follow it "all the way down".

Test out to 5KHz-6KHz.... there's important "zing" overtones to there. Test past 20KHz if possible; some amps get nuts and that can affect transient tone.

You have a dummy load, of course?

[Fresh_Start]

Yes, this dummy does have a load

Sorry, just couldn't resist that response once it popped into my head. 

I'm not interested in continuous sweep for purposes of tone stack experiments.  It was very interesting that distortion showed up at different levels and with different wave form shapes for various frequencies.  Obviously, I did not go up high enough in frequencies.  I'll go back and look at the 5-6KHz range too. 

EQ generating phase shift made sense after I thought about it... just hadn't thought about it before.

Never would have occurred to me to look way up at 20KHz.  I will though!

Thanks for your thoughts,

Chip

[simonallaway]

I think it'd be interesting to use the techniques you've already described to measure/plot the frequency response of a given amp, and then use that as a reference to measure the subsequent change when using a speaker/cabinet.

You'd need a decent microphone I suppose (or at least account for its own response).

Even if the measure was based on just 10 frequencies, 100Hz apart, you'd be able to say with some degree of certainty that "speaker A has 3dB more in the mids that speaker B". And of course how this response changes depending on how hard you're pushing the rig.

[eleventeen]

The problem with making those types of measurements is the sheer number of things that influence the results. All of which you kind of have to remove or otherwise account for if your goal is make an accurate measurement. And in the end, it's what you like more than how low your THD is or how accurate you've measured it anyway!

First, you start with the anechoic room....then the $2,000 recording-grade microphone and the $700 lab-grade mic preamp...and on and on.

[Fresh_Start]

Thanks guys, but I'm at the top of the learning curve and it's a l-o-n-g way down!

Chip

[PRR]

> measure was based on just 10 frequencies, 100Hz apart, you'd be able to say with some degree of certainty that "speaker A has 3dB more in the mids that speaker B".

No. you can't.

That was one of my first attempts at amp measurement... the curve was VERY ragged and more ragged the more measurements I took.

Room response will vary +/-10dB from 1KHz to 1.01KHz.

And a 'scope isn't much use.

Two ways to measure speakers in rooms:

1) continuous sweep with averaging. This was the standard when we had knob-operated signal genertors and mechanical pen-plotters. The mike output is rectified and then low-passed to get a running average as the frequency changes.

2) band-passed hiss. You can buy Test CDs with octave and third-octave bands of hiss. You can generate such signals on DAW. Just hold a standard sound level meter at some reasonable angle and distance and note the average reading for each band.

> how this response changes depending on how hard you're pushing the rig.

Audio systems "should" be linear. Response the same at any level.

This is nearly true.

In "clean" PA systems working at very high level, you can measure bass resonance change (damping) after some seconds at high power. This is "bad" so they try to keep it "small", adding more woofers to reach the desired level without so much heat in the coils.

Guitar speakers are often worked HARD. Indeed the bass may be reduced (and the midrange be garbled) when the coil is driven beyond the magnet gap. Since this "problem" can be cured easily, we must assume we "like it that way". Since the amount of non-linear depends what you drive it with, and the musical "sound" of the nonlinearity is quite subjective, there may be no useful way to measure it. We are left with descriptions such as you see on guitar speaker sales-sheets.

[jjasilli]

This should feel at home on your bench:  http://cgi.ebay.com/Eico-model-369-Sweep-Generator-/280566508379?pt=BI_Cellular_Optical_Television_Test_Equipment&hash=item4153112b5b  Maybe it even works!   

[eleventeen]

Naaah, that's way up there in frequency, RF-FM.

PC-based audio sweep generators can be had for free, as I posted above.

[Fresh_Start]

Thanks guys.  I'll worry about a sweep generator after I've learned to walk a bit.  Right now I'm still crawling

Besides, I've got one amp to fix up and another to build.  No time for more anti-Q test equipment!

Chip

[simonallaway]

Room response will vary +/-10dB from 1KHz to 1.01KHz.
And a 'scope isn't much use.

I think I had a picture in my mind of seeing how a sine wave looked at different frequencies. i.e. are distortion characteristics the same throughout a given range. I fully understand that I don't know what I am talking about.

2) band-passed hiss. You can buy Test CDs with octave and third-octave bands of hiss. You can generate such signals on DAW. Just hold a standard sound level meter at some reasonable angle and distance and note the average reading for each band.

I'd forgotten about that technique. I recall by band's sound guy having some "magical" device that would produce such noises and would spend 10 minutes or so measuring the response of whatever room we were in. It helped with resonances, but filling a bar with noise at 100db didn't go down well with the customers. I have an old radio shack spl meter...perhaps I'll try it just to see what happens.

We are left with descriptions such as you see on guitar speaker sales-sheets.

Exactly...which I hate, because they mean nothing. "British sounding"? what does that mean? I'm British and I don't know 

[Fresh_Start]

Exactly...which I hate, because they mean nothing. "British sounding"? what does that mean? I'm British and I don't know

WE know what "British sounding" means on this side of the pond  but I don't want to be rude...

Just kidding.  I love listening to different accents.  One of the funniest things is to listen to an American who's been in England for a couple of months... we tend to start mimicing what we hear every day but it comes out sounding sillier than John Cleese in a comedy sketch! 

Marketing descriptions of speakers are worse than useless.  I once bought a Weber California for a specific application based on the product description.  Let's just say it was not a success and the damned thing is still sitting in its box 7 or 8 years later.

Cheers,

Chip

[simonallaway]

Just kidding.  I love listening to different accents.  One of the funniest things is to listen to an American who's been in England for a couple of months... we tend to start mimicing what we hear every day but it comes out sounding sillier than John Cleese in a comedy sketch! 

That happened when I took my kids back to England for a bit (born and raised here). They didn't even know they were doing it!

Slightly on-topic, but speaking of accents...when I first needed to buy solder in this country, I went into a local parts place. I asked for solder but was met with blank stares. So I went and found it on the rack and pointed at it. The salesperson said "Oh you mean solder?". But to my ears he was pronouncing it without the "L"...a kind of "soh-der". Really weird. 11 years on and I still cannot pronounce it the american way. 

[PRR]

> we tend to start mimicing what we hear every day but it comes out sounding sillier

Ah'm at that awkward stage. Born in Missouri, too much time in New Jersey. Now in Maine I'm starting to say "loom" (stuff you put on your rocks to try to get grass) and "hahtop" (tar/stone mix).

It's going to take years to downshift from talking NJ/NYC speed to ME/MO speed.

> The salesperson said "Oh you mean solder?"

Around Philly we used the weak "l". But somebody from Georgia or Alabama came in asking for "sodder" or "sodder wore". The saleshelp were blank. My Missouri birth helped me know he wanted a lead/tin wire.

> a Weber California

"California" is code for "JBL flavor". Excellent for some things, just annoying for other things. That's for true E120/E130s, and even when JBL made them the high-strung mechanism didn't always come out the same (one of my four same-date E130s was "different"); with Weber's smaller production the assembly variations may be even larger.

[Fresh_Start]

Off topic, but I can't help it.  In the south, folks say "ink pen".  It took me a couple of years to figure out why this was so.  It seemed redundant to me until I realized that "pen", as in a writing utensil, and "pin", as in pointy piece of wire for sewing, etc., are pronounced identically.

Good luck with the Maine accent.

Chip