Continued Learning on Oscilloscope.

[Ed_Chambley]

Can someone please direct me the documentation of Oscilloscope use.  The simple things I understand well, but I now have an oscilloscoe, Tektronics, 2 channel with a lot of added function and would like to increase my understanding of how to read frequency an harmonics.


What is some good reference material in this area?  The Tek Manual begins advanced since the model is on the more featured models.


Direction would be helpful.  I have run the gambit of youtube, not ai an looking for more in-depth documentation and I cannot find any on the NET.  Have no problem with purchasing or downloading.  Any and all suggestions are welcome.


Thanks,

[PRR]

> how to read frequency... cannot find any on the NET

Either you have been very unlucky, or hasty.

The 'scope is calibrated in Time. You ask Frequency. Seconds per Cycle, Cycles per Second.... they are the same thing only upside down.

The 'scope is calibrated in time per cm. One cm is the distance between major marks. Usually 10 marks across the width.

SO if the sweep is 1mS/cm, a wave that just fills the width of the 'scope is 10mS long. 10mS is 10/1000 Seconds. Or in frequency: 1000/10 Hz, or 100Hz.

You can verify this by putting your finger on the probe. This picks up all the hum (and buzz) in the room. Power frequency is well controlled on utility power systems. Knowing if you are in a 50Hz or 60Hz land you should be able to estimate the mS of a wave and thus the Frequency.

Alternate: get a calibrated frequency source. Put the 'scope in XY mode. Known and unknown freqs to the two inputs. You get a jumbled pattern. As the two frequencies approach "simple fractions" they make simple patterns, Lissajous .

http://tubebooks.org/technical_books_online.htm Test & Measurement

[eleventeen]

Your scope most likely has a little test point on it called "CAL" that you can clip your probe on to. That node is supposed to be a 1 Khz square wave signal of one volt. That's where you start. You have to integrate in your thinking the sweep setting your scope is set to and how many waves are displayed. One cycle of a 1 KHz square wave takes 1 millisecond....because....there are 1000 of them in one second, right?  Thus if your scope is set to 1 ms, then an up & down of a single cycle of a square wave should span one of the ten squares etched on the scope screen (aka the graticule). Being familiar with how you use the scope for measuring freq is getting familiar with the reciprocal math involved changing from "milliseconds (or microseconds) per division" *as displayed* to "cycles per second" *as calculated*. Do this with the CAL test point 50 times on 50 separate occasions and you'll ultimately arrive at a state where you can transform "how many divisions" at "which sweep rate" produces a number that is the frequency in Hz. It is no sin to have a cheapo calculator there if you want. See how the trace changes when you double the sweep rate (will display MORE cycles) and then double it again, and again. (this won't really be doubling, usually the sweep rates go 1 > 2 > 5 > 10 >20, but you get my drift. Then go the other way. The good exercise to try is to predict what you SHOULD see when you make a particular change in sweep rate.

If you think about it, you do this whenever you use an ordinary DVM. Almost always, you have an idea of what voltage you should see at a particular point BEFORE you touch your probe to it. You are measuring plates on a 12AX7 preamp, you expect to see 120-200 volts, for example. If you saw 67 volts (which is possible but not common) you would suspect something was wrong. If you saw 350 volts, you'd think "that's high". Even though there are all kinds of 350 volt points within a typical tube amp.


You can do this with a heater winding on a tranny, as well. Unlike the CAL node, you will require a ground connection. (On the CAL it is of course internal) It is better that way than using your finger to pick up 60 Hz floating in the air because it is much more stable. But, 60 Hz is not that ideal of a frequency to play with


You do the same thing in the vertical direction. If the V amp is set to 1 volt per division, the 1 volt square wave should be 1 division high. Don't forget that in most circumstances you are using a 10:1 probe.


You are NOT going to be able to "measure" the kinds of frequencies that come off a plucked guitar string in much of a satisfactory way, because the base (not bass) note is so dominant and the upper harmonics decay so rapidly. You can certainly measure the fundamental frequency of such a string. You can see upper harmonics, but they are so transient you won't really be "measuring" them.


In order to measure "frequency response" of a network (such as a tone control or tone stack) in a visually coherent way you really need an audio sweep generator. Otherwise you have to have a normal frequency generator and plot input vs output voltage as single points at various frequencies.


By the way.....you can use onlinetonegenerator.com for EITHER fixed or sweep tone generation, though I can't seem to get the sweep gen to endlessly repeat (which is what you want) But there is plenty to play around with using just the function generators there. 

[jjasilli]

The stuff on the scope display is relative, which can be very confusing.  Maybe the 1st thing to do is use the 'scope as a voltmeter.  Start with a known DC voltage.  Get it to display on the screen.  It will be a horizontal line.  It will display at different levels on the screen, but will always be true to the scale you select. 


Then graduate to a known AC voltage & frequency, say 6VAC @ 60 Hz from a filament tranny.  Get it to display on the screen.  It will be a sine wave.  Note how you can change its position and size by the settings you select.   Also note that the scope will show peak-to-peak voltage; AND the peak voltage will not be RMS, so it will be 1.414 X that shown on your voltmeter!  Then learn how to reconcile the known 60Hz frequency with what you what you see on the screen. 


Other places for insight are vintage articles and publicationS like the one suggested by PRR.  Also vintage scope manuals like from Heathkit, Eico, etc. 

[shooter]

I was able to go to Techtronic's site and get a copy of my operators manual, that I lost
I have forgot more about my scope than I wanna re-learn
They are indeed a great tool, get the basics down as 2nd nature, then the bells n whistles are fun to play with.
 

[Ed_Chambley]

Thanks for the input everyone, I will check out all that you guys wrote along with the documentation PRR posted.  I am not a novice using one and was taught a lot by an older friend who has passed.  Shooter, I have the manual as well.


Oh and I was not Hasty and did find a lot of info on scopes on the net, just not frequency related as it seems like this topic is often considered a known.


Let me do some testing and I will come back when I have additional questions.  I like the approach of using the heater string since they are known.  Get suggestions measuring known.


Thanks guys!

[Ed_Chambley]

I will do some reading first, but let me clarify more of what I am wanting to see.  I saw a video of a guy showing an amp that was Push Pull and had no driver tube and he was showing the harmonics in a sine wave with the wave decreasing.  It was an unusual cheap tube amp and he was showing the harmonics of that amp compared to an amp that has a driver tube.  In this instance the difference was huge.


I have to get the thinking so I will read a while.

[Ed_Chambley]

I found things like this:
https://www.youtube.com/watch?v=OQMnjaMSKn8








But what I am wanting to learn is where in the circuit to make measurements and also wanting not to trash the scope with voltage.  Other than taking a reading at the output, what can be learned by using the scope along a tube amp circuit?

[sluckey]

Which scope do you have?

[shooter]

what can be learned by using the scope along a tube amp circuit?

You can see each stage and how it acts on the signal, did it start clipping before the TS, what does "that" sound like?  Not so good, make adjustment, repeat.  You can see gremlins, slow oscillations, noise spikes.  Then you can make changes and see the effect.  IOW, you add one more sense, sight, with hearing.

I'm still trying to use my FFT mode to view in frequency, instead of time for the horizontal base. 

[Ed_Chambley]

Which scope do you have?

Steve, it is a TeK 2440.  Very nice and I got the Tek probes as well.  I also bought some less expensive ones as well.  I am simply looking for examples of things others do inside amps with the scope.  I know some, but would like to get more knowledge.

[sluckey]

I thought you were gonna get an analog scope. I've never seen a 2440 but I worked quite a bit with a 2446. Very nice.

You can use the cursors to measure frequency and have it displayed onscreen. Just select cursor function 1/TIME, move the cursors to bracket one complete cycle, and the screen will display the frequency in Hz.

[Ed_Chambley]

I thought you were gonna get an analog scope. I've never seen a 2440 but I worked quite a bit with a 2446. Very nice.

You can use the cursors to measure frequency and have it displayed onscreen. Just select cursor function 1/TIME, move the cursors to bracket one complete cycle, and the screen will display the frequency in Hz.

I had one. Japanese 2 channel and was very nice, but I got this one offered at a price you would gotten it.
Thanks for the tip.

Thanks to all. If anyone has anything else, please contribute.

[shooter]

I’m currently *tuning* my last build.  below are 3 scope captures, the first 2 are preout, clean and max, the 3rd is speakerout, max drive.
What they show that a meter won’t is the gross clipping, and distortion that comes from flogging a pair of PSE 6V6’s with 40 to 60Vrms!  A good meter would show me the Vrms, but NOT the *way* their distorted.  The scope also shows me a clean, (no un-wanted, un-expected weirdness), signal out of the pre-amp.  Now when I make changes I can be confident if *weirdness* happens, I caused it verses, it was there before.

[Ed_Chambley]

I’m currently *tuning* my last build.  below are 3 scope captures, the first 2 are preout, clean and max, the 3rd is speakerout, max drive.
What they show that a meter won’t is the gross clipping, and distortion that comes from flogging a pair of PSE 6V6’s with 40 to 60Vrms!  A good meter would show me the Vrms, but NOT the *way* their distorted.  The scope also shows me a clean, (no un-wanted, un-expected weirdness), signal out of the pre-amp.  Now when I make changes I can be confident if *weirdness* happens, I caused it verses, it was there before.

Yes, I appreciate you posting.  Lets say you were viewing a Sine wave and it were normal on top, but when it swings the lower portion makes a V instead of a U so what you have is a U over a V shape.  I am getting this from a 5879 preamp tube.  Any idea?

[shooter]

makes a V

Here's my way of *solving* unknowns like that;
decrease input signal, get better?
If yes, then the tube is going into "distortion", what kind or why, I'm clueless.
IOW, I try and *look* at what the signal passed through, an RC network, simply from grid to plate.  Swap the tube to eliminate that, then you're left with bias, circuit, etc.
 The other thing I'm finding out, when it does weird things, how does it sound?  Some of the weirdest waveforms seem to get Musicians all kinda happy.

[HotBluePlates]

...  Lets say you were viewing a Sine wave and it were normal on top, but when it swings the lower portion makes a V instead of a U so what you have is a U over a V shape.  I am getting this from a 5879 preamp tube.  Any idea?

The scope shows you what is happening, but it's on you to figure out "why". You may have to theorize a "why" and a way to manipulate that "why" then test it out and see on the scope (or by listening) if it had the desired effect.

The U on top and V on bottom sounds like it might be similar to Shooter's Max Sig screen shot, if you kept increasing signal amplitude. I'm gonna guess distortion, where you top U is a signal excursion that's being limited/distorted from reaching the same peak in the same shape as the bottom V. If you listened to the amp through a speaker with the same signal-level setup, I'm guessing you would hear the distortion as well.

So then the "why" part... Well it depends on where you were probing that stage. You saw the top of the waveform flattened a bit compared to the bottom... If you were probing the plate output, the positive swing of the signal is where the tube current is falling, which leaves less voltage drop across the plate load resistor and more voltage on the plate.

If you look at a set of curves for a pentode, you'll see the grid lines crowd together at low plate current, so there are smaller current changes for a given amount of grid voltage change. That sounds like distorting the cut-off side of the output wave. You might also conclude this is mostly even harmonic distortion (one side bent).

You might make a note of the input signal size to that stage, and compare it to the bias and the output signal size at the plate. Does the peak input exceed the value of the bias voltage? If yes, then you're running the grid positive, getting distortion which coincides with the negative peak at the plate. What is the ratio of output signal/input signal? That will be your voltage amplification (gain). If you're getting less gain than predicted for the stage, you might ask yourself, "have I biased this stage properly? Do I need to move the idle bias so one side doesn't go into distortion too early?" But then the tradeoff might be than you idle to a point where the top & bottom of the waveform distort, which is odd harmonic distortion (though even may also be present). That will sound different.

Or maybe you checked the measured stage gain and then noticed you're slamming the stage with a huge input signal, larger than that stage would typically see. Now you know you are observing a condition on the scope that wouldn't be as-severe in actual use.

Anyway, point being that the scope is just another tool to observe what the amp is doing, but it requires a good bit of work on the operator's part to translate what the scope display means. I don't know if there is a good place out there to learn this stuff in one shot. If there is, I'd like to know about it and learn something myself.

[Ed_Chambley]

I am simply probing the plate at V1 5879.  I also have a 5879 in V2 for a parallel and series.  It is even order harmonics until you run through both tubes.  Parallel I love all the different tones and textures and cleans are very nice.  If you push it hard using only one of the 5879 then the distortion is very nice.  In parallel it is even better since both tubes are biased much differently.  It is the series connection that disrupts the wave and makes it almost look like a triangle wave.  I have a little more tweaking here.


I have been working on this amp design fir a long time now.  The wave form does not change anywhere in the circuit, even at the output. 


I have gotten some idea of the visual in comparison.  I know this is gaining experience, but I need experience in both areas.  One is the use of the new scope and its features and secondly is understanding the wave forms and begin to understand what I see and relate it to what I am hearing.


Basically, I am simply compiling a list of different tests others use and know.   If I replicate the same I will learn the same.  It is that "translation" on the operators part I am working on.  That is why I titled the thread as continued learning on the oscilloscope.  I really know enough to use it on traditional circuits and was what I was taught.  Basically using one for troubleshooting a Blackface or any amp with a primarily clean the scope has made this a lot simpler.


Just wanting to learn more so I am reading some of the old lit PRR pointed me to along with the manual.  Also Tektonics website in addition to the older texts is most helpful.


HBP, you mentioned 2 areas on a pentode to change cap values, but most schematics have the screen tied to the cathode.  Can you direct me to a schematic which does not directly tie the screen direct to the cathode.  I am assuming there must be another cap in the works.

[sluckey]

Can you direct me to a schematic which does not directly tie the screen direct to the cathode.

Your AC-15 is one example. I've never seen a screen connected to the cathode. Maybe you mean suppressor grid (G3)?

You will probably never use more than about 1% of the capability of that scope while probing inside a guitar amp.

[Ed_Chambley]

Can you direct me to a schematic which does not directly tie the screen direct to the cathode.

Your AC-15 is one example. I've never seen a screen connected to the cathode. Maybe you mean suppressor grid (G3)?

You will probably never use more than about 1% of the capability of that scope while probing inside a guitar amp.

Thanks, Got it now.


I know the scope is more than necessary for a tube amp, but learning more about it will help me with my industrial equipment if I learn and that is why I started this thread.  There are a lot of us who can benefit from more understanding of oscilloscopes.


For instance, before my neighbor brought his over a few years ago and showed me how to troubleshoot a problem in a Princeton Reverb and we found the issue in about 5 minutes, whereas I had been using listening amp which I find not to be a great tool.  I had been chasing this problem for days only to find out the 100 ohm hearter balance resistors, one of which had drifted a lot.  Replaced the resistor and the repair was done.


This got my attention and right then I saw the tremendous value in having one.  After looking back at all the time I wasted not having one and you can get one like a 2 channel Iwatsu analog for less than $50 if you watch craigslist.  To top it off, I will bet I have changed enough good components chasing problems than I paid for the new one I got.  After selling the one I had, I have $75 in the Tektronix 2440 and one used probe sells for more than that.


Thanks for the AC15 reminder.  I know this one well and immediately what HBP had mentioned a while back.

[jjasilli]

I suggest the following hypothesis.  It's a given that the uneven curve shape is a form of distortion.  So: either the tube is bad; or the input signal itself is distorted; or it's a result of bias/loadline.  If the latter, the tube is probably biased "low" on the loadline, so that the top curve has plenty of room to be itself, but the bottom curve is getting tweaked.  The fact that the same distorted curve shape is everywhere in the amp, suggests that this is happening in the first gain stage (assuming a pure input signal).


EDIT: or there's something wrong with the scope, which is unlikely but easy to test.

[Ed_Chambley]

Can you direct me to a schematic which does not directly tie the screen direct to the cathode.

Your AC-15 is one example. I've never seen a screen connected to the cathode. Maybe you mean suppressor grid (G3)?

You will probably never use more than about 1% of the capability of that scope while probing inside a guitar amp.

Also,  simply want to learn more and guys like you know this stuff and we don't have a lot of learning discussion (fixing and building tube amps is the main discussion).  The forum has loads of other topic areas and I am beginning to work with Vfet transistors and other circuitry.  To me this stuff is amazing.  I still am amazed that from an antenna you can have music come from a speaker.

[Ed_Chambley]

I suggest the following hypothesis.  It's a given that the uneven curve shape is a form of distortion.  So: either the tube is bad; or the input signal itself is distorted; or it's a result of bias/loadline.  If the latter, the tube is probably biased "low" on the loadline, so that the top curve has plenty of room to be itself, but the bottom curve is getting tweaked.  The fact that the same distorted curve shape is everywhere in the amp, suggests that this is happening in the first gain stage (assuming a pure input signal).

You are correct! I have a decade box of caps and the cathode resistor is a 5k pot.  That is how I got the wonderful clean.  This is a 5879, 5879 to a 12Ax7, CF tone, 12AT7 Driver, 2 KT66.  It is getting the second 5879 to make a better overdriven tone.  I have a Gain control on the tube and I have gotten what I would call a nice punchy crunch, but if I push it the scope begins to look like devils horns.


Too much from one stage so I have to find out if I can get a tad more from the other stages without messing up the clean. I may try to figure out a way to morph it from a triode to pentode, but I hate doing anything to add overdrive that is not foot switchable.


And now I have had to take a break from it, but I have plenty of notes.  2 of my Tombstone Table Top radios have gotten very noisy and the wires were so brittle I am having to rework 2 rats nests.  So I thought I would take some time with the oscilloscope since rewiring a radio is sort of mindless work.


The scope self tests and checks out.  Also using it on my other amps actually show the same results as my previous scope.  I also still an old single channel scope I have kept to compare if I am concerned with the functionality of each.

[shooter]

if I push it the scope begins to look like devils horns

And that trace just might be *the sound* that makes a musician break out cash!  I'm an electronics guy, 38 years, when I started looking at "musical" waveforms I was horrified, I would been fired long ago had my equipment produced that, but this a whole different *system* with different normal so *devil horns* might be good.

[Ed_Chambley]

if I push it the scope begins to look like devils horns

And that trace just might be *the sound* that makes a musician break out cash!  I'm an electronics guy, 38 years, when I started looking at "musical" waveforms I was horrified, I would been fired long ago had my equipment produced that, but this a whole different *system* with different normal so *devil horns* might be good.

That is funny, but true.  The amps I play make a very nice wave, but distortion does have its place.  Like I mentioned getting to know looking at sounds like takes experience, but I am a technical sort of guy.  I prefer to read manuals over watching TV.  I have been able to keep my amps up for years, it wasn't until I did my first scratch build on a 5f6a and added reverb.  Then I was hooked.  That same amp today has another gain stage too.

[shooter]

I prefer to read manuals over watching TV.

Yup!

[shooter]

Ed, you might find this interesting…or not.

Still gathering data on my last build, I’m on the plate of the 6V6.  I rotated the bass pot and about #7 the oddity showed up, (image 1), then by #9 on the dial, it went away.  I zoomed in on the peak, (image 2) and stretched the horizontal from 250uS to 1uS to better see the oddity, (image 3). 

The problem is consistent, so I scoped the out of the TS, clean, into 6V6’s clean, swapped 6V6’s same, used a different probe, walla, gone!  No soldering required, but I’m bummed that my 100x tectonics probe is bad!

[Ed_Chambley]

Ed, you might find this interesting…or not.

Still gathering data on my last build, I’m on the plate of the 6V6.  I rotated the bass pot and about #7 the oddity showed up, (image 1), then by #9 on the dial, it went away.  I zoomed in on the peak, (image 2) and stretched the horizontal from 250uS to 1uS to better see the oddity, (image 3). 

The problem is consistent, so I scoped the out of the TS, clean, into 6V6’s clean, swapped 6V6’s same, used a different probe, walla, gone!  No soldering required, but I’m bummed that my 100x tectonics probe is bad!

Sure I am interested.  Thanks, this is why I started the topic.  Just smile discussions on wha you see verses what is heard. In this case, it does suck that a Tek probe has gone south.

[shooter]

This is the 3rd amp I've taken the time to dig into the waves n how they sound, this is the 1st amp that sounds completely great, with no distortion, and completely NON-musical, with anything above 10% distortion.  The odd thing, the distorted waveforms aren't much different than the others, just guessing it's *in the tubes* and iron, both others were X7's EL84 and KT88 with heavy iron, this is U7's and 6V6 with light iron.
 I'm gonna tweak out some drive and call it good, since it actually meets my original *design* intent as a bedroom MP3 amp that you can plug in and play-along with.

Tek probe has gone south.

Yup!, I played with it some, and it would actually cause the PA tubes to break into bad oscillations and make horrid sounds!, I tested my homebrew 100X and it's good enough for now.

[MakerDP]

Ed, if you are still looking for some valuable scope testing info, I have found this to be very useful....

http://www.r-type.org/articles/art-125.htm

It is very short but it shows:

o Where  to test in a sample circuit.
o Sample waveforms and - most importantly - WHAT causes them to look like that!
o The benefits of square-wave testing. You can tell a LOT about an amp's signal by looking at what it does to a square wave.

Don't let its brevity fool you into thinking it is somehow incomplete. It's chock-full of great stuff.

I agree with you about YouTube. If someone made a series of videos on using a scope to test amp circuits it would generate a LOT of interest on the various tube amp forums. Maybe I'll put a bug in Uncle Doug's ear about it...