# Can I do this with an oscilloscope?

1. Dec 24, 2016

### Planobilly

Hi,

I am frequently concerned with audio "noise" coming from a guitar amp. I can inject a 500hz signal for example and obviously see that signal at the output transformer. There are other frequencies also being produced by the amp at the same time which are typically hard to see or not displayed at all. For example low amplitude 120Hz.

Is it possible to build some sort of filter that would filter out all frequencies greater the say 150Hz? The signal generator would be driving the amp at 500hz but I don't want to see that signal on the scope.

I am wanting to look for 60Hz, 120Hz signals and measure their amplitude. I also want to look for frequencies greater than than 20KHz.

Thanks,

Billy

2. Dec 24, 2016

### jim hardy

That'd be a real sharp cutoff low pass filter.

These will do a fine job

but if you're always after say 500 hz, i'd say you can build a
with Q 50 or so, put it in a plastic box with BNC's and a 9v batery, call it your o'scope accessory.
I used the one in fig 52 here for a very narrow bandpass , worked quite well . Cascading them gives sharper cutoff.

Last edited: Dec 24, 2016
3. Dec 24, 2016

### Averagesupernova

You could play with some of the available software that shows a spectral display on a PC. Just feed something into the mic input. Be careful here, you want to make sure to keep high voltage out of your PC.
-
If you want to troubleshoot noise generated by the amplifier you would just short circuit the input and go from there. I see no reason for a filter if you can utilize a spectrum analyzer.

4. Dec 24, 2016

### Planobilly

LOL Jim...you are going to pull me into transistor electronics yet!!

Seriously though, I assume the concept at least is logical. Build a device of this nature would be fun. The one on ebay is not too far out of my budget.

Thanks,

Billy

5. Dec 24, 2016

### Planobilly

Hi Averagesupernova,

I actually have a recording program ( Sonar Professional) on my main computer system up stairs that is chock full of spectrum analyzers. Plus many types of filters. I had assumed that it might prove useful. I have not tried it yet because of the issues of moving the amps from the workbench down stairs to the computer system upstairs which is pretty large and complex and can not be easily moved.

I also had some concern about introducing ambient noise into the microphone. I don't really think that will be an issue that I can not overcome as I have a lot of ways to control the microphone.

I had in mind to run a microphone cable down stairs as I can remotely control the computer from this computer on my work bench. Too many projects..too little time...lol

Thanks,

Billy

6. Dec 24, 2016

### tech99

Why do you not set the reference level using 500Hz tone and then switch it off and measure the noise?

7. Dec 24, 2016

### Planobilly

Hi tech99,

If I understand what I think I know, there is the noise that is being generated at quiescent conditions and other noise being generated as a function of amplification. The injection of some frequency like 500 Hz which is arbitrary to begin with, 1000Hz being a more common reference frequency is used to drive the amp. As the amplitude of the reference frequency is also adjustable other conditions that may exist at quiescent but are not noticeable can only be seen or heard with signal input.

The whole idea is to identify frequencies that should not be there and define where they are coming from. Listening to a amp, one may be able to hear an unwanted low frequency of 120 Hz but it may not be being produced by the usual places it normally comes from such as the power supply. Faulty tubes, caps, and many other conditions produce low frequency noise. These faults may or may not be apparent at quiescent conditions.

I guess the simple answer to your question is that when you turn off the 500hz signal you are no longer causing the tubes to amplify.

Cheers,

Billy

8. Dec 24, 2016

### tech99

Hi Billy
The tubes are amplifying even with no drive, so most noise sources will be visible. It is possible however, though unusual, for something to arise only during drive conditions, for instance a Class B amplifier might become unstable. The other source of "noise" is if the amplifier generates harmonics of the 500Hz tone. This will be a consequence of non linearity and you should be able to see it by looking at the shape of the 500Hz at the output, where it should be a perfect sine wave. Any instability of the amplifier should also be visible on the 'scope.
A technique for measuring total noise and harmonic distortion utilises a notch filter tuned to 500Hz at the output. This notch filter is traditionally a Wien CR bridge. You can check the measurement system by using the filter directly on the signal generator without the amplifier in circuit.

9. Dec 24, 2016

### Planobilly

Yes I understand what you said. The conditions I am speaking about are sort of unusual.

Well, I have been setting around for the last two hours playing the new amp and every now and then it has started to make a small almost inaudible static sort of sound only on certain notes I play which it does not make when it has no signal. If I am lucky, I will get it to reproduce this fault hooked up to the scope. This will most likely be easy to see because it will look like a voltage spike of some sort. Finding where it is coming from may or may not be easy. This situation could be caused by several things actually. The guitar cable or guitar. May also be a cap going bad somewhere.

Also I have FFT on my scope to look a harmonics. It is not a super good feature on the Owan scope I have but it is what I have to live with at the moment.

Thanks,

Billy

10. Dec 24, 2016

### Svein

Sounds more like an instability in the amplifier - possibly a periodic high frequency oscillation.

11. Dec 24, 2016

### dlgoff

My thought is that those notes are being clipped somewhere. From Wikipedia Clipping (signal processing):

12. Dec 24, 2016

### Planobilly

First thing I did was re-seat the tubes. That seem to fix the issue or at least it has not returned yet. I doubt anyone but me would have even noticed the sound. It happened at random intervals.

There could be high frequency oscillations going on, that would not be unusual with these old amp circuit designs. There are some ways to make them more stable but generally at the expense of good tone. This could also be environmental EMI. I do have some noise from a modem I am aware of.

Not much to do until it comes back, if it does.

Cheers,

Billy

13. Dec 25, 2016

### UsableThought

A few comments about the comments so far - FYI, my perspective by the way isn't a physics/EE perspective, nor even a "really good amp tech" perspective. However, as a hobbyist, in the past 2 years I've done a lot of work chasing down & eliminating various sorts of noise in tube guitar amps (and a bit in SS), and I've done a huge amp of reading on this as well. I also frequent forums elsewhere where the perspective is exclusively to do with building, repairing, or modifying guitar amps - both SS and tube, but most of the time tube. I've previously mentioned a couple of these forums to @Planobilly & recommended visiting them; they can be helpful if you are able to separate out the 95 percent witless comments from the 5 percent really good EE/tech comments.

Anyway - starting with the question of whether a scope or something else is best used for identifying & diagnosing problems with more or less continuous noise (e.g. 120Hz ripple, 60Hz hum, etc.):
I agree with @Averagesupernova that separating out different frequencies is a lot easier using an FFT app that can draw a spectrum. A scope is far more difficult for this but can be used. You can avoid frying the computer soundcard by 1) plugging the app into a portable USB sound card with a line level input (better than mic for this purpose), and 2) building a variable voltage attenuator (with a diode clamp too) to place after whatever probe you use & before the USB sound card; I can point to a good recipe if anyone is interested. OR, you can just be really careful; but the attenuator is more useful & safer.

Also I say "probe" because you would want to use this device in conjunction w/ a probe you build w/ a DC blocking cap & ground clip, similar to a scope probe but much simpler; this allows you to directly check nodes in the amp. This typically excludes the power amp due to the very large voltage swings on the plates there (although you can build yourself a 100-to-1 attenuator if you really wish for that); but it includes the preamp signal path right up through the phase inverter or driver. Also you can carefully look for AC (ripple and/or audible noise resulting from diode ringing) on the rectifier output. Thus you can often find exactly where a noise is entering the signal path. It's a great way to speed up diagnosis & supplement the old "divide and conquer" routines. The late & wonderfully generous Bill Machrone, former editor in chief of PC Magazine, for many years had a small business selling mod kits for a specific kind of Fender amp, the Blues Junior; his site has a neat page showing how to build such a probe: Troubleshooting Your Amp

Also Billy, no need to try & move your big computer; just get a laptop, tablet, or other small device; they make the appropriate apps for these devices at between $10 and$20 or so.

However . . .
. . . I wouldn't chase after the noise w/ an FFT setup quite yet, because I agree w/ @Svein that since the noise is associated w/ playing a note, it is not your typical mains hum/rectified mains buzz/diode ringing sort of situation, but most likely some kind of instability. There are many sorts of short-lived oscillations that can occur that will cause what are called "ghost notes." Billy, since you seated the tubes & the noise seems to have gone away, you've tried the #1 thing that typically gets recommended on the guitar forums just to start out with. Microphonic tubes; poorly seated tubes; tube sockets that need tightening and/or cleaning; these can all sometimes cause a small instability.

BTW, @dlgoff mentioned clipping; this noise that Billy was hearing was almost certainly not clipping; clipping is very different & is not experienced as a sound like "static", nor as separate from the originally played note. Also the high frequencies associated with clipping vary depending on the type of clipping and how hard it is, but generally speaking clipping is experienced as fairly musical in nature. Anyway that's a whole other topic.

Actually, although tracking down oscillation can be difficult - the causes are potentially quite numerous - getting rid of it need not be at expense of tone. Some tube guitar amps by design may live near the edge of instability, but they don't actually have to be unstable to sound good. Instability generally occurs w/ older amps and is introduced by either aging components, or mistakes made with either component values or layout during repair or modification. Poor layout or lead dress can lead to unintentional positive feedback; a filter cap going bad can cause problems due to increased leakage; apparently coupling caps can cause issues; etc. Once identified such issues can be corrected w/ no alteration to the amp's originally intended behavior. FYI, here's a nice page someone put together on sources of instability: http://www.ozvalveamps.org/stability.htm

Also regarding EMI, low-frequency EMI would be constant & not just when notes are played; however if it's a really old amp, it may not have grid stopper resistors on the input triode(s) and thus might be admitting RF frequencies. If I remember computer modems do have some RF, but the operation that causes interference (switching of some sort?) is apparently intermittent & unlikely to cause instability. Generally people seem to diagnose modem interference in their homes as an intermittent clicking sound down in the audio spectrum.

BTW if you ever do want to add grid stoppers, so as to avoid noise from cell phone calls, AM radio stations, and other sources of RF, these can be added quite easily (put them right up next to the input pin on the socket) & again won't hurt tone a bit. In theory a grid stopper can add a small amount of resistor hiss for when the amp is cranked up, but today's metal foil resistors are very quiet; plus if it's a really high gain preamp & you still are concerned about hiss, you can take the further step of reducing the value of the grid stopper down to say 10K by adding a very small cap, say 300pF, after the grid stopper & running its other end to ground; this improves filtering even w/ the smaller 10K resistor such that once again you are cutting off all frequencies above say 20KHz (obviously you can do out the math if need be).

Anyway glad the issue seems to have stopped w/ the reseating of tubes. Hopefully it won't come back any time soon. From that page I linked to above about causes of instability:

Last edited: Dec 25, 2016
14. Dec 25, 2016

### Planobilly

Hi Usable Thoughts,

Thank you for taking the time to post all your comments. I will take the time to study all you said in detail.
There are some functions on my scope I don't know how to use. I started to make some photos but there is so many screens I need to make a video so anyone could make sense of it. I will post it when I get it made.

I turned the amp on this morning and the intermittent crackling noise came back. It occurs on the low E string and A string. It is the typical sort of noise I have heard many times associated with issues with tube sockets. The sort of noise that is caused by arcing. As this is a new amp built with the best components I could find so some of the typical issues with old amps do not apply. Also the AB763 Fender circuit has been used in thousands of amps in one form or another and the Deluxe Reverb is not known for instability.

After I finished the amp there was a problem with the reverb recovery circuit causing hum and unusual noises. I measured the value of the components and found nothing wrong. I finally got tired of dealing with the issue and replaced the plate resistor, cathode resistor, cathode cap and the .0033 coupling cap. That solved the issue and the amp started working without issue. I don't know which of those components had actually failed or why they failed. Bad component? Overheated them when I soldered them in? Who knows. This problem is part of what prompted me to ask questions about troubleshooting noise.

It is entirely possible that I got a bad batch of caps. They were all best quality brands, Sprague Atom, Nichicon, Cornell Dubilier 715/PS Film capacitors.

So...the issue is more or less constant now and I hope I can reproduce the fault with signal injection which is a lot easier to deal with than playing the guitar and looking at the scope. I don't think this is circuit design instability. If I am lucky it will be a failing component. Worst case is a bad solder joint or something unusual like a bad stand by switch that is arcing. If the fault is in fact arcing it can be hard to find.

If I can find what stage or tube is associated with the issue the problem can be fixed even if I have to take the shotgun approach and just replace the components associated with that stage.

I am going to take the chassis out now and start looking. I will post the results. I really want to get this issue fixed so I can get back to my original questions about using scopes and other methods for tracking down noise in general.

Best wishes to everyone for the coming New Year,

Billy

15. Dec 26, 2016

### dlgoff

You are correct. I just checked out this page where there's an example of Peak Clipping.

16. Dec 26, 2016

### Planobilly

Here is a video of the issue.

I have made a lot of changes and tried a lot of things with no results. I really have no idea at this point what could be causing this issue. I guess it is possible the amp had done this from the beginning. It is hard to imagine I would not have noticed it after I resolved the hum issue in the reverb recovery circuit. At that point the amp sounded great as I tried the two new different type of speakers I was testing.

Perplexing to say the least.

Billy

17. Dec 26, 2016

### jim hardy

Bill this is speculation

try "divide and conquer". Amp or speaker ?

i'm not sure whether i heard the buzz on my cheap headphones
but i see a small speaker trying to play bass notes at high volume

as you know
distance= rate X time

to make a 100 hz bass note a speaker cone must move forward for 1/200 second and backward for 1/200 second.
How far does it move during that 5 milliseconds ? Depends on how loud it's asked to be.

Small speakers have limits to travel.
Outward limit is where the voice coil starts to exit the magnet area, or the surround gets stretched so tight it stops motion.
Inward limit is where the voice coil former , usually a thin plastic or aluminum cylinder, hits the bottom of the magnet structure. That makes a loud 'clack' and a series of 'clacks' is a 'buzz'.
I dont know what hitting the "out" limit would sound like, i'd think it might not even be noticed.. But the "In" limit is a physical collision.

To isolate if that's what you are hearing , try a bigger speaker, or close the back of that amp making it a sealed enclosure that'll limit cone travel naturally.

Good stereos have a LF rolloff button that prevents subsonic frequencies from reaching the speaker, that's in case somebody is using vented enclosure with a high powered amp that's capable of driving the speaker beyond its limits. It'll also keep your speaker cone from trying to follow a warped LP record..
Sealed encosures limit cone travel because they compress the air behind the cone.
Vented ones don't so you have to be aware of your amp's capability .
Guitar amps i've seen have open backs. So there's no inherent limiting of cone travel. If you "souped up" that amp it might need a speaker with 'longer throw' .

What's low end response of that little amplifier? Do you think it's capable of overdriving your speaker? Gently place a finger on the cone while playing that note and see how much motion you feel...

one step at a time.

old jim

Last edited: Dec 26, 2016
18. Dec 26, 2016

### Averagesupernova

Interesting. I never thought of that. A former coworker and I used to discuss whether the subsonic filter did anything at all. In that case it is likely it could.

19. Dec 26, 2016

### Planobilly

Hi Jim,

Sorry the video did not "show" much. It was meant just to let you hear the sound.

I have played the amp though a quad Marshall cabinet with four 12" Celestion with the same results. The little Orange amp is a small transistor amp I use to hear the results of a probe placed at different points in the circuit. The speaker you hear in the video is a 12" Weber DT12 30 watt guitar speaker.

The issue is definitely in the amp. The greater question is about what the amp should normally sound like. Is this buzzing normal for this design? I don't think so but I have not found any defective component as of yet.

I also plugged the amp into a another Weber 12" 65 watt speaker designed with a Neodymium magnet. The speaker is more efficient and has a different bass response. That speaker reduced the buzz by a small amount.

The buzzing noise happens on both channels. It happens when channel one is disconnected and the reverb stage and tremolo stage are disconnected.

That only leaves V2A and B, V4B, and the phase inverter and of course the power stage. I went back and checked all the voltages and ripple values plus the bias on the output stage tubes. I replaced coupling caps and cathode caps and resistors on V2 A and B. I measured the resistor values on the rest of the amp. All this without finding any issues. No change.

If I have too I guess I can start changing coupling cap values to change frequency response and reduce the capacitance of the cathode cap. All that is a bit over my head to calculate but I can figure it out if need be.

Jeezs...I just want to build amps, not worry about being in the corner with any frequency ...lol I know, I know, silly joke

Cheers,

Billy

EDIT: The amp is around 20 watts, and yes it would be capable of overdriving a small speaker. The whole video is with the amp volume turned up less the 50%.

20. Dec 26, 2016

### jim hardy

Okay, use your scope's FFT spectrum analyzer to find the frequency of that string.

Look for difference in the harmonics in amp output at low and high volume. Distortion should appear as bigger harmonic content

Use your signal generator to feed that frequency into the amp continuous so you can look with scope for distortion stage by stage, examining both the raw signal and its spectral content via FFT .

Sometimes a triangle wave will make distortion more obvious to the eye.

Remember when Svein explained what is "Transient Intermodulation Distortion", overdriving intermediate stages within a closed loop ? I'd learned that phenomenon by another name, so was mistaken about what it meant.
Could be that's going on, is there a schematic of this little guy someplace ?