Audio Generator or Function Generator for amp testing & troubleshooting?

[Enzo]

COnsider where you have the scope trigger set. If your waveform keeps swapping phase, perhaps the trigger is set too close to zero. Try turning the trigger control. You don;t need an extrenal trigger, your scope triggers off the waveform perfectly well.

Digital and analog scopes can both have unstable displays, and they can both be rock steady. It is more about what they are connected to and how they are adjusted.

Plug your phone test signal into an amplifier and listen to it. Does the tone sound rock steady? Or can you hear the occasional shift or glitch? I don;t have such a phone, but I could belueve that while the phone is doing this app, it might also be doing other tasks, and those interruptions might cause little changes. AFter all, the app is not lab gear.

[MarkF786]

Enzo,

Thanks for the tip. Though I had previously tried adjusting my trigger above zero without success, your post led me to look at the problem again. What I found is that if I set the trigger coupling to HF reject (which attenuates all signals with a frequency higher than 150 kHz, though I don't know why I'd be seeing frequencies that high from an iPhone headphone output) it increased the stability, then as I dropped the voltage, I also adjusted the trigger sensitivity. Now I can get down to around 50mV rms and still keep a stable waveform (whereas before, I needed around 500mV before my scope went haywire). Still, I see the built-in frequency counter jumping around in value, while if I use the built-in probe test connection on the scope. or my analog synth, a Moog Voyager. the frequency stays rock solid.

I realize I could use my Moog as an audio generator - it's square wave isn't perfect but it's better than the iPhone app, and I can get a sine wave by using it's LPF to remove all the upper harmonic from any of the waveforms - but it's not very portable.

Mark

[Enzo]

As I suggested, the phone is not an audio generator, the phone is a computer doing that AND all the othe phone things it has to do. So tiny little interruptions to its audio routine might not be audible, but they sure as heck can trip your scope trigger. You wouldn;t hear an interruption of some 150kHz arifact, but your scope can see it. And on a totally different tack, that phone has no ground reference, so static buildup or god knows what could be creating little spikes.

DO you have anything that will spit out a steady signal? How about a little power transformer, one that makes 6vAC or 10vAC, even a wall wart that makes AC. SO it is 60Hz instead of 1kHz. It will be as stable qas the mains. COnnect that to the scope. STable? If one source is stable and the other is not, then the scope is not the issue, the signal source is. If everything dances around, then your scope is not set up or your electrical invironment is real noisy.

[Steve Conner]

The phone is also a high powered radio transmitter, in fact two of the things for cell and wi-fi service. Hardly the kind of thing you want around a sensitive analog circuit.

Might be worth trying "flight mode" to turn the radios off and see if that steadies the glitches.

Having said that, emission of ultrasonic crap is a known problem with the kinds of audio DACs ("sigma delta") used nowadays. That was why I ended up building an oscillator, because the crap coming from my USB soundcard ruined the measurements I was trying to do with the distortion analyser.

[J M Fahey]

I use all three of them: "real" analog RC oscillators, function generators that start with a triangle wave and round it and various frequency MP3s living in my MP3 player.
*Best*, hands down , is the "real" RC oscillator stuff.
When some say that the 1 % distortion of a function generator or the sampling artifacts of a digitized signal are irrelevant compared to the , say, 5% distortion of a typical tube amp, think again.
More importantbthan the "number" is the "type" of distortion.
5% tube distortion comes from predictable, smooth curve non linearity, which bends sinewaves in a way that can´t be distinguished on a scope screen.
Function generator sinewaves have the original triangle lurking in, and very much alive.
Not noticeable if you, say, drive a power amp straight, but given the huge (and that´s an understatement) treble boost , bright, presence, mid cut, whatever , (or all of above) present in most guitar amps the ugly spikes come back strong ... unless you have an expensive function generator.
Or the wave kinks or bends sideways in a strange way.
RC generated sinewaves do not, and more accurately show whatever effect that particular stage is having on it.
My "pocket" kit is made out of a function generator, 40Hz to 8KHz for general purpose testing, and a quite pure 400Hz/1KHz (switchable) RC generator to check nonlinearities, setting bias, etc.
Both homemade in cigarette case size boxes, 9V battery powered.
Plus 40Hz to 8KHz 1/3 octave MP3s *always* available in their own folder in my MP3 player.
Plus a very useful "Multifrequency" signal courtesy of Rohde and Schwartz.
Sounds like a Dalek (Robot for those outside UK) raspy voice, which has "a little of everything", good to hear what tone controls actually do.
So, in a nutshell, if you can get a real RC oscillator, used, for a low price, go for it.

[Gtr_tech]

Enzo,

Thanks for the tip. Though I had previously tried adjusting my trigger above zero without success, your post led me to look at the problem again. What I found is that if I set the trigger coupling to HF reject (which attenuates all signals with a frequency higher than 150 kHz, though I don't know why I'd be seeing frequencies that high from an iPhone headphone output) it increased the stability, then as I dropped the voltage, I also adjusted the trigger sensitivity. Now I can get down to around 50mV rms and still keep a stable waveform (whereas before, I needed around 500mV before my scope went haywire). Still, I see the built-in frequency counter jumping around in value, while if I use the built-in probe test connection on the scope. or my analog synth, a Moog Voyager. the frequency stays rock solid.

I realize I could use my Moog as an audio generator - it's square wave isn't perfect but it's better than the iPhone app, and I can get a sine wave by using it's LPF to remove all the upper harmonic from any of the waveforms - but it's not very portable.

Mark

You can turn off the oscillators on the Moog and set the filter to self oscillate (full resonance). That should give you a decent looking sine wave suitable for testing.

[Gtr_tech]

B&K 3025 function generator TESTED GOOD | eBay

I have 3 of these and they work just fine for t-shooting. B&K is kinda the Chevy of test gear. Snubbed by bench gear snobs, but works just fine for the rest of us.

[trobbins]

For simple troubleshooting and investigation I've used discarded (by my kids) USB music players and even old portable cd players - just record a sine sample using best mp3 format and cross-check for any inherent level issues re distortion - the main advantage to me being no ground loop issues that may arise from a mains powered instrument. Of the two mains powered instruments I have, one has a battery power source input which I use for more convenient testing when I want to eliminate hum loop contributions.

If you're doing comparitive measurements with a spectrum analyser, then as long as the base source distortion is fairly low, it is fairly easy to observe changes in harmonic levels, without them having to be 'zero' at the input.

[imaradiostar]

Geez...this is a complex and long thread.

If you're LISTENING to the signal, play a guitar or music through the amp!

If you're scoping the signal and looking for aberrations it's really tough to beat a simple sine wave audio generator. My main one is a Heathkit tube-based model. I have another old tube one (Knight?) with a sticker on it that says something to the effect of "the john f Kennedy center for retarded persons" which is totally un-PC but interesting to note. I have a GIANT Jackson brand tube dinosaur with a dial that sweeps frequency and a needle on the front that lets you know where you are. It has a giant low impedance class a output section and frankly would probably make a decent low wattage amp with some internal tweaking.

My point about my ancient generators- they're tube based, easily recapped/rebuilt, produce a pretty pure sine wave, and have knobs/switches that allow me to easily go from 80Hz to 1kHz and back to 40Hz. They may still be around when I'm dead and they can produce more signal than any computer based generator which you sometimes need on a tube amp.

I guess I'm saying don't overthink it.

I'd go with an HP dinosaur (3325a or something) over my tube generators but in the end- the tube based ones look better on my bench!

jamie

[Enzo]

Old HP 200 series make nice clean sine waves, and there are a lot of them out there.

I found one for $5. One bent plate shorting the tuning cap, straightened it out, et voila, working oscillator.

And nothing wrong with all that Heathkit stuff out there. Go to the next few Ham Radio gatherings in your area.

[mozwell]

Hi
i use this simple circuit for a sine wave generator. it is fairly low distortion sine wave output.
It is based on harmonic cancellation, as i recall.
the circuit is from Silicon Chip (Australian electronics magazine).
You need a dual 5K pot for frequency, but its easy to build, and will give a signal up to several volts.
For amps & pedals, i find its more than enough. Its cheap, which is the main reason i built it rather than purchase one, especially as i only rarely use it.
Use TL074 etc for the op amp, dont use LM324, as they are not high speed enough for audio.

[amgard]

I use a little generator I built myself based around an ICL8038 sig-gen chip. Gives (sort-of)sine, triangle, and square (open collector) - cost less than £10 all in - amply good enough for most occasions.

There's a BIG advantage with Enzo's suggestions of a music feed or a liitle Casio organ though - when soak testing an amp with a generated steady tone - it CAN drive you crazy