Need help understanding dirty signals.
Pic below shows input signal vs output signal.
Where my brain loses the plot is when one starts to see dirty lines.
Is it a scope thing. It would be nice if there was a database or list explaining what this noise is. How its caused. And the different kinds of noise on signals.
In this input vs output the noise on the in (Yellow line) seems random. But I have encountered other types of noise where the wave form is made up of very clean clear small zigzag lines. Then there are those where the wave carries minor spikes inside the curve. Those I found seem to appear when there are dirty of failing caps in the circuit. Sorry if this is a very basic question. But it would seem I need to grasp these concepts to make any progress with this hobby.
Pic below shows input signal vs output signal.
Where my brain loses the plot is when one starts to see dirty lines.
Is it a scope thing. It would be nice if there was a database or list explaining what this noise is. How its caused. And the different kinds of noise on signals.
In this input vs output the noise on the in (Yellow line) seems random. But I have encountered other types of noise where the wave form is made up of very clean clear small zigzag lines. Then there are those where the wave carries minor spikes inside the curve. Those I found seem to appear when there are dirty of failing caps in the circuit. Sorry if this is a very basic question. But it would seem I need to grasp these concepts to make any progress with this hobby.
Attachments
I guess the image is not stable, but it runs. if so, your oscilloscope is not trigger. Try other triggering options or press the button auto.
The yellow trace with the noise is the smaller input signal. The cleaner blue trace is the output signal. The noise could be RF or an oscillation or an adjacent oscillation, but it isn't getting through the amplifier.
Agree, the noise is on the input signal. Connect your source to the scope (forget the amp for the moment) and try to get that clean.
It most probably is a grounding issue, so it can be that when you connect the source directly to the scope all is clean.
If so, you should review your wiring including the grounding between source, amp and power supply
Can you make a pic of that wiring and a drawing?
Jan
It most probably is a grounding issue, so it can be that when you connect the source directly to the scope all is clean.
If so, you should review your wiring including the grounding between source, amp and power supply
Can you make a pic of that wiring and a drawing?
Jan
Make sure your scope ground is close to the input signal input.
If its not there will be ground noise getting in.
If its not there will be ground noise getting in.
If you can try to connect the grounds of the generator and scope to the same point. The input would be the logical choice, Often mains grounds are connected and nothing can be done about that. Some power supplies may have a facility for disconnecting it from that.
Another factor that has me scratching my head a bit. Having used them for years I would be using DC coupling on the scope unless I had good reasons not to and preferring to use the scope's shift to bring the waveform into the view if needed. The AC coupling is there for when this can't be done,
Another factor that has me scratching my head a bit. Having used them for years I would be using DC coupling on the scope unless I had good reasons not to and preferring to use the scope's shift to bring the waveform into the view if needed. The AC coupling is there for when this can't be done,
My scope ground is hooked up to the RCA Input Jack. Can't get any closer.
Will try and draw a schematic of how everything is hooked together. I should also try the same test with my old Fluke 105B and see if this noise shows up there. Or I can try bumping up my input signal to 3 volts peak to peak. And see if that helps.
Will also from now on take pics of various types of noise encountered and solutions if possible. Should have taken a screenshot off the spikes caused when a cap goes bad. Will put in a request to john audio tech if he could cover it in one of his videos. i.e. types of noise and distortion seen frequently when working with audio circuits on scopes.
I had a similar problem with input noise, except it was composed of narrow spikes occurring at a few hundred kiloHertz. Was it the wall wart for the little signal generator? Nope. Dimmer to the room lights? Nope. Finally unplugging the standing desk about three meters away cleaned the scope traces right up. Its power supply was not quite as silent as one might hope, but standing desks are not found that often in electronics labs so I wouldn't be surprised nobody had complained of this before.
All good suggestions above. I'd like to add that the amplifier in the scope has a wide bandwidth, a high input resistance, perhaps a higher gain to overcome the loss of a 10X probe, and protection components at its input. All of these can contribute to noise. The amplifier under test likely has a lower bandwidth than the scope amplifier and hence can amplify and output a less noisy sine wave at 500Hz. The larger amplitude of the amplifier output makes the scope channel showing the output less noisy.
If your scope can average multiple traces, the value of the signal at any particular time will be the same, while the value of the noise at that time will be zero. This can give you a better look. But to use that feature, you need to trigger the sweeps from the "least noisiest" signal available (the output of the amp, or even better, the trigger out of the signal generator).
Tom
If your scope can average multiple traces, the value of the signal at any particular time will be the same, while the value of the noise at that time will be zero. This can give you a better look. But to use that feature, you need to trigger the sweeps from the "least noisiest" signal available (the output of the amp, or even better, the trigger out of the signal generator).
Tom
Concur with tom, 100MHz scope, input is at 200mv/div which with a 10X probe means the scope is seeing 20mv/div with 100MHz. If you can average, it should get cleaner. Hmm, just noticed, looks like output has 20MHz bw limit but input does not if I read the bottom legend of the scope correctly.
No the bandwidth limit is on for both channels. For some reason the legend does not show it. And I did notice that stepping back helped some. This post is more about trying to learn what dirty traces are telling us. As in whats really doing on i.e. why is it when we zoom into any signal we see that the main signal is made up of smaller zig zag signals. Just trying to understand how noise is carried on a signal. And what it looks like on the scope screen.
Im now also confused why somebody said use d/c mode I thought you always use a/c mode for audio signal tracing.
Im now also confused why somebody said use d/c mode I thought you always use a/c mode for audio signal tracing.
I think the intended meaning of "d/c" there is "direct coupled" rather than Direct Current. It is used here as an abbreviation for a direct coupled rather than capacitively or inductively coupled input mode. It's usually selectable and as the title suggests, it allows DC as well as AC voltage components to be amplified together and the result shown on the screen. This could come in handy for locating problems with the test device as well as the application, as we can see here.
Last edited:
You haven't shown that the input source in itself is clean. Without that it all remains speculation and a waste of time.
Your call.
Jan
Yes me. Electronic labs for donkeys years. I mentioned scratching my head. I assume I was told to do that by my betters. AC when you must,
Thinking about it one of the reason is probably being aware of DC levels. Even on a class A with a single rail supply that needs to be known and also when tracing signals through a circuit. A single rail supply amp will also be ac coupled to what it's driven by and driving so the capacitor is already there.
Further scratching. AC coupling also reduces the bandwidth of the scope. A comment from a maker https://www.picotech.com/library/oscilloscopes/ac-coupling
AC coupling is very important when measuring an offset AC signal, but the added attenuation and blocking adds problems for some AC signals, such as the Square Wave, or low-frequency sine wave as they can start to attenuate in weird, unpredictable ways.
As a rule of thumb, stick to DC Coupling. When measuring a waveform such as a power supply ripple, or a biased transistor amplifier input where the AC waveform is offset by a large voltage, switch to AC Coupling to provide an aid to reading these AC waveforms effectively.
Another point is that the makers always provide way more offset adjustment than needed to just get the trace to travel along the 0 line. On many that is rather fiddly to set very accurately.
The links suggests using AC coupling to check a supply. Sensitive range are usually needed for that and it can cause noise problems in odd way. A wild for instance the power output stage is boosting the current levels and that is causing problems with the whole grounding in the circuit, Noise can be seen on the output. The current levels there are higher than at the input,
Mains loops as all of the gear is hooked up via that? Not usually a problem but didn't someone mention something on the mains was causing them problems? I can think of one instance where it was a problem. A rather high kw motor driver that was a absolute pig to look at even without that problem. Doing things that shouldn't be done help.
wise man told me to understand noise first build noise free signal
So busy building this light bulb
So busy building this light bulb
So did you measure the noise of your source? Results?
What will you (and we) think if you build this and you still have the noise?
Jan
What will you (and we) think if you build this and you still have the noise?
Jan
Jan yes I have measured the FFT noise on my source. Its fine. I tried various cheap function generators. But found the best and cleanest signal to be put out by my Samsung Galaxy Note 9. (Exynos Ver).
I then connected my source. To the transistor amp Im building which is lets call it the Rod Elliot PA3 inspired.
The plan for this amp is to use it as a test bed for various transistors. (Im still fixated with my first question I had when I joined this forum. Which is I have a bag full of transistors. How do I pick the right / best / lowest noise / lowest distortion transistor for the job.)
Lets change the story up a bit here.
So clean source to amp. And output has a lot of Harmonics.
So now what ?. Start tracing the signal from input to see at what stage the harmonics start ?. Does this sound like a legit plan.
Or is there a better way to do this. ?
I understand transformers and voltages can play a role here so I have 4 transformers Toroidal / EI from 12 to 32 volts that can be switched in and out to see if it helps drop harmonic distortion.
This is a new amp, a new question and has nothing to do with the first picture. I notice that as soon as the signal hits the first BD140 the harmonics start. Now what ?. Replace the BD140 with a diffirent brand / transistor ?
Pull out a component to break the forward signal path and only proceed once we clear the harmonics at this point / Try with a different PCB / Bread board.
Or go back to the drawing board. Make another simple bread board circuit for a single transistor amp. And use that to test various transistors for harmonics.
I then connected my source. To the transistor amp Im building which is lets call it the Rod Elliot PA3 inspired.
The plan for this amp is to use it as a test bed for various transistors. (Im still fixated with my first question I had when I joined this forum. Which is I have a bag full of transistors. How do I pick the right / best / lowest noise / lowest distortion transistor for the job.)
Lets change the story up a bit here.
So clean source to amp. And output has a lot of Harmonics.
So now what ?. Start tracing the signal from input to see at what stage the harmonics start ?. Does this sound like a legit plan.
Or is there a better way to do this. ?
I understand transformers and voltages can play a role here so I have 4 transformers Toroidal / EI from 12 to 32 volts that can be switched in and out to see if it helps drop harmonic distortion.
This is a new amp, a new question and has nothing to do with the first picture. I notice that as soon as the signal hits the first BD140 the harmonics start. Now what ?. Replace the BD140 with a diffirent brand / transistor ?
Pull out a component to break the forward signal path and only proceed once we clear the harmonics at this point / Try with a different PCB / Bread board.
Or go back to the drawing board. Make another simple bread board circuit for a single transistor amp. And use that to test various transistors for harmonics.
Words, words, words ...
Sorry.
Show input source FFT or scope graph, show output FFT or scope graph.
Schematic.
Otherwise, the replies will just be words, words, words.
Try to think it through logically, focus on the work at hand.
You are throwing up solutions before knowing the problem.
We need to identify the problem. Then a solution is easy. Always.
Jan
Sorry.
Show input source FFT or scope graph, show output FFT or scope graph.
Schematic.
Otherwise, the replies will just be words, words, words.
Try to think it through logically, focus on the work at hand.
You are throwing up solutions before knowing the problem.
We need to identify the problem. Then a solution is easy. Always.
Jan
If the scope has a noise filter or bandlimit option, try engaging that. You might also find a sweet spot by adjusting the trigger level, too.