Measuring 800hz ripple looks like noise

[Davewalker5]

Measuring 800hz ripple from a power supply looks like noise

I didn't know the output of a power supplies noise was ripple, I thought ripple was a sawtooth looking waveform

60hz ripple looks like a sawtooth waveform

400hz and 800hz ripple looks like noise

When measuring Ripple:
I'm using a Tektronix TDS 2022 O-scope

Acquire Menu : click the Average to take away the noise from the ripple

Does the trigger Holdoff help stablized ripple noise when measuring 400hz or 800hz ripple?

 

[(*steve*)]

There are multiple sources of noise. If you have a spectrum analyser you can see that the noise spectrum starts from essentially 0Hz and hoes all the way up as far as the spectrum analyzer can analyze.

There will be peaks. Many of these will be related to your mains frequency, probably a significant one at double the mains frequency, and others at multiples of this due to the sawtooth (ish) nature of ripple.

Then there are other sources. A switchmode power supply will have noise at its operating frequency. If it's discontinuous (i.e. it turns on and off because the load is low) it will have a lower frequency component at that switching rate.

And you have many noise sources. Transistors, zener diodes, even resistors generate noise. These will often add the random wideband noise that you see.

Another source is external interference. You may notice noise at the same frequency as a local FM transmitter. This is not from the power supply, but picked up (more than likely) on the leads between your power supply and your spectrum analyser.

If you're using an oscilloscope, you really only see the gross detail, so that means the largest peaks. And for you, that is 60Hz (are you sure it's not 120Hz?) and the 400/800 Hz.

If you have several power supplies, compare them. You'll almost certainly note that some have less noise than others, and probably that the dominant frequency of the noise is different.

You might even like to try a regular battery. Do you see noise? Where might this be coming from?

 

[Davewalker5]

If you're using an oscilloscope, you really only see the gross detail, so that means the largest peaks. And for you, that is 60Hz (are you sure it's not 120Hz?) and the 400/800 Hz.

Yes the output ripple of a 60hz power supply is 120hz ripple, looks like sawtooth waveform

The output ripple of a 400hz and 800hz power supply looks like noise, it doesn't have the sawtooth waveform

I didn't know ripple was an AC noise on a DC reference

I thought ripple was an AC waveform that looked like a sawtooth waveform riding on a DC reference

What is the difference between the two different types of AC ripple?

Or what other types of AC ripple is there?

 

[KrisBlueNZ]

What is this power supply a part of? What does the connected circuitry do?

 

[(*steve*)]

I didn't know ripple was an AC noise on a DC reference

Think of it like waves on the surface of a swimming pool. The depth of the pool is the DC voltage, the variation at the top is like AC, but on top of the DC.

I thought ripple was an AC waveform that looked like a sawtooth waveform riding on a DC reference

Ripple is normally the name given to a noise source that is regular and repetitive and predictable. In this case you know that 120 times a second it will reach a peak in some manner and then fall back to a trough in some (possibly different) manner. And you know it's going to happen roughly the same next time. It may vary with load, etc., but it's a regular waveform at a given operating point.

Or what other types of AC ripple is there?

It's useful to differentiate ripple (which is a regular signal) from random noise. Random noise is a variation in the output signal that is not predictable in any other way than by talking about averages. You might note that it has an average RMS voltage of some voltage, but you cannot tell from instant to instant what the actual voltage will be.

Other sources of noise are distinguished by their origin. For example if you noted a signal at 990kHz, and the amplitude looked noisy, but the bandwidth of the signal was very narrow, you might well wonder if you had a nearby AM radio station transmitting at 990 on the AM dial. You would call this RF interference, because you know the source of it.

You might notice spikes of noise that correspond to turning a large electrical appliance off and on. This may be transmitted through the mains, and so that is mains borne interference.

There are all sorts of noise sources. If you are really dedicated you could include:

• Thermal
• Shot
• Flicker
• Inter-modulation
  
• Crosstalk
• Impulse
• Interference
  
• Burst
• Avalanche
• Industrial
• Atmospheric
• Solar
• Cosmic

...and probably others

In most cases you're interested in the dominant sources (and in your case Solar and Cosmic noise are unlikely to feature heavily)

 

[Davewalker5]

400hz and 800hz ripple pictures

I still can't believe this is ripple, it looks like noise to me , take a look

 

[Davewalker5]

This is 120hz ripple, this is what i'm use to seeing as a ripple waveform , it's sawtooth like

I think ripple should only look like this , take a

 

[(*steve*)]

I would call the first one ripple with some additional noise component. The second one looks more like noise.

I think you should only look like this:

 

[Davewalker5]

But why do they call this ripple and not noise?

 

[(*steve*)]

But why do they call this ripple and not noise?

Read post #5 in this thread

 

[Davewalker5]

Ripple is normally the name given to a noise source that is regular and repetitive and predictable.

So ripple is noise?

but it's a regular waveform at a given operating point.

operating point which is the DC reference it's riding on?

differentiate ripple (which is a regular signal) from random noise.

So ripple is noise?


This is why I turn the Average function on the O-scope to take away the noise from the ripple but it still likes like noise to me

 

[(*steve*)]

So ripple is noise?

Yes.

operating point which is the DC reference it's riding on?

No it's the current operating state of the device. For a power supply this might be "when set to 5V output and 100mA is being drawn". Some aspects of noise may well change if you change the output voltage or current.

So ripple is noise?

What, are you expecting a different answer?

This is why I turn the Average function on the O-scope to take away the noise from the ripple but it still likes like noise to me

The average function -- in your case an average over 16 samples -- will tend to cancel out any signal that is not regular and repetitive nature. Is ripple like this?

Before you ask me that question, note that I described it in post #5 as "regular and repetitive and predictable".

 

[Davewalker5]

The average function -- in your case an average over 16 samples -- will tend to cancel out any signal that is not regular and repetitive nature. Is ripple like this?

So what should I set the samples to? on the average function

No it's the current operating state of the device. For a power supply this might be "when set to 5V output and 100mA is being drawn". Some aspects of noise may well change if you change the output voltage or current.

If you overload the device it can make the current higher and the noise higher

Overloading a device causes the device to work harder stressing it out

 

[(*steve*)]

So what should I set the samples to? on the average function

The average function is not magic. It CAN'T know what is signal and what is noise.

All it can do is attenuate that part of the signal which does not regularly appear at the same point in the waveform.

If you overload the device it can make the current higher and the noise higher

Who said anything about overloading? In this case I just said changing the load. I could also have reduced it.

Overloading a device causes the device to work harder stressing it out

I'm pretty sure that by this definition you are overloading me.

 

[Davewalker5]

All it can do is attenuate that part of the signal which does not regularly appear at the same point in the waveform.

True, I just thought most techs used the average function when measuring ripple

I guess you don't or you set the sample other then 16 samples

Who said anything about overloading? In this case I just said changing the load. I could also have reduced it.

Yes the ripples waveshape depends on the load

 

[(*steve*)]

True, I just thought most techs used the average function when measuring ripple

It depends on how much other noise there is.

I guess you don't or you set the sample other then 16 samples

Or use an analog scope.

That's a difference between analog and digital scopes. Digital scopes tend to make the noise that's present more visible than it would be on an analog scope.

Yes the ripples waveshape depends on the load

Probably.

 

[Davewalker5]

That's a difference between analog and digital scopes. Digital scopes tend to make the noise that's present more visible than it would be on an analog scope.

Why does digital scopes tend to make the noise more visible?

Maybe this is the problem , because the test could have been for an analog O-scope

 

[(*steve*)]

Why does digital scopes tend to make the noise more visible?

Watch this.

Maybe this is the problem , because the test could have been for an analog O-scope

What's an O-scope?

 

[Davewalker5]

Steve, that guy on the video is claiming that digital oscilloscopes show the noise because analog oscilloscope are hiding it.

You have to go into menus and menus to adjust and turn things on in digital oscilloscopes to make it like an analog oscilloscope

But this guy is claiming that digital oscilloscopes don't create noise or make the measurements noisy

Also turning ON the bandwidth on the Oscilloscope cuts out the high frequencys of the noise and turning ON the average will clean up the noise

both I have done as a tech

 

[(*steve*)]

That guy is Dave Jones. Call him Dave.

Remember there are different types of noise. Averaging "hides" various types of random noise -- but ripple isn't random noise.

And you don't "turn on the bandwidth" you reduce it.

So it answers your questions?