Why my DC-DC output has a 100Hz noise on it?

[citroen]

My DC-DC converter is a boost voltage mode converter.
When I test the output voltage ripple using AC couple, there is always a 100Hz noise on it, no matter has a load or not. Please see attachment
I don't know where does the noise come from(from the power,ground, the converter itself or ...), Does anyone ever meet the same problem?

 

[nitishn5]

Is it some pickup from some rectified ac? Like in a tubelight?

 

[BradtheRad]

Sorry, clicking your attachment brings up an error message 'invalid attachment'.

Is your converter powered by mains AC? 50 Hz? That would be a prime candidate for sending noise through your converter. After being rectified by a diode bridge, it's 100 Hz.

 

[citroen]

Sorry, clicking your attachment brings up an error message 'invalid attachment'.

Is your converter powered by mains AC? 50 Hz? That would be a prime candidate for sending noise through your converter. After being rectified by a diode bridge, it's 100 Hz.

Tks BradtheRad
I have reuploaded the attachment.
And my converter is powerde by a DC not an AC.

 

[goldsmith]

My DC-DC converter is a boost voltage mode converter.
When I test the output voltage ripple using AC couple, there is always a 100Hz noise on it, no matter has a load or not. Please see attachment
I don't know where does the noise come from(from the power,ground, the converter itself or ...), Does anyone ever meet the same problem?
100hz-noise.png

Hi Citroen !
And what kind of DC to DC converter you're referring to ? can you post your schematic ? then we can judge about where those LF ripples are coming from . is that any isolated DC/DC converter ? or perhaps it is something like a buck converter ? it can be because of gang of reasons ! for instance lack of a good filter after rectification or ... etc . for clear help , clear information is required .
Best Wishes
Goldsmith

 

[mtwieg]

I'm guessing that it's some sort of burst mode operation caused by light load. What kind of controller are you using?

 

[neddie]

Another thing to check is that your scope is set correctly. Are you not maybe aliasing the signal you are seeing.
The frequency may be much higher , but if your timebase is set incorrectly (too low) on a digital scope , you may "see" a signal
that is not actually there!!
Just a thought.

 

[citroen]

I'm guessing that it's some sort of burst mode operation caused by light load. What kind of controller are you using?

Tks mtwieg.
It is a Boost DCM voltage mode controller.

- - - Updated - - -

Another thing to check is that your scope is set correctly. Are you not maybe aliasing the signal you are seeing.
The frequency may be much higher , but if your timebase is set incorrectly (too low) on a digital scope , you may "see" a signal
that is not actually there!!
Just a thought.

Tks neddie.
I have set the bandwidth of scope, using GHz or 20Mhz has the same result.

- - - Updated - - -

Hi Citroen !
And what kind of DC to DC converter you're referring to ? can you post your schematic ? then we can judge about where those LF ripples are coming from . is that any isolated DC/DC converter ? or perhaps it is something like a buck converter ? it can be because of gang of reasons ! for instance lack of a good filter after rectification or ... etc . for clear help , clear information is required .
Best Wishes
Goldsmith

Tks Goldsmith.
I am testing a boost DCM voltage mode DC-DC controller.
And I will post the peripheral schematic later to make the question more clear.

 

[mtwieg]

I meant what is the specific part number of the controller. Preferably a schematic as well.

Do you notice if the error signal also has the same 100Hz signal on it?

 

[goldsmith]

I am testing a boost DCM voltage mode DC-DC controller.

Hi Citroen !
Well , where your input DC source is incoming from ? is that coming from an AC source and then a rectifier and then a capacitive filter ? or perhaps it is coming from a pure DC source ?

And As Mtwieg mentioned , check the error signal too . and let us know is there the same LF component there too ?

And I will post the peripheral schematic later to make the question more clear.

Good idea !

Best Wishes
Goldsmith

 

[schmitt trigger]

"Tks neddie.
I have set the bandwidth of scope, using GHz or 20Mhz has the same result."

What Neddie is saying, is not the channel's analog bandwidth (which one sets in a menu screen) but the sampling rate, which can create alias distortion.
The timebase in your image is very slow (10ms/div), or 100ms across the screen. If your power supply is switching at 100 Khz, then you require a record length of at least 20,000 samples to prevent alias. Of course, if the switching frequency is higher, your record length would be increased proportionately.

I'm not saying that, in this particular problem, this is the root cause, but only alerting people that DSOs can lie to you if one is not careful.

 

[citroen]

I meant what is the specific part number of the controller. Preferably a schematic as well.

Do you notice if the error signal also has the same 100Hz signal on it?

It is just a experimental controller.
And the error signal has the same 100Hz signal on it.(test the feedback point)

 

[schmitt trigger]

Another possibility is feedback un-stability.

But without additional information (i.e. schematic, or controller type) everything so far is only speculation.

 

[mtwieg]

We'll definitely need a detailed schematic of the experimental controller in order to help.

 

[citroen]

Hi Citroen !
Well , where your input DC source is incoming from ? is that coming from an AC source and then a rectifier and then a capacitive filter ? or perhaps it is coming from a pure DC source ?

And As Mtwieg mentioned , check the error signal too . and let us know is there the same LF component there too ?

Good idea !

Best Wishes
Goldsmith

Input source is a pure DC power supply.
And attachment here is the peripheral schematic I am using now.


- - - Updated - - -

Another possibility is feedback un-stability.

But without additional information (i.e. schematic, or controller type) everything so far is only speculation.

Tks schmitt trigger.
I don't have any equipment that can test the loop stability.
So I test the load regulation, and there are not any rings during the recovery stage.
I think the loop is stable.

 

[schmitt trigger]

0.1 uF for the output capacitor, seems to be an awfully small value for a boost convert, specially since the boost converter's output current is always discontinuous. Not saying that this is the case, but that is very odd.

What frequency are you switching at? Is your boost controller current or voltage mode...it appears voltage mode since you don't have any current feedback. Does your converter go above 50% PWM, and if so, does it have slope compensation?

 

[citroen]

0.1 uF for the output capacitor, seems to be an awfully small value for a boost convert, specially since the boost converter's output current is always discontinuous. Not saying that this is the case, but that is very odd.

What frequency are you switching at? Is your boost controller current or voltage mode...it appears voltage mode since you don't have any current feedback. Does your converter go above 50% PWM, and if so, does it have slope compensation?

Switching at 500KHz
It's voltage mode, and don't have slope compensation.

 

[mtwieg]

Input source is a pure DC power supply.
And attachment here is the peripheral schematic I am using now.
View attachment 90828

- - - Updated - - -



Tks schmitt trigger.
I don't have any equipment that can test the loop stability.
So I test the load regulation, and there are not any rings during the recovery stage.
I think the loop is stable.

You need to show the internal workings of the controller. A black box doesn't help at all.

 

[FvM]

Looking exactly at the waveform in post #1, the ripple isn't 100 Hz rather than about 95 Hz. So it's unlikely to be a mains interference.