Help with PSU noise

[Gary Barnes]

I hope someone can give me some pointers with a problem I have with an audio circuit powered from a 24VDC switch-mode PSU. The 24V circuit also has some incandescant lemps attached which add a burden of 230mA to the 24V supply. Without the lamps in circuit I can hear a small amount of high-frequency background noise. With the lamps in circuit, the noise disappears as though the lamps are providing some sort of damping. If I halve the lamp burden the noise re-appears, but not so loud.

The PSU exhibits the standard sort of noise bursts normally associated with a switcher, and the only change I can see is that the leading edge of these bursts increases with the lamps out of circuit, but the increase is not on every burst, so I suspect that what I am hearing is the modulated result.

I want to change the lamps to LEDs, and as this circuit needs to be battery-powered I can't really afford a 230mA burden.

I've attached some waveforms of the PSU +VE rail, AC coupled, 200mV full scale, 500ns/div.

Any clues as to eliminate the noise would be greatly appreciated.

With Lamps



without lamps

 

[Arouse1973]

Hello Gary
Is this power supply something you purchased? Do you know much about the design and whether it had audio use in mind when it was designed. I am initially thinking a T-LC filter.
Adam

 

[hevans1944]

If you are going with battery-powered portable operation, why not just replace the switch-mode PSU with a nice quiet 24 V battery?

That high-frequency ringing you see can be very difficult to remove. You could try an LC low-pass filter to knock down some of it, but beware of "sneak paths" amongst the ground or common that manage to bypass the filter. Also there appear to be enough harmonics to fully staff a mariachi band, so it may be very difficult to remove both conducted and radiated PSU noise. This is one reason many radio amateurs avoid switch-mode PSUs. Some PSUs are quieter than others, so if it is an option, perhaps you can try another 24 V PSU and compare noise figures.

 

[Gary Barnes]

Hello Gary
Is this power supply something you purchased? Do you know much about the design and whether it had audio use in mind when it was designed. I am initially thinking a T-LC filter.
Adam

Hi Adam,

This is a purchased PSU notherwise audio specific.

 

[Arouse1973]

Ok so maybe a simple LC or T-LC will help. it's worth a try Gary.
Adam

 

[Gary Barnes]

If you are going with battery-powered portable operation, why not just replace the switch-mode PSU with a nice quiet 24 V battery?

That high-frequency ringing you see can be very difficult to remove. You could try an LC low-pass filter to knock down some of it, but beware of "sneak paths" amongst the ground or common that manage to bypass the filter. Also there appear to be enough harmonics to fully staff a mariachi band, so it may be very difficult to remove both conducted and radiated PSU noise. This is one reason many radio amateurs avoid switch-mode PSUs. Some PSUs are quieter than others, so if it is an option, perhaps you can try another 24 V PSU and compare noise figures.

Hi,
I'm not really bothered about the
ringing. It's well outside the audio band. There does seem to be some sort of audio frequency modulating the amplitude of these pulses which only occurs when the PSU burden is low.

 

[Arouse1973]

That's probably because at low load the frequency of the switching will be lower because it's doesn't have to refresh the energy lost from the output capacitor.
Adam

 

[Gary Barnes]

Ok so maybe a simple LC or T-LC will help. it's worth a try Gary.
Adam

OK Adam I'll give this another go. I did try an LC filter but I may have to play with values, and T-LC might be better.

 

[Arouse1973]

What values did you use?
Adam

 

[Gary Barnes]

A

That's probably because at low load the frequency of the switching will be lower because it's doesn't have to refresh the energy lost from the output capacitor.
Adam

aha! An explanation, thanks!

 

[Gary Barnes]

What values did you use?
Adam

Er... Just checked and not what I thought I was! I'll have another go. Thanks for all your help

 

[Arouse1973]

No probs. 150uH and 220uF cap and another 150uH inductor. The two inductors in series with the output and at the join junction of the inductors the capacitor to zero volts. Or just try one inductor and cap.
Adam

 

[Gary Barnes]

No probs. 150uH and 220uF cap and another 150uH inductor. The two inductors in series with the output and at the join junction of the inductors the capacitor to zero volts. Or just try one inductor and cap.
Adam

OK Adam I'll give it a go and let you know how I get on. Very kind of you to give me your time.

 

[Arouse1973]

No probs Gary.
Adam