Power Supply Noise Reduction

Thread Starter

business_kid

Joined Feb 5, 2013
22
I have a Samsung Laptop PC with power supply which I want to use for streaming. When I do, and amplify the sound, there is an audible noise - very irritating on silent sections, but tolerable otherwise. Unplugging the power supply kills the interference, so I conclude it is the external power supply. The supply output is 19V/3.16A, so I imagine I have some voltage tolerance. The interference sounds like white noise - no set frequency, but audible.

Has anyone found a good approach to this problem? My first idea (Space is tight) was a choke between 1st & second electrolytics to give a low pass effect, but I have not tried this yet.
 

ian field

Joined Oct 27, 2012
6,536
I have a Samsung Laptop PC with power supply which I want to use for streaming. When I do, and amplify the sound, there is an audible noise - very irritating on silent sections, but tolerable otherwise. Unplugging the power supply kills the interference, so I conclude it is the external power supply. The supply output is 19V/3.16A, so I imagine I have some voltage tolerance. The interference sounds like white noise - no set frequency, but audible.

Has anyone found a good approach to this problem? My first idea (Space is tight) was a choke between 1st & second electrolytics to give a low pass effect, but I have not tried this yet.
Maybe the overall design relies on filtering provided within the laptop itself - presumably the laptop has a soundcard built in, so the supply must get adequately filtered at some point on its way.

Some discount stores have "generic" 19V and possibly ones with selectable output voltage and an assortment of power jack attachments. I bought one specifically for bench testing projects, and not had the problem you describe.

Maybe some homebrew filtering between the power brick and your application could help. Any scrap appliance with an SMPSU in it, will have various capacitors and inductors you can salvage for experiments. There's usually at least one common mode choke, that with capacitors at each end might stop the mush.
 

Dr.killjoy

Joined Apr 28, 2013
1,196
I am a complete newbie here.. But I have read alot of people put an inductor on the output of the smps to tame the noise .. I am not sure how it works or why but its worth a shot .. Oh I could be completely wrong and should be taken out back and beaten..
 

AnalogKid

Joined Aug 1, 2013
10,971
An inductor in series with the ps output followed by a capacitor to ground will be more effective than just a bunch of capacitors alone, although that is the place to start. Try a small group of capacitors with the values 0.01 uF, 0.1 uF, 10 uF, and 100 uF. The first two are ceramic, the last two are aluminum or tantalum electrolytic. All caps should be rated for at least twice the ps output voltage. Important, keep the leads as short as possible. This should have a noticeable impact on your noise. If you need more, then we talk inductors.

ak
 

ian field

Joined Oct 27, 2012
6,536
Also worth trying a ferrite clamp on the output power lead, the effect can sometimes be miraculous.
When I repaired monitors for a living, I harvested any ferrite clamps from the one's that got scrapped.

After one particular PC motherboard upgrade, FM reception became impossible - I started clipping ferrite clamps randomly on any mains leads I could get at. This reduced the hash a lot, but not perfect.

Then I found a couple of sets of ferrite clamps for large flat cables, for each set I wound a speaker cable a few times round one half the clamp, then clamped the 2 halves together - that finally fixed the interference.
 

blocco a spirale

Joined Jun 18, 2008
1,546
I have various plug-in chargers in the car that all use cheap and noisy switching regulators and they seriously affect FM radio reception. Adding ferrites to the cable makes a significant improvement but the interesting thing is that, counter intuitively to me at least, positioning the ferrite as close as possible to the DC-DC converter does not provide the best results; they seem to work best at around 15 to 20cm along the cable from the source.
 
Top