0
\$\begingroup\$

I am looking into fitting ferrites to my SMPS power supplies output leads (wall warts.)

I have been getting some strange noises on my valve stereo amplifier since using a couple of LV DC PSUs as above. One for Bluetooth and another for a Cambridge audio D.A.C.

Is it a case of just fitting the clip-on cylindrical variety or taking a given number of turns thru each device?

I have tried a single pass but this seems to give no obvious improvement. Also will a number of (effectively) ampere turns affect the SMPS loading?

If I remove the SMPS supplies all is well so am sure the high frequency noise is due to these (nssty) SMPS devices. If this is the case I will have to invest in a proper linear PSU.

\$\endgroup\$
  • 1
    \$\begingroup\$ You might need 10 or 100 nF from each DC output line (both pos and neg) to solid earth to clear common mode noise. \$\endgroup\$ – Andy aka Oct 12 at 16:50
  • 1
    \$\begingroup\$ I would suggest some ceramic 0.1uF to 0.47uF capacitors where the power comes in and where it is connected to along. A good way to eliminate common mode noise is with a coupled inductor. \$\endgroup\$ – bunker89320 Oct 12 at 16:55
  • \$\begingroup\$ RF noise is often Common Mode and can leak into differential mode by poor PSRR which needs high GBW. show amp design details and filter. Most likely your filter is too high LPF and poor CMRR \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Oct 12 at 17:14
  • 1
    \$\begingroup\$ If you can hear the noise, a clip on ferrite may not filter it (couldn't hurt to try, though). They are really for RF noise. Not audio frequency noise. Maybe a multi-stage LC low-pass filter would be more useful. Or a large, multi-turn toroidal choke such as is used for conducted emissions compliance. \$\endgroup\$ – mkeith Oct 13 at 4:13
0
\$\begingroup\$

The classic 'clip-on' ferrite is intended simply for the power cable to pass through once with the hole diameter ~equal to the cable diameter.

If you're able to pass the cable through more than once, there's an advantage since the effectiveness of the filter ~~proportional to the number of 'turns'. If you can pass the cable through multiple times, so much the better ! There's no 'loading effect' at 50/60 Hz to concern yourself with as the ferrite only has meaningful effect at radio frequencies.

Ferrites designed for this application are a bit different to the ferrites used in switch-mode transformers etc. They are 'lossy' and convert some energy into heat. This makes them more effective at suppressing interference than an inductor alone.

You might like to investigate ferrite ring cores that allow you to wind many turns on them if you're having trouble reducing your EMC problem.

It's not asked but you may also get some benefit from a ferrite on the DC output cable too.

Some more info is available here ..... https://en.wikipedia.org/wiki/Ferrite_bead#:~:text=A%20ferrite%20bead%20%28also%20known%20as%20a%20ferrite,ferrite%20ceramic%20to%20build%20high-frequency%20noise%20suppression%20devices.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ WOW! Thanks to all you guys who have posted Useful advice. It does seem that I will need some time to run some tests, applying suggested possible fixes. \$\endgroup\$ – Oldnewt Oct 14 at 11:03
  • \$\begingroup\$ WOW! super response from several contributors, can get going on some trials and tests as soon as I have gathered some parts. One of the above options may at least reduce the noise. I have now managed to borrow a digital Osc/scope which will be an asset. Thanks again tto all. \$\endgroup\$ – Oldnewt Oct 14 at 11:07
  • \$\begingroup\$ A further thought. As you're having some trouble suppressing the interference, try fitting more than one clip-on ferrite. to the cable(s). Also, since it's your stereo experiencing the interference, is the device powered from the SMPS connected to the amplifier's inputs ? There is another route for interference to enter you see, through the audio ground. You don't necessarily even have to have that input selected and turned up. \$\endgroup\$ – Graham Stevenson Oct 14 at 11:22
  • 1
    \$\begingroup\$ today sourced a multi output linear PSU. Home built by local electronics wizard who is copying the circuit for me to build. His device cleared the HF noise problem at a stroke. \$\endgroup\$ – Oldnewt Oct 19 at 13:48
  • \$\begingroup\$ I suspect that your problem was actually the audio frequency leakage current from the PSU working its way into the amplifier, not an EMC issue at all. \$\endgroup\$ – Graham Stevenson Oct 19 at 19:34
0
\$\begingroup\$

It seems as I suspected, removing the SMPS from the loop immediately solved the issue. Should have twigged this earlier as I had a similar problem some time ago with adding a budget brand b.tooth adaptor to my Denon compact RCD39DAB. Never trusted SMPS since their inception back in the early 1990s, where if a psu fails all else goes with it! Will stick with the more reliable linear supply, as no been running fine for 20 hours not problem. Thanks to all for help and useful suggestions!

| improve this answer | |
New contributor
Norman roscatha is a new contributor to this site. Take care in asking for clarification, commenting, and answering. Check out our Code of Conduct.
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.