I have a long (2m) but thick wire between transformer-based power supply and low frequency power amplifier. Know your best advice ahead, but no, I cant get rid of it.

I can put it in two configurations:

enter image description here

What should I choose and why?

Looking at some transformer-based PSU's like for mixer equipement I can see a long wire with AC, plus a bridge right in a mixer, but not in the PSU. But looking at SMPS especially for notebooks - they are mostly shielded cables, but runs DC - shottky bridges are close to transformer.

  • \$\begingroup\$ The thing about transformer AC powersupplies is in many locations they are only allowed to be sold for devices that require AC to operate. so if the maker wants to aviod using a switched-mode powersupply it helps to require the device to use AC for operation. \$\endgroup\$ Commented Jan 20, 2019 at 6:56

2 Answers 2


With LONG leads between the transformer and the rectifier, lots of EMI will be generated and radiated, as the diodes in the rectifier turn on quickly (microseconds, if not faster) and the rapid change in current thru the long-wires provides external magnetic fields that are very difficult to shield with just copper foil.

And place small capacitors (100uF) right by the rectifier.

Something like this


simulate this circuit – Schematic created using CircuitLab

  • \$\begingroup\$ Great, just the opposite answer :-) so you think better is to set up (I have full bridge) just nearby transformer? \$\endgroup\$
    – xakepp35
    Commented Jan 20, 2019 at 15:15
  • 1
    \$\begingroup\$ Larger capacitor is of no worry, they said. So I thought for inductor. To "localize a field" one day i put a whole transformer secondary (thinking it is a great inductor). Tda2050 welded inside, and current found its path through its case to the tab, melting it in bright orange color dot. \$\endgroup\$
    – xakepp35
    Commented Jan 20, 2019 at 15:28
  • \$\begingroup\$ Suppose your power amplifier uses 10amps, or 100 amps peak thru the rectifiers, and the diodes turn on in 1uSecond. We'll assume the power transformer permits that speed of turnon. Thus the dI/dT is 100amp/1uS or 10^+8 amps per second. Assume your power cable is placed 10cm from the vinyl-record playback turntable, with 1milliVolt signal at strong audio output. We want 80dB SNR, thus need 100 nanoVolts interference level, or smaller. Assume the vinyl RIAA preamp has 1cm by 10cm PCB loop. Interference will be computed from Vinduce (magnetic field) is MUo * MUr * Area/Distance * dI/dT. \$\endgroup\$ Commented Jan 20, 2019 at 21:31
  • \$\begingroup\$ That formula simplifies, using MUo as 4*pi*1e-7, MUr as air or copper = 1, to Vinduce = 2e-7 * Area/Distance * dI/dT. We have all the numbers. What is Vinduce? 2e-7 * (Area = 1cm * 10cm) / (Distance = 10cm) * dI/dT = 1e+8. We now have 2e-7 H/meter * 1cm * 1e+8 = 2e-9 * 1e+8 = 0.2 volts. Yet our budget is 100 nanoVolts. People over in diyAudio.com develop not only 1milliVolt vinyl-playback systems (moving magnet) but 100 uV systems for moving-coils. Their trash-budget is ~~ a nanoVolt of Hum, etc. How to do this? They have 1 meter spacing between XFORM/Diodes and the low-noise circuits. \$\endgroup\$ Commented Jan 20, 2019 at 21:37
  • \$\begingroup\$ Another detail is to SLOW DOWN the turnon time of the diodes, with small resistors or small inductors in series, People often place 1nF capacitors in parallel with each diode, for the same reason. \$\endgroup\$ Commented Jan 20, 2019 at 21:41

It does not really matter for your use case.

I'd still go with putting the rectifier behind the cable, because it leaves you with better maintenance options, e.g. if you want to insert a switch or relay into that cable later.

  • \$\begingroup\$ So will I do. Just curious about possible diffirences. Is it significantly worse to switch a DC? \$\endgroup\$
    – xakepp35
    Commented Jan 20, 2019 at 4:47
  • 1
    \$\begingroup\$ @xakepp35, yes, switching DC is somewhat harder. \$\endgroup\$ Commented Jan 20, 2019 at 5:09
  • 3
    \$\begingroup\$ And there's always the possibility of getting the polarity wrong when repairing or replacing wiring with DC. Explaining a repair or replacement to someone else is also harder with DC. AC is just less worry. \$\endgroup\$
    – jonk
    Commented Jan 20, 2019 at 5:51
  • 1
    \$\begingroup\$ DC systems are fine - just pay attention : having trained as an auto-electrician I have seen some excellent errors and consequences by people who start with "but I thought it worked like this... "... \$\endgroup\$
    – Solar Mike
    Commented Jan 20, 2019 at 6:16
  • \$\begingroup\$ @SolarMike swapping polarity for tda2050 or 7294 power legs, as well as for large electrolytic caps are my favorite stuff to do . I love fireworks :-) \$\endgroup\$
    – xakepp35
    Commented Jan 20, 2019 at 15:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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