1
\$\begingroup\$

I've got power supply (transformer based) 9V powering modelling guitar effect processor. That noise gets amplified and it hurts.

Humming noise is 100Hz rectified mains. Processor takes 300mA 9V (500mA rated). Unfortunately I cannot tell ripple voltage (don't have an oscilloscope).

Is there any equation for calculating capacitance needed to eliminate the noise?

When I was studying step-up/down switching regulator IC datasheets, I saw the formula: coef**current*time_on/Ripple_voltage. But is contains coefficient, which is different from datasheet to datasheet.

\$\endgroup\$
  • \$\begingroup\$ No because there isn't an obvious criterion for how much reduction you need. So try 1000uF, if that helps but not enough, try 10000 uF. \$\endgroup\$ – Brian Drummond Dec 2 '17 at 23:48
  • \$\begingroup\$ Get a regulated PSU instead. It will eliminate the problem. If you can find an unregulated unit that gives out > 12 V at full load you could easily add a regulator. I think the 78xx family had a 7809 which, with a couple of decoupling capacitors would be all you need. \$\endgroup\$ – Transistor Dec 3 '17 at 0:06
  • \$\begingroup\$ @Transistor, the power supply I'm talking about consists of transformer(21.6V AC out, no load)->full_bridge_rectifier-> elec.cap->78L09->22nF.ceramic_cap. And it still making humming noise \$\endgroup\$ – Qeeet Dec 3 '17 at 0:13
2
\$\begingroup\$

Any cap above 670uF would do their job, like 1000uF, and also remember to use always a higher rated voltage, like 16V in this case.

I calculated it in this way: With 0.3A and 9V power consumption we can treat your device like an equivalent resistance of 30 Ohms. Like in RC low-pass filters, we could determine their filtering frequency using 4 times tau to get away any 50 Hz or other harmonic of your mains frequency.

This is 4*tau=4*R*C -> 0.08=4*30*C -> C = 667uF

I always have in my toolbox a pair of handy 2200 uF 16V in case I need it :)

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ isn't formula f=1/(2piRC)? I so, I got 110uF \$\endgroup\$ – Qeeet Dec 3 '17 at 0:43
  • \$\begingroup\$ No @Queet, that might give you 50% ripple \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Dec 3 '17 at 1:25
2
\$\begingroup\$

The power supply I'm talking about consists of transformer->full_bridge_rectifier-> elec.cap->78L09->22nF.ceramic_cap. And it still making humming noise

The 7809 will require 3 V headroom so check that you've got 12 V DC on the electrolytic. You could - depending on your multimeter type and quality - get a rough idea of the ripple by measuring the AC voltage across the capacitor.

According to the 78L09 datasheet it is rated for only 100 mA. You're load is up to 300 mA. That might be the problem.

|improve this answer|||||
\$\endgroup\$
  • \$\begingroup\$ I forgot to mention, that power supply with 78L09 I use for small pedal and it is noisy. I have a transformer capable of delivering 800mA 12V, I want to use it to make a power supply for my processor, which require 300mA. Sorry, maybe I oversimplified the question, which created some misunderstandings \$\endgroup\$ – Qeeet Dec 3 '17 at 0:29
  • 2
    \$\begingroup\$ 78Lxx stands for low current. The 78xx series is rated at 1 A but watch out for temperature limiting. Yes, a garbled question can waste a lot of people's time. It's worth spending a few moments proof-reading and seeing how a post could be misinterpreted before hitting the submit button. \$\endgroup\$ – Transistor Dec 3 '17 at 0:35
2
\$\begingroup\$

Formula for a full wave rectified supply is that peak-to-peak ripple

\$V_R=\frac {I_{OUT}}{2f_{mains}C}\$

The regulator should take care of most of the ripple unless you drop below about 11-12V input to the regulator. If you have a 12VAC transformer it will deliver about 15.5V peak under load, with nominal line voltage. If you allow line to be 10% low, that's 13.8 or so your ripple should be less than about 2V p-p, so 1000uF should be fine. Double that wouldn't hurt.

If the hum is kind of raspy rather than pure, the diodes in the bridge may be contributing. Put 10nF 100V across each pair of leads (4 caps).

|improve this answer|||||
\$\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.