[sdiy] Ferrite beads or 10ohms resistors in series with +/-15vdc ??

[Ian Fritz]

Just to widen the discussion a bit, cross-talk between modules can also 
come from ground loops.  The solution for this is to use a star-type power 
supply distribution.  My system has a central raw supply starred out to 
several cabinets that each have their own regulators.  Inside the cabinets 
power to each 19" panel is starred out. All this with fairly heavy 
wire.  For the modules associated with each panel I don't worry about 
ground loops since the lead lengths are so small.

The disadvantage of decoupling with RC's on each board is that it is 
inconvient to make the filter go down to the LFO range.  People often 
complain about bleedthrough of their LFOs.  I believe problems at such low 
frequencies are almost always from ground loops.

I think decoupling at the board level may help sometimes if high-frequency 
spikes are getting on the supply lines.  Either small resistors or beads 
are effective for this kind of noise.  I alway put them on boards I sell, 
because folks seem to think they are necessary.  But like David, most of my 
modules just have big caps where the power comes in and a few more bypass 
caps sprinkled here and there on the board.  A study I saw from one of the 
chip manufacturers showed that bypassing at every chip is not necessary.

Ian


At 10:27 AM 1/22/2010, Harry Bissell wrote:
>simple question easy answer
>
>The 10 ohm resistor forms an RC filter with whatever filter or decoupling
>capacitance there is. It also introduces a loss, as you noticed. The trade 
>of is bigger
>resistances work better at low frequencies (nearer to DC), but they lose more
>voltage. BIGGEST issue is that the drop is proportional to current draw of 
>the circuit.
>
>The Ferrite bead is almost "zero" ohms at DC, so there are usually no 
>discernible losses
>in voltage. At high frequencies (MHz) they are far superior as they are 
>essentially 'lossy'
>inductors.
>
>At audio frequencies, the ferrite beads may or may not be effective. If I 
>was trying to stop
>audio from entering or leaving a board, the resistors will probably work 
>better.
>
>The right technique is to use the RC supply filters at the chip or 
>subcircuit level. Maybe
>a sensitive preamp chip alone could survive the voltage drop from even a 
>100 ohm resistor...
>and the RC will work muck better at the lower frequencies.
>
>Another good technique is to use onboard regulation right where you need 
>the low noise supply.
>Zener diode shunt references can be very effective...
>
>H^) harry
>
>
>
>----- Original Message -----
>From: Jean-Pierre Desrochers <jpdesroc at oricom.ca>
>To: synth-diy at dropmix.xs4all.nl
>Sent: Fri, 22 Jan 2010 11:48:03 -0500 (EST)
>Subject: [sdiy] Ferrite beads or 10ohms resistors in series with +/-15vdc ??
>
>
>Hi list,
>
>Browsing a lot of DIY module schematics all around the net
>I noticed some builders use ferrite beads in series with their bipolar 
>supply lines
>coming to the PCboard. Some other use 1/4w 10ohms resistors the same way.
>That makes me wonder..
>I'm on the making of a MIDI2DAC 6 voices polyphonic module using +/-15vdc 
>and a separate 5vdc
>supplies for digital stuff. I started using 10ohms resistors in series 
>with my 15vdc lines
>but rapidly found that these dropped my actual 15vdc voltage too much..
>I'm using 12 X 12bits DAC's here, a douzen of opamps  and don't want to 
>disturb my 1v/oct outputs
>because of not stable supply voltages..
>So I changed the 2 resistors for 2 ferrite beads instead
>and now my 15vdc supply lines are precises but do I get
>a good noise barrier still compared to 10ohms resistors ?
>
>Your thoughts welcome.
>JP
>
>
>
>
>
>
>
>_______________________________________________
>Synth-diy mailing list
>Synth-diy at dropmix.xs4all.nl
>http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>
>--
>Harry Bissell & Nora Abdullah 4eva
>_______________________________________________
>Synth-diy mailing list
>Synth-diy at dropmix.xs4all.nl
>http://dropmix.xs4all.nl/mailman/listinfo/synth-diy