(3) dirty/clean ground again [sdiy]

harrybissell at prodigy.net harrybissell at prodigy.net
Tue Aug 20 22:06:08 CEST 2002

>At 11:40 PM -0400 08/19/02, Glen wrote:
>>What about the idea of using inductors in place of the resistors? Would
>>>that be even better?
>An inductor would prevent sudden changes in current which, based on what
>Harry and Rene have said, is responible for the sudden changes in voltage
>at the supply pins, but that current is what the op-amp wants in the first
>place.  So it's sort of like throwing the baby out with the bathwater.

The capacitors on the load side must supply all the transient current
needs... in fact the voltage drop of the resistors (or inductors)
must be allowed for.. i.e. you don't care about the drop.

>>I know that decoupling with a series resistor is a popular technique, and I
>>have used it in some circumstances, but it has one major drawback that must
>>be considered. In fact, some time ago I read an article that vehemently
>>insisted that this technique should never be used! The reason is that the
>>series resistance effective increases the output impedance of the power
>>supply regulator to the value of the resistor. So instead of your carefully
>>designed and constructed low-impedance power supply, you now have a 100Ohm
>>(or whatever) power source.

Yes, absolutely right. OTOH you now have a lowpass filter on the supply... so any higher frequency noises are attenuated. This can remove the annoying fast transients from (lets say) LED multiplex
switching, from being audible in a preamp.

The ferrite beads won't help this.

Signals with low slew rates are MUCH harder to couple capacitively
into other traces as well...

>>I've had good luck decoupling with ferrite beads. Odd that no one has
>>mentioned this possibility in this thread.
>Using inductors, like ferrite beads, is good at reducing RF interrence that
>might be on the power supply lines, but it should be followed by a large
>enough capacitor to provide for sudden demands in current.
>>In my last VCO project I put beads on the supply lines where they enter
>>the board and also from that point to several different sections of the
>>>circuit. This is the first VCO I have made where the switching transients
>>>don't couple back into the expo converter section.
>That might also be a good idea sub-circuits using noisy high-impedance
>devices like logic, but I don't know if it would work for noisy devices
>like LED's or timing caps that draw large amounts of current.  It might
>just slow them down a bit while smoothing out their effect on the power
>supply.  Otoh, I wouldn't be surprised if it created oscillations of its

The best way to deal with that is to separate clean and dirty POWER
as well as the grounds.

I just design the LED boards to get their power direct from the 
regulator... like a "star ground" this is "star power"

It all comes down to... "Can you hear it... do you care ???"

H^) harry

Powered by Wild On Web:
Free POP Mail Access - Access your E-mail from Anywhere in the World!

More information about the Synth-diy mailing list