[balou]

Hi,
I have troubles understanding how the power rail isolation of the PPA exactly works. It's a JFET on each rail, which will limit the max current draw of each channel.

Will it limit current spikes? But I thought that current flowing is given by U = R * I, R being fixed, and U being the voltage from the power supply. So if the power supply is halfway stable, it shouldn't make that big of a difference. Or does it?

 

[balou]

bump, anyone?


and to ask yet another question... how exactly does the m3 power supply work?

Quote:

Capacitance multiplier for opamp power The power supply to the opamps is further isolated and stabilized with a complementary capacitance multiplier. A capacitor is kept charged by two JFET current sources at each side, whose capacitance is then multiplied by the DC current gain of the complementary pass transistors to provide very low noise, low impedance DC rails to the opamps, unperturbed by the MOSFET supply rails.

 

[AudioCats]

All the JFET's in parralle provide more current than the max needed current (the opamps don't need that much).

What was done there, correct me if I am wrong, is using JFETs as constant current source to fill the capacitors. The function of the rail spliter is to make sure the middle of the caps are at exactly the middle of the rail voltage. If the capacitors are full, the extra current not used by the opamp goes through the rail spliter to - rail (basically wasted).

Did I get it right?

 

[tangent]

First, forget Ohm's Law. That only applies to passive components.

Think of the JFETs in this configuration like a Slinky: as the voltage above it changes, it "stretches" to make up the difference, leaving the voltage below it stable.

As for "current spikes", there aren't any, to speak of. The op-amps have a constant quiescent current, and are running their outputs into a very high impedance buffer stage. Thus, their operating current is all but constant across its entire range of operation.

 

[balou]

thanks for the replies.
AudioCats, so the JFETs kind of provide a constant, fixed current flow through the cap?
but why should it have to go through the rail splitter? doesn't the current just flow from v-, cap-, cap+, v+ (or vice-versa, depending if you look at the electron flow or the other thingy I can't remember)

tangent: I had to first look up what a 'slinky' is

.
what I don't understand, how does a JFET 'stretch'? that sounds more like a voltage regulator than a current regulator

and concerning the current spikes: I thought about the current spikes induced by other components on the power supply, be it from the buffers, or noise from the power grid which went through the psu

 

[tangent]

Quote:

Originally Posted by balou /img/forum/go_quote.gif tangent: I had to first look up what a 'slinky' is .

Surely that toy is world-wide by now? Do they call them something else in Switzerland?

Quote:

how does a JFET 'stretch'? that sounds more like a voltage regulator than a current regulator

What do you believe voltage regulators are made of, if not transistors? You need to do some studying before you ask more questions like this.

Quote:

I thought about the current spikes induced by other components on the power supply, be it from the buffers, or noise from the power grid which went through the psu

Yes, definitely go crack open an electronics book, if you aren't even sure about the difference between current and voltage. Some suggestions here.