Rail Splitter PSU Circuit...

[Damian Mitchell]

Hi...

I've built one of these http://www.diyaudio.com/forums/solid-state/234665-virtual-ground-regulated-rail-splitter.html (Second Schematic on the page)...

To Power one of these:- http://www.ebay.co.uk/itm/291399023398?_trksid=p2057872.m2749.l2649&ssPageName=STRK:MEBIDX:IT (Needs +12, -12, +5)

So I've put it together (with a 78L05 on the positive rail), and a 2.4v Zener in place of the voltage reference - as I don't have the correct part here...

I'm using a 24V/0.7A Switched mode "wall wart" to supply the circuit.

The circuit works OK, but with no load connected it is drawing 300mA, and the LM317/LM337 chips are getting hot.

The Zener has just under 2.1V across it, which I guess may be the problem?

Has anyone got any ideas?

 

[Damian Mitchell]

Solved it... Changed the Zener for a 2.7V one

It now has 2.5V across it, and no load draw is 15mA

 

[(*steve*)]

I've built one of these

I read that with one eye while I was looking at your picture with the other. :-D

OK, then I read the rest...

Has anyone got any ideas?

There is a reason why an LM336X-2.5 is specified in this application, and that is because it is both a very accurate voltage voltage reference, and it is twice the 1.25V that appears between the ground and output pins of an LM317/LM337.

What you have also noticed is that the voltage across a zener diode is less than the rated voltage when only a small current flows through it. This is a feature of all zener diodes, but is especially significant for lower voltage zeners.

You would be better off with the voltage reference specified. What you will find is that your circuit will draw more and more current as the voltage across the zener falls, and the rail will be less stable if it rises.