> I've used zeners and regular diodes in the past to raise the ground
reference for a linear voltage regulator.
> When will this come back to bite me in the backside?
> Needed a 9v regulator but only had a 5v so used a zener? or a couple
diodes to raise the ground a couple volts.
> Still working but wondered how bad a practice this is?

SUMMARY:

This is an entirely viable method BUT is not without its shortcomings.

Where possible, stocking up on eg LM317s and a few resistors will allow you
to make a much better regulator which is readily settable to at the voltage
of your choice.

DETAILS:

1.    If you use diodes or a zener then the regulation will be made worse by
the soft voltage /current knee on the zener or diode. Temperature regulation
will also probably be worse. Regulation substantially inferior to that from
a modern IC regulator can be expected. Often this will still be good enough.

2.    Older regulators like eg 7805/LM340 powered their internal circuitry
by current passing in through Vin and out through the ground lead. This
current varies in an ill defined manner and is passed by the diodes/zener
and adds to the poor regulation in 1. above. The impedance of the zener is
however relatively low and this MAY be superior to the more common method of
using a voltage divider - see 3. following.

3.    The more usual; method of increasing the Vout of a fixed regulator is
to provide a resistor R1 from Vout to 'ground" pin and another resistor R2
"ground" pin to true ground. This voltage divider needs to be "stiff" ie the
current provided from Vout to ground via the divider needs to be largish
compared top the current in R2 compared to the current in R2 due to internal
regulator electronics as in 1. above.

4.    There may be regulator protection issues (reverse voltage on shut down
etc) which may be made worse by this arrangement and which need to be
examined on a case by base basis.

Some more modern regulators (some now several decades old :-) ) power
themselves via the Vin to Vout voltage drop and send only a "reference"
current down what was the ground pin. As this current is reasonably
constant, the voltage divider method above can be used with reasonable
success.

This is the principle which  the very popular LM317 series regulators use.
They are designed as a low voltage regulator (1.25v?) which is meant to be
boosted up to a higher voltage of choice by the use of 2 resistors as above.
this method is liable to be superior to the diode or zener method and
cheaper. There is now no soft zener knee to be wary of and the tempco of the
resistors is liable to be superior to that of the zener (unless you have
fortuitously chosen a value which has approx zero tempco)(which only happens
at a specific voltage).



            Russell McMahon

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu