[sdiy] zener regulators and negative supply voltage.

Mon Dec 23 23:24:22 CET 2002

Hmmm... I'll take exception to that  ;^P

Zener (shunt) regulators DO have a very useful place in our
designs.

No shunt regulator is good for high power. The series resistor
would get way too hot...and there is a lot of power wasted at
all times.

OTOH... there is NO linear regulator as fast as a zener. Zeners
will never oscillate, will not overshoot.  Zeners do not need
some minimum load, or capacitance... to be stable.

I've used zeners twice this week. One was to make a reference supply
for the Frequency and Pulse Width pots on a new modular...er...
module (VCO).  Using the power supply rails would give the chance of
noise and frequency instability.  A simple one resistor, one zener,
one cap solution makes a decoupled, ripple free solution. I allowed
about 5mA for the zener and the pots together. 

Another was a overvoltage clamp for a footswitch input. WHO knows what
someone might stick in there... using clamp diodes might allow an external voltage to backfeed my 5V supply and fry my micro (ouch)... so
I used a pull-up resistor (high enough value you can't push the supply up... a series limiting resistor... and a shunt 5V zener. The zener protects against voltages above 5V and negative voltages are clamped to 
0.7V at the same time (for free).

Get to know the zener. Zeners are our friends. Shunt regulators are
a very good solution for low power and guaranteed fast response.

H^) harry

Subject: Re: [sdiy] zener regulators and negative supply voltage.

>Well not really wrong, but your probably are making things hard for your
>self. Killing the ripple with a zener and a RC or LC filter no longer makes
>sense for most applications. Exceptions include high voltage supplies for
>tubes and exotic things like that.
>Did you look at linear regulator ICs like the LM317 or the 78XX and  79XX
>series?
>The 78xx and 79xx are for a fixed voltage and easiest to use.
>The xx indicates the voltage and the 78 is for positive and 79 for negative
>voltages.
>The 317 and cousins can output any voltage set by the ratio of two
>resistors.
>Jim Patchell has a good example for the 78/79 and in depth explanations of
>calculations on his site.
>http://www.silcom.com/~patchell/powersupply/powersupply.pdf
>http://www.silcom.com/~patchell/powersupply/powersupply.html
>There are zillions of other examples of this on the web too.
>However if your application does not drain a lot of current you may forget
>Jim's calculations.
>470uF or 1000uF for the input cap and 47uF or 100uF for the output cap
>should work in ok. Smaller caps may be ok too, but 470uF or 1000uF are no
>longer that large in size anyways.
>Don't forget the100n decoupling caps though they are important!
>Adding 22n to 47n caps in parallel with  your rectifier diodes (or bridge)
>will help kill the switching transients from the diodes.
>But this is a fine tuning detail and not needed.
>Note that the regulator ICs need a input voltage that is at least 2V higher
>than the output voltage to work!
>Also note that the current rating for the ICs if only valid if you add a
>cooling body to the IC.
>Without cooling most ICs can handle a little over half the rated current,
>depending on the input voltage.
>Just to be on the save side I usually use the 1A rated ICs even if I only
>need 300mA.
>HTH
>Theo
>From: Arun Bohm 
>> Yeah I am trying to build a power supply. Where am I going wrong? So far
>> I've set up a bridge and capped it. Just trying to get the ripple out.
>I've
>> always only worked with breadboards at the work bench so this will be my
>> first attempt at actual fabrication.
>> Thanks.
>>
>>
>>
>>
>>
>> >From: "Theo" 
>> >To: "Arun Bohm" , 
>> >Subject: Re: [sdiy] zener regulators and negative supply voltage.
>> >Date: Sun, 22 Dec 2002 23:11:39 +0100
>> >
>> >Maybe I get you wrong, but I smell something burning.
>> >What exactly do you want to build?
>> >If it is a power supply, you might be on the wrong track.
>> >
>> >Theo
>> >
>> >From: Arun Bohm 
>> >
>> >
>> > > I would just reverse the reverse bias for a negative supply voltage,
>> >right?
>> > > is it smart to use an inductor to elliminate ripple or would that
>create
>> > > transients at the power out.
>> > >
>>
>>
>> _________________________________________________________________
>> MSN 8 helps eliminate e-mail viruses. Get 3 months FREE*.
>>
>http://join.msn.com/?page=features/virus&xAPID=42&PS=47575&PI=7324&DI=7474&S
>U=
>>
>http://www.hotmail.msn.com/cgi-bin/getmsg&HL=1216hotmailtaglines_virusprotec
>tion_3mf
>>