zener decoupling?

[Peter]

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

--- Peter ---

 

[James Meyer]

On Sun, 16 May 2004 08:46:09 -0400, "Peter" <peterm@wanad00.com> posted this:

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

--- Peter ---

Take a look at the slope of the zener's characteristic curve. It isn't
perfectly flat. That means it has an equivalent output impedance that will
reduce the voltage when loaded and change the voltage with changes in the
applied Vcc.

Then there is the broadband noise associated with the zener mechanism
its self.

I would have to say, yes, the bypass capacitor is needed. Not for the
same reasons as for an active regulator, but needed nonetheless.

Jim

 

[CFoley1064]

Subject: zener decoupling?
From: "Peter" peterm@wanad00.com
Date: 5/16/2004 7:46 AM Central Standard Time
Message-id: <moJpc.680$1f2.17709@weber.videotron.net

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

--- Peter ---

Hi, Peter. A typical 1N4733 (5.1V 1W zener) has a zener impedance of 7 ohms at
a test current of 49 mA. At 20 mA it's not specified, but it's going to be
quite a bit greater. Let's say it's 25 ohms. Now let's say you've got an
average of 20mA zener current, but current spikes mean your current varies from
5mA to 30mA. With a 25 ohm zener resistance, that means your output voltage
will vary significantly. Now, this is a "first cut" approximation based on
extrapolating a typical zener test current across a wide range of currents, but
you get the idea here. If you're talking about a need for a +/- 10% regulation
power supply, you're already pretty close here. A cap would be good practice,
I think. One of the biggest newbie mistakes is cutting things close on the
power supply, and then wondering why it doesn't always work right. Best to do
things well to begin with, and then if something goes wrong, you won't have to
waste time wondering. If you're powering a processor, you've got a load which
is actually a series of current pulses which happen with every processor clock
pulse. You shouldn't use a zener to power a processor, but if you do, or if
you bought into something that does (I'd return it for a full refund just on
principle -- the person who put it together obviously needs to get it
together), try to get an electrolytic with good high frequency characteristics,
or use a tantalum cap. And don't forget to put a good high frequency cap right
next to the PIC IC power pin, too.

Good luck
Chris

 

[Reg Edwards]

I would have to say, yes, the bypass capacitor is needed. Not for the
same reasons as for an active regulator, but needed nonetheless.

============================

Hint. If a zener is used as a precision voltage reference the noise
decoupling capacitor should have the highest possible insulation resistance,
eg., polyethylene, polystyrene.

 

[John Larkin]

On Sun, 16 May 2004 08:46:09 -0400, "Peter" <peterm@wanad00.com>
wrote:

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

--- Peter ---

The zener will be quite stable without a cap, but there's still only
20 mA available. The PIC, like any CMOS gadget, can pull sizable
transient current spikes, and there's nothing to supply that current
in a shunt zener regulator. So put a bypass cap (0.1 uF, vaguely) near
the PIC chip.

But you might get away without one. Scope the PIC Vcc with a fast
scope and see how nasty any spikes might be.

And zeners are even more venerable than 7805s!

John

 

[Spehro Pefhany]

On Sun, 16 May 2004 08:46:09 -0400, the renowned "Peter"
<peterm@wanad00.com> wrote:

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output

No, the 7805 does not need a capacitor on the output to be stable.
However the high-frequency output impedance may be fairly high if the
output does not have a load.

and also, I think, to
decouple current spikes from the load.

That's possible.

But is this capacitor really needed
with a zener regulator?

If you put it near the micro and assuming a 20mA shunt current, yes it
will work, at least at moderate clock frequencies.

As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

The DC regulation with a 5V1 zener will not be very impressive, and
spikes due to capacitive loading will not help. If your micro will
operate at 3V it will probably be okay with fairly large
negative-going spikes on the supply, given a fairly large minimum
zener current (say 10mA). I have not actually measured the spikes,
this is just engineering feel from how these things behave when
(mis)used in similar circumstances.

That said, it would have be very unusual circumstances before I'd go
to the time and trouble of evaluating doing this in a product-
capacitors are VERY cheap, and evaluation and characterization is not
(and trouble is MUCH more expensive again).

I have designed a product with no capacitor on a 7805 ouput (feeding a
big slow CMOS mixed signal chip with a fairly constant drain and a
slow micropower CMOS precision op-amp. Made many thousands of them.
But I did put a spot for one on the (punched single-sided) board. ;-)

Best regards,
Spehro Pefhany
--
"it's the network..." "The Journey is the reward"
speff@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com

 

[Peter]

Reading all your replies convinced me the cap is still needed or, at least,
can be seen as a good insurance against weird and/or nasty problems from
current spikes.

Thanks for your quick replies!
--- Peter ---

"CFoley1064" <cfoley1064@aol.com> a écrit dans le message de
news:20040516105128.10191.00001577@mb-m14.aol.com...

Subject: zener decoupling?
From: "Peter" peterm@wanad00.com
Date: 5/16/2004 7:46 AM Central Standard Time
Message-id: <moJpc.680$1f2.17709@weber.videotron.net

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt
regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really
needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should
be
fast enough to deal with current spikes. I know it may sound silly to
ask,
just one small cheap cap to add as it surely won't degrade the
regulation.

But... is the cap really needed with a zener shunt regulator?

--- Peter ---

Hi, Peter. A typical 1N4733 (5.1V 1W zener) has a zener impedance of 7
ohms at
a test current of 49 mA. At 20 mA it's not specified, but it's going to
be
quite a bit greater. Let's say it's 25 ohms. Now let's say you've got an
average of 20mA zener current, but current spikes mean your current varies
from
5mA to 30mA. With a 25 ohm zener resistance, that means your output
voltage
will vary significantly. Now, this is a "first cut" approximation based
on
extrapolating a typical zener test current across a wide range of
currents, but
you get the idea here. If you're talking about a need for a +/- 10%
regulation
power supply, you're already pretty close here. A cap would be good
practice,
I think. One of the biggest newbie mistakes is cutting things close on
the
power supply, and then wondering why it doesn't always work right. Best
to do
things well to begin with, and then if something goes wrong, you won't
have to
waste time wondering. If you're powering a processor, you've got a load
which
is actually a series of current pulses which happen with every processor
clock
pulse. You shouldn't use a zener to power a processor, but if you do, or
if
you bought into something that does (I'd return it for a full refund just
on
principle -- the person who put it together obviously needs to get it
together), try to get an electrolytic with good high frequency
characteristics,
or use a tantalum cap. And don't forget to put a good high frequency cap
right
next to the PIC IC power pin, too.

Good luck
Chris

 

[John Fields]

On Sun, 16 May 2004 08:46:09 -0400, "Peter" <peterm@wanad00.com>
wrote:

Hi every one!

I have a very simple PIC circuit supplied with a 5.1V zener shunt regulator
(20mA shunt current). No output pin are loaded more than few mA.
Now if I were using the venerable 7805, a small cap at the output would
usually be needed to stabilize the regulator output and also, I think, to
decouple current spikes from the load. But is this capacitor really needed
with a zener regulator?
As far as I know, the zener does not need a stabilizer cap, and it should be
fast enough to deal with current spikes. I know it may sound silly to ask,
just one small cheap cap to add as it surely won't degrade the regulation.

But... is the cap really needed with a zener shunt regulator?

---
The short answer is, "No, and neither is it needed with a 7805."

The long answer is... well, "How far into it do you want to get?"

--
John Fields