Zener clamping output of regulator

The small print says it doesn't regulate to within its specified tolerances on loads less than 10% and the 3.3v part can be 20% high.

If it is absolutely critical that parts never see a voltage above 3.3v even at very low current draw then you need a better regulator. A series resistor between its raw output and the zener may help.

--
Regards, 
Martin Brown

V

t not cool to put a zener at the output to clamp the voltage of the regulat or?

Totally *not* cool. Zeners need to operate at a certain current..(well at least over a small range.) I assume there too much current flowing through the zener and this puts it above the voltage 'spec'. Low voltage zeners h ave a very soft knee... look at the curve in the spec sheet you posted. (a nd observe it's logarithmic in current.)

George H.

George,

Sure zeners need a certain current, this little regulator can ouput a max o f 75mA, if it started to output over 3.3V I would assume the zener would tu rn on and start sinking some of that current and clamp at 3.3V, but why wou ld it let the voltage across the zener go higher? 75mA is not enough to mak e the zener go to 3.6V, I'm thinking the zener isn't turning on at all... b ut I don't see why not.

.3V

it not cool to put a zener at the output to clamp the voltage of the regul ator?

t least over a small range.) I assume there too much current flowing throu gh the zener and this puts it above the voltage 'spec'. Low voltage zeners have a very soft knee... look at the curve in the spec sheet you posted. (and observe it's logarithmic in current.)

Oh I didn't see that.. then maybe just replace the zener with a suitable resistor to ground. (to get some minimum current draw from the regulator.)

George H.

of 75mA, if it started to output over 3.3V I would assume the zener would turn on and start sinking some of that current and clamp at 3.3V, but why w ould it let the voltage across the zener go higher? 75mA is not enough to m ake the zener go to 3.6V, I'm thinking the zener isn't turning on at all... but I don't see why not.

Well try making an I-V plot of just the zener. That might help you. Go from maybe 1uA to 10mA. (or add a 300 ohm resistor to ground)

George H.

Those things are not regulators at ALL. They're kind of the opposite of a regulator, in fact.

No, it's not cool. If you need a regulated output, you can use a 5V module and regulate it down to 3.3V.

I've never seen such garbage.

3V

it not cool to put a zener at the output to clamp the voltage of the regula tor?

It's cool for those so-called zeners. Check out the spec sheet, IZT is 50uA ! Then further down the graphs for zener impedance vs current, it is almost 10K ohms at that current! It would only take a 1mA or so to elevate the 3.

3V to 3.6V. One thing for sure, absolutely WRONG part to use as a clamp.

Excellent, I feel like this is the answer but I'm having a little trouble f ollowing it

So, 50uA is the test current used to get the 3.3V nominal value right? I t hought you would usually run a zener with a few milliamps to get it to clam p...

How does this 10k impedance come into play? If it's 3.3V at 50uA, and then it gets another 1mA running through it, doesn't the impedance just come dow n since there's more current? I think the answer is here but I just don't see it

thanks!

3.3V

s it not cool to put a zener at the output to clamp the voltage of the regu lator?

uA! Then further down the graphs for zener impedance vs current, it is almo st 10K ohms at that current! It would only take a 1mA or so to elevate the

3.3V to 3.6V. One thing for sure, absolutely WRONG part to use as a clamp.

That's an unregulated converter (and unlimited; basically a two-transistor chopper, transformer and rectifier). As others have said, if you need better regulation, do something else. For example, use the 5V model (assuming you also have a 5V input somewhere), followed by an LDO.

Zeners are rather poor accuracy, particularly low voltage devices (under

7V), which have large dynamic resistances.

If you also need insurance that voltage does not exceed 3.3V plus change (3.60, say), use a TL431 to trigger an SCR. You will also need a current limit; an LDO's built in current limit is very crude, and the isolation module will smoke rapidly if overloaded (the regulated flyback type isolators are current-limited, but not available in small packages). At this point, you might be better off building your own isolator or regulator.

Tim

--
Deep Friar: a very philosophical monk. 
Website: http://seventransistorlabs.com 

"panfilero"  wrote in message  
news:8ca28e60-6b89-4fc9-9b2d-5add67574641@googlegroups.com... 
> Hello, 
> 
> If I had a little isolated 3.3V voltage regulator like this guy 
> 
> http://www.recom-power.com/pdf/Econoline/RM.pdf 
> 
> and I put a 3.3V zener right across the output to clamp the output to  
> 3.3V 
> 
> like this zener 
> 
> http://www.diodes.com/datasheets/ds30410.pdf 
> 
> but I saw 3.6V across my zener.... any thoughts on what is going on? is  
> it not cool to put a zener at the output to clamp the voltage of the  
> regulator? 
> 
> much thanks!

So what I've gathered from this (I'm more interested in understanding why the zener is at 3.6V than in getting a better 3.3V clamp on my circuit)

is that, zeners with voltages under 5V have high impedance associated with them which cause them to fluctuate wildly around their test point. so, since my 3.3V zener has a Izt = 50uA, and according to the datasheet that gives me an impedance of 10k

so if I'm seeing 3.6V, that means my zener may be seeing around 30uA above 50uA

i got that from

(3.6V-3.3V) = (50uA + I) x 10k

I = 30uA

sound right?

If you want a hard clamp, use a TL431. ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

following it

thought you would usually run a zener with a few milliamps to get it to cl amp...

n it gets another 1mA running through it, doesn't the impedance just come d own since there's more current? I think the answer is here but I just don' t see it

Figure 4 pertains to your DDZ9684 3.3V nominal. Although they don't show it specifically, you can guesstimate from the 9678 and 9681 curves that appro ximately a doubling over the IZT test current gets the 300mV additional VZ above nominal. So about 100uA will produce 3.6V in the 3.3V. This product l ine acts like ideal zeners at about Vz of 8V and above. You can't use the D DZ9684 as a clamp in this application.

The TL431 is quite an impressive little part really. I rediscovered it recently. Was looking for a cheap, lower noise reference than the LDO I made the mistake of using (few mV p-p) and the LM4431 I have used before (couple hundred uV p-p).

The TL431 noise is unspecified on the datasheet I had, but I measured it as 45uV pk-pk in 1 second, not bad at all. And 5 cents!

--

John Devereux

My favorite junk-box parts: LM339, MC34084 and TL431.

I've even used a TL431 as an OpAmp ;-) ...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Your MC34084 looks to be a bit obselete I'm afraid.

It's a funny little thing, sort of precision zener, comparator, transistor-with-2.5V vbe. Oscillator! :(

--

John Devereux

Probably. But I have about 100 in my junk box ;-)

...Jim Thompson

--
| James E.Thompson                                 |    mens     | 
| Analog Innovations                               |     et      | 
| Analog/Mixed-Signal ASIC's and Discrete Systems  |    manus    | 
| San Tan Valley, AZ 85142   Skype: Contacts Only  |             | 
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  | 
| E-mail Icon at http://www.analog-innovations.com |    1962     | 
              
I love to cook with wine.     Sometimes I even put it in the food.

Not probably, they dropped it about 15 years ago. But this is a good substitute:

the zener is at 3.6V than in getting a better 3.3V clamp on my circuit)

h them which cause them to fluctuate wildly around their test point. so, s ince my 3.3V zener has a Izt = 50uA, and according to the datasheet that gives me an impedance of 10k

e 50uA

No! Let me suggest this again. Take your zener, some resistors, a power s upply and your voltmeter. Measure the voltage drop across the zener at a b unch of different currents (say separated by a decade, 1uA, 10uA, 100uA...u p to 10mA. maybe you'll want to fill in the points and do 3uA, 30uA etc.. t oo.) Now plot the data (you may need a semi-log plot.) Now tell me at what voltage the zener 'turns on'.

(If you've never done it you may want to do the same thing for a 'regular' diode. (and please don't tell me the regular diode turns on at 0.6 Volts))

The dynamical resistance of the zener is the local slope of the V-I curve a t that point.

George (a man who loves data) Herold

Yeah but we were talking jellybeans I think; that is $2.50/1k.

I like AD8655, it's my sort of "blameless" opamp; fast, low noise, low bias, high drive, rail to rail I/O. 5V though.

Not quite jellybean but not too bad, $0.71/1k.

--

John Devereux