variable resistances (Re: [sdiy] Multi-Input filters)

[René Schmitz]

Hi everyone,

At 21:18 12.02.03 +0100, ASSI wrote:
>On Wednesday 12 February 2003 16:12, Roman wrote:
>> there's even a parameter specified for zeners, called dynamic
>> resistance, that says what resistor would be connected to ideal
>> brickwall zener to have response like real zener. And it changes with
>> current. Personally I'd pick small zeners from 2V7 - 3V6 range. They
>> have smoother knee, and should be easier to use.
>
>Zener diodes (or any junction in breakdown mode) are really noisy (so 
>much that they're some of the best noise generators you have around). 
>If you don't need a large voltage drop on them, run from them as fast 
>as you can and use forward biased diodes instead. If it's just the low 
>voltages you mention, you may try one or two LED instead. I don't 
>really know what their noise properties are, but at least they should 
>not be worse than Zener diodes.

You're right about the noise of course. And LEDs make a pretty good 
voltage reference, especially if you run them from a current source.

Low value Zeners aren't really zehners. The lowest available "zeners" are 
just a bunch of forward biassed diodes really. Thats also why the knee 
is softer. 

Then there are two mechanisms in "real" zeners, one is the zener effect the 
other is the avalanche effect. They have tempcos of opposite sign, and there 
is a particular voltage region where the tempco is very low. That happens 
arround 5 to 6V. 

Cheers,
 René




-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159