Do low noise zeners exist?

[Samuel Groner]

LM329 quotes 7 μV wideband noise figure. Since when was in order of magnitude or two not a significant improvement?

Two orders of magnitude? Check your math or look at figure 14 of the datasheet! The LM329 is 75 nV/rtHz (7.5 µV in a 10 kHz bandwidth), so either 2x better or 2x worse than the discrete Zener, depending on current.

If you find a temperature stable voltage reference which has two orders of magnitude lower noise than these figures (without filtering and at reasonable cost/complexity), make sure to get a patent before posting here...

Samuel

 

[Kingston]

Check your math or look at figure 14 of the datasheet!

I see. The max number isn't awfully convincing either.

 

[Samuel Groner]

The max number isn't awfully convincing either.

Indeed! Probably more a sign that they're not testing this figure than that actual specimen are that bad. Bot not guaranteed is not guaranteed...

Samuel

 

[tv]

If you want "low noise", you can try a BJT VBE multiplyer with the CB bypassed with a suitably large cap. For a tru-hole circuit, you could cook a SMT "thingy" that would fit in place of your Zener, imho, BC817-40 is your friend.

Caveat: Temp. stability not good, and more importantly, it will have a very "soft knee" wrt. voltage regulation. Which could be somewhat improved if you lift the BJT emitter with a diode junction or two, or a small SMT LED, but hey, we're already overdoing it, right?

 

[JohnRoberts]

The max number isn't awfully convincing either.

Indeed! Probably more a sign that they're not testing this figure than that actual specimen are that bad. Bot not guaranteed is not guaranteed...

Samuel

I had one major IC company refuse to even specify nominal noise levels for their 3 terminal voltage regulator. I only asked because their regulators were so noisy it was showing up in my audio path (long story). The other IC brands were quiet, so I black balled the outlier.

JR

 

[ricardo]

Yes, that is indeed a simple solution to keep the zener noise down.
The only problem is where to place the extra components on the existing PCB... (Space is limited!)

It's only ONE resistor.  You can move the cap across the Zener to after the resistor.

But if you wan't a really LN component to replace the Zener, try a resistor so that the voltage across it is roughly the same as as with the Zener.  I've done this on a commercial device in Jurassic times and it works well.

As long as your P48V preamp meets IEC standards 48V +/-4V until 10mA, there is no problem if that point isn't perfectly regulated.  It will vary with Idss of the FET but for a DIY one off, no big deal.

 

[RuudNL]

Yes, I thought of that myself. (To replace the zener with a resistor.)
The main goal is to reduce the voltage for the FET stage.
The regulation is not very important.
A voltage change of 0.2 V (or so) won't cause the FET to operate out of its workpoint.
But at the moment I am happy with the three 2.7 V zeners in series.
(I've put them zig-zag in a short piece of shrink tube and they fit on the original PCB.)

 

[tv]

Two small TH 1.5mm (1.6?) white leds would solve your problem even in a crowded circuit, without any mechanical acrobatics, nice and sturdy. (IMHO)

 

[RuudNL]

Yes, you are right! I had 2.7 V zeners at hand, but no white LEDs.
I did some tests with yellow LEDs and found that they are really low noise!
But I will keep the white LED solution in mind next time.

 

[tv]

I can't draw a picture ATM, but you could glue them back-to-back (easy if you obtain the small rectangular ones), so you would in effect "get" a comparable "lead spacing" to, say, a "box" poly cap.

Now, you'd have to solder the "upper" leads together (and if needed, wrap in a shrink-tubing), so it would look remotely similar to a - noname poly cap.

But, you could be smart and take also Ricardos advice BUT "mod it" - use a low-value SMT resistor (22-47 Ohm) an solder it across the upper LED leads, the ones which would otherwise get soldered together (i.e. shorted).

Super-glue it and shrink-wrap it, ET VOILA, super quiet photo-fancy ghetto stabilizer, DIY styley.

Mannnnng..

 

[tv]

Like this..

 

[12afael]

This topic is very old but very interesting.  I found this article with more information about noise of different ways of regulation.
http://www.dicks-website.eu/noise_in_voltage_references/part1.html

better to keep everything in one place.

BR
Rafael

 

[RuudNL]

Interesting!

 

[jensenmann]

Rafael, your link gave me an afternoon full of reading. Great resource.