Some noise measurements for LEDs and zener diodes

Jump to Latest
#201
Noise figures for paralleling op amps

Here is an excel file that I use for calculating noise of paralleling op amps.
The model is based on Analog Device's Application Note ( or TI's document, can't recall now).
I used 625 for Rs because it is my R-2R DAC's output impedance.
 
#203
B&W_arthur said:
bcarso, I considered ordering some 2sk3557 a few days ago. But I did not.
I know that bf862 has lower noise but I am afraid that its high Tf (750MHz)
is susceptible to RF oscillation. Can you let me know more about the circuit
in which you use bf862?
Click to expand...
Unfortunately it is for a client and proprietary.

Early breadboards did oscillate too, even after some ferrite beads in the gate lead, but some of that I believe was due to parasitic inductances owing to the SOT-23 adapter. I've since mounted them on leads coming up from 0.1" grid plated-through pad-per-hole material and cut down the parasitics, and found I no longer needed the ferrites.
 
#204
Hi Christer... or someone who can explain..
i need to build a regulated psu.
the zener strings need to be 30V-35V for negative rail and positive rail,
if i replace those zener with red led with 1.85V each led,
that will need 18x red led each rails.

if i use 18x red led with 1.85v
compared with 3x 10v zeners.

which one will results the better.

thanks.
 

Attachments

  • zener.png
    zener.png
    15 KB · Views: 1,214
#205
As you have fairly constant current in your Vref chain and not current draw into the mosfet you could just:

1) Have a 30 - 36V zener with a 10R on top followed by the electrolytic.

or alternatively

2) Use a current source to drive into a wirewound resistor to give you a very quiet Vref with the the electrolytic in parallel

I tend to use option 2 mostly

mike
 
#207
wengtech said:
Hi Christer... or someone who can explain..
i need to build a regulated psu.
the zener strings need to be 30V-35V for negative rail and positive rail,
if i replace those zener with red led with 1.85V each led,
that will need 18x red led each rails.

if i use 18x red led with 1.85v
compared with 3x 10v zeners.

which one will results the better.

thanks.
Click to expand...
Q2 is upside down.
 
#209
for 16x LED strings,
according to this calculator :
======================
red LED that measures 0.27 uV,
we get that the noise voltage of the LED itself ought to be:

sqrt(0.27^2 - 0.19^2) = 0.19 uV

A chain of 16x LEDs should thus give a total noise voltage of

n = sqrt(0.19^2 x 16) = 0.76 uV

is that a big noise? is that audible?

thanks.
 
#210
wengtech said:
for 16x LED strings,
according to this calculator :
======================
red LED that measures 0.27 uV,
we get that the noise voltage of the LED itself ought to be:

sqrt(0.27^2 - 0.19^2) = 0.19 uV

A chain of 16x LEDs should thus give a total noise voltage of

n = sqrt(0.19^2 x 16) = 0.76 uV

is that a big noise? is that audible?

thanks.
Click to expand...

I thin a 12V zener has got much lower noise than using 16 x LEDs
12V zener reference requires about 5 LEDs which is definitely greater than 0.4uV.

Each Red LED on an avg.

EL202HD (red):
#1 @ 1mA: 0.31 0.32 0.31 0.31 0.32 uV
#1 @ 5mA: 0.26 0.26 0.27 0.27 0.27 uV
#1 @ 20mA: 0.39 0.36 0.37 0.36 0.37 uV
#2 @ 1mA: 0.39 0.37 0.38 0.38 0.35 uV (Vf = 1.82 V)
#2 @ 5mA: 0.32 0.30 0.30 0.30 0.31 uV (Vf = 1.89 V)
#2 @ 20mA: 0.41 0.40 0.41 0.41 0.46 uV (Vf = 2.09 V)

see the Zeners especially from Vishay:

BZX55/C12 (0.5W 12V):
#1 @ 1mA: 0.35 0.37 0.37 0.39 0.39 uV
#1 @ 5mA: 0.30 0.28 0.28 0.28 0.30 uV
#1 @ 20mA: 0.24 0.25 0.25 0.26 0.25 uV
#2 @ 1mA: 0.32 0.33 0.32 0.33 0.32 uV (Vr = 11.32 V)
#2 @ 5mA: 0.26 0.26 0.27 0.32 0.26 uV (Vr = 11.37 V)
#2 @ 20mA: 0.25 0.26 0.28 0.24 0.30 uV (vr = 11.42 V)

for 12V you get about .28uV at best
so use one 12V zener from BZX55/C12 and then add leds for that for the nearest value. you probably require 2 more leds with zener.
 
#213
Gentlemen, in the context of this thread please have a look at:
Noise in voltage references

Some nice measurements of LED, Zener and integrated references made with a
spectrum analyzer and also normalized values for comparison (near bottom of page).

Note the dc voltage of a 2V7 Zener was less than 2 volts in his set up, and I am a little
bit in doubt about this point. However Dick Kleijer has no e-address on his home page,
so I could not write to him about this confusion.

For a useful comparison of different items in terms of noise and voltage please see
Morgan Jones' post 66 here (and bad "Zener noise" reputation again in question).
Some noise measurements for LEDs and zener diodes
 
#214
I have seen manufacturer data (but I missed what) showing less noise for the "gray" area (mixed avalanche-zener mode) than for the rest.
Remember that zener devices are variable and for different manufacturers probably the noise relations change.
Some time ago I made a simple shunt -13V supply using 12V zener and a Si BJT. I noted too much flicker noise when I measured the rail with FFT (sans load). I changed this zener for two 6V2 specimens and the flicker vaished, and even the HF part reduced a little.
So much for me. I prefer to test in actual circuit, since I use brands that I have at hand or exotic ones.

Normally most IC's references are far more noisier than a good selection of zener or LEDs but for consistency (and another things like absolute voltage accuracy, unimportant for a lot of audio projects) people choose ICs.
 
#215
In the home page referenced in post 213 Mr Kleijer tested only one Zener, but it had less noise than a red LED.

In the meantime he replied why this single Z2V7 specimen only had a voltage of about two volts in the test: it was specified for 80 mA current and used with less, so smaller voltage.

He also sent me some interesting curves from a diode reference book, these may indicate that a 15 V avalanche type apparently has the steepest operation (note the different scale on both diagrams).
 

Attachments

  • Z.div.jpg
    Z.div.jpg
    69.4 KB · Views: 701
#217
as_audio said:
In the home page referenced in post 213 Mr Kleijer tested only one Zener, but it had less noise than a red LED.

In the meantime he replied why this single Z2V7 specimen only had a voltage of about two volts in the test: it was specified for 80 mA current and used with less, so smaller voltage.

He also sent me some interesting curves from a diode reference book, these may indicate that a 15 V avalanche type apparently has the steepest operation (note the different scale on both diagrams).
Click to expand...
Very interesting!
With my diodes I have at hand, some time ago, I measured the slope with a curve tracer, and for low voltage true zener ones the slope always are curved like in the ones show in your post, and with high dynamic resistance.
But the one with lowest dynamic resistance I measured are a old 6V2/1W from Philco with nice round case (like some old Fairchild). All the higher voltage ones measured higher than it.
About tempco, some literature about it mention the "mixed" ones (5V6-6V2) having almost zero tempo. Having built some simple LDO PSU using 5V6 zener in series with a Ge Vbe (the transistor for first gain stage in CE configuration) resulted in a basically zero tempo in practice, and almost zero tempo for a Si Vbe.
Again the tempco will varies with manufacturer I suppose.
For curiosity, I measured some old NOS SiC LEDs from ex-USSR and these have very curved slope with high resistance, and the "normal" old russian red LEDs (these with partial metal cased) have surprisingly low dynamic resistance, but the winner are always the old HP monolithic early red displays (vertical line like the 15V zener of your post; <1 ohm).
 
#218
Oops... bad spelling. Tempo here = "tempco" = thermal coefficient. I never quantized the values, only noted from experiences using zeners/diodes.

And thanks for all people measuring these things; maintain your work! Is good to have all data to compare.
 
You must log in or register to reply here.
Top Bottom