Simpler PI...

[Aziz]

Emitter degeneration is beneficial for 1/f noise. It is a proven fact for oscillators and it can't hurt to have some, provided it does not ruin the noise performance. In your setup emitter is in a feedback path so the emitter impedance is actively lowered by feedback in case of a signal, but not so much for the error (1/f noise etc.)
Current noise is not a problem for low impedance sources, and also the source - a coil - has a finite resistance and hence noise. There is something called "optimum impedance" that occasionally pops here and there, and it is simply ein/iin (voltage noise/current noise) and your input impedance is ideally matched if it is near it.

Hi Davor,

yep, since we can measure and calculate e_n and i_n we can calculate the optimum signal source impedance (Ropt = e_n/i_n). Optimum impedance matching would ensure optimum signal source power transfer to the amplifier system.

But we would require a transformer, which is making the KISS-design more complex. Or we can make more turns count RX coils, which comes close to the optimum source resistance (->heavy coil).

The 1/f noise isn't much critical for frequencies above 1.5 kHz. The thermal noise often already dominates at this frequency.

But I wouldn't add a series resistor to match the impedance!!! That would be really stupid. There is a nice myth about noise figure and optimum source resistance (please ask Mrs. Google).

Cheers,
Aziz

PS: I have asked Mrs. Google. She responded with:
http://www.google.de/url?sa=t&rct=j&...62922401,d.Yms
(F*!!!!!, don't talk B.$. to me! Why are you keen on spying me b1tch!)

I had to edit the NSA Inc. link. And here is a nice pdf document (see "the noise figure myth" section)
www.electro.fisica.unlp.edu.ar/temas/pnolo/p1_AN-104.pdf

^sif

[Aziz]

PNP Version

Hi all,

I have built the PNP version of the ultra-low noise AC amplifier recently (see attachment). (For the sceptics: Yep, it's real and is working a treat.) It will be interesting to see, whether the PNP BJT's do really generate less noise.
Any bets? My betting office you know..

BTW, I have changed my measurement method. It is simple and efficient.
I have specified my RX coil + cabling resistance to maximum 10 Ohm. I'm measuring the average noise floor rise between my system noise floor level (ADC sound card and absent amplifier) to the equivalent 10 Ohm input resistor thermal noise noise floor level for the bandwidth of 1.5 kHz to 48 kHz. The noise floor level will be averaged from 1.5 kHz to 48 kHz. The lowest noise floor level rise for a given gain should be the best one. High rise -> high noise, low rise -> low noise.

Here are some noise floor level measurements for the NPN BJT's:


This is the schematics of the PNP version of the AC amplifier:


Measurements will be done soon. May the best PNP BJT win.

Cheers,
Aziz

[Aziz]

Some PNP BJT Measurements.

Hi out there!, but be quiet! We are approaching new limits with hyper-ultimate quietness.

This is the PNP BJT measurements of the noise floor rise:


I've forgotton to measure the 3x BC 337-40 (NPN):

3x BC 337-40 + NE5534A, no driver, RG: 470, 4.7, RC=470, IC=4.1 mA

-119,43

-118,08

0,02

1,35

You see, that the PNP type BC 327-40 is more quiet.

And if you combine both NPN and PNP BJT's, you can lower the noise more and can have a bit of tempco compensation. Note, that the DC offset voltage error is positive for NPN and negative for PNP.
This is the ultimate hyper-low noise AC amplifier: *LOL*


But DC offset voltage error doesn't bother me at all.

^sif,
master of quietness

[WM6]

Bravo Aziz!

Never give up!

[Davor]

Here is a pdf with collected scans from 1980 RE/P magazine by John Roberts on his solution to a low noise pre. There are many similarities with Cohen pre: http://www.circularscience.com/des_art.pdf

[Aziz]

Hey guys,

you will find nice LTspice files of the latest ultra-low-noise AC amplifiers here:
(post #1743)
http://www.geotech1.com/forums/showt...852#post186852

Cheers,
Aziz

[Aziz]

Hi all,

I've done two more measurements. WM6 has sent me two different BJT's for testing. Thanks for the transistors and the nice op-amp book.

Here are the results:
NPN:

3x 2SD882 (NEC) + NE5534A, no driver, RG: 470, 4.7, RC=470, IC=4.1 mA

-119,34

-118,13

0,02

1,21

WM6's BJT

PNP:

3x 2SB772 (NEC) + NE5534A, no driver, RG: 470, 4.7, RC=470, IC=4.1 mA

-119,33

-118,17

0,02

1,16

WM6's BJT

All results till now (zipped Excel file, at the bottom of the table, see the new measurements 1.5 kHz - 48kHz):
AC-Amplifier-Noise-Measurements-09a.zip

Well, the noise floor level increase is 1.16 dB for the PNP type. I think, the better ones should be preferred.
Cheers,
Aziz

[WM6]

Thank you Aziz. So till yet BD175 and BD176 preform the best.

[geoscash1]

Anyone else build this besides Waikiki Sweep ?, if so how well does it work and what coil specs are you using..
http://www.geotech1.com/forums/showt...217#post184217

[scs]

Schematic is with unnecessarily high static consumption.

[WM6]

30mA, not low really. Thanks for remark. What you suggest to avoid this?

[Davor]

You may always use higher resistances, say x10, all except 1k in MOSFET gate. TL082 has ~16nV/sqrt(Hz) of noise, so resistances in range of 10k can't ruin anything. Raising the gate resistor would make MOSFET switch slower though.

[Beenthereuk]

Now that I have a working PI detector (Pickini V3), I'm going to have another go at this to see if I can reduce it further and boost its sensitivity to make a decent pinpointer.

Wakiki, did you build your version?

[6666]

Now that I have a working PI detector (Pickini V3), I'm going to have another go at this to see if I can reduce it further and boost its sensitivity to make a decent pinpointer.

Wakiki, did you build your version?

Did you ever do any more work on this project ?