I need a good text on low noise amplifier design ASAP

meter high impedance loops would be pretty awkward physically.

'm having trouble finding good information that I can use. Obsolete transi stors and other unobtanium aren't helpful.

ay, but I didn't see much in the previous versions that I could use for thi s. AofE rarely lets me down.

ollower to drive the output, but I am not seeing any solution that gives re asonable gain without more noise than signal. (noise as reported by spice o n V(Out) of the circuit referred to V1, which is the input signal to the am plifier.)

If you post your LTspice circuit, you might get some comments/ suggestions.

Is your 300 nV/rtHz at the input or output of your amp? Re-reading, 500 kHz is in opamp range. There are numerous low voltage opam ps these days, maybe one of those would work? (I don't have any part numbers to recommend.)

George H.

probably feel silly, but that's life. Despite a number of close calls, I still have a pulse, so I'm still learning.

free. There are ConsultantBucks available to solve this problem, feel free to email me privately for more information. I'd prefer to solve it myself , but I need it solved.

Looking at old CCIR (now ITU-R) measurements, the summer time noise level at 500 kHz in the summer is at least 70 dB above thermal noise in continental US, so with a -60 dB antenna, the recovered signal is still +10 dB above a decent amplifier noise. Of course, if the ferrite antenna gain is worse than -60 dB, you are out of luck.

I really suggest building some amplifier and check if the noise levels, when the antenna input is switched from a resistor to the ferrite antenna. If the noise level increases, the antenna noise figure is sufficient.

Take a look at the Rick CampBell KK7B direct conversion receivers, which uses a common base amplifier to present a 50 ohm load to the SBL-1 diode ring mixer. Similar stages are also used with very low impedance record player pick-ups.

What is the impedance of the ferrite antenna pick-coil ? How much less than 50 ohms ?

I really do not understand how you can have such noise as 300 nV/sqrt(Hz). Even if this is output noise density, it still sounds far too much for a low impedance source. I suggest building something and check if it receiver noise level is below band noise from the ferrite antenna.

e meter high impedance loops would be pretty awkward physically.

a

I'm having trouble finding good information that I can use. Obsolete trans istors and other unobtanium aren't helpful.

way, but I didn't see much in the previous versions that I could use for th is. AofE rarely lets me down.

follower to drive the output, but I am not seeing any solution that gives r easonable gain without more noise than signal. (noise as reported by spice on V(Out) of the circuit referred to V1, which is the input signal to the a mplifier.)

Well it goes as the gain^2 so 3nV/rtHz at the input and a gain of 10 would do it.

George H. I suggest building something and

Output.

As a practical test, the receiver shows no delta in noise when the antenna is disconnected/grounded, so amplifier noise is definitely the issue.

Will do

About 110 ohms. Not too high for common base circuit

There is no noticeable change in noise level when the antenna is connected, or disconnected/grounded.

No, no, no! Output noise voltage grows with the simple gain.

*Power* grows with gain^2.

Jeroen Belleman

Thought so.

Jeroen Belleman

How about something resembling this common base "oldie but goodie" with coupled inductors:

(Gt = m^2, there should be an RF bypass cap at the top of "n")

Sorry, also the cap on the base lead should be in parallel with the resistor.

I was thinking, that if the antenna tap-off coil is low impedance, a step-up into a jfet might help. A series inductor and a cap to ground, series resonated, might work.

But a BF862 is under 1 nv/rthz noise, so may be good enough without matching. Atmospheric noise may well dominate at 475 KHz. The world is very noisy down there.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

A step-up circuit like that is good when the input is coming from a low impedance source cuz you can get a good impedance match while still keeping the driving source resistance to the transistor high.

like a center tapped transformer with input on the primary, output tap to ground, one half to nfet source and one half to gate, and a parallel LC in the load, maybe?

I was thinking that it's low enough freq that you could just parallel the 4 amp inputs in whatever the quad of the NE5532 is (10MHz GBW) and Bob's your dad's uncle. But I don't think it'll work at ~3V

Lots better IM3 performance than a BJT, too.

A pair in push-pull would have lower IMD because the even orders would approximately cancel.

Cheers

Phil Hobbs (coming to you from a beach bar in Cozumel) :)

Lots better IM3 performance than a BJT, too.

A pair in push-pull would have lower IMD because the even orders would approximately cancel.

Cheers

Phil Hobbs (coming to you from a beach bar in Cozumel) :)

Lots better IM3 performance than a BJT, too.

A pair in push-pull would have lower IMD because the even orders would approximately cancel.

Cheers

Phil Hobbs (coming to you from a beach bar in Cozumel) :)

Lots better IM3 performance than a BJT, too.

A pair in push-pull would have lower IMD because the even orders would approximately cancel.

Cheers

Phil Hobbs (coming to you from a beach bar in Cozumel) :)

With really putrid wifi. ;)

Sorry about the multiple posts!

(On the other hand, it does give me a chance to rub it in a bit more.) ;)

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs 
Principal Consultant 
ElectroOptical Innovations LLC 
Optics, Electro-optics, Photonics, Analog Electronics 

160 North State Road #203 
Briarcliff Manor NY 10510 

hobbs at electrooptical dot net 
http://electrooptical.net

Don't try to fool us. It's the gin and tonic interfering with your communications.

--

John Larkin         Highland Technology, Inc 

lunatic fringe electronics

** That is absurd - receiver noise is always specified as "equivalent input" by diving the measured output noise by the voltage gain.

I bet your gain is many thousands - right?

300nVrtHz in a 200Hz band = 4.2uV

How the heck are you measuring that ??

FFS post a schem somewhere.

.... Phil

** That's bonkers.

Receiver noise is always specified as "equivalent input" by dividing the output noise level by the voltage gain.

In any case, 300nVrtHz in a 200HZ band = 4.25uV

How exactly are you measuring such a tiny signal ??

Your whole story makes no sense.

FFS post your schem somewhere we can all see.

.... Phil