Design of sensitive low noise receiver in 500Hz-20MHz band

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

I want to design a receiver of very weak signals with amplitude of about few (not more than 100) uV within 500Hz-20MHz band (or, alternatively, at least 5MHz-10MHz) and try to decide which active device to use.

What you would suggest - bipolar transistor, FET, or op amp?

Thank you!

P.S. It's important that the project could be manageable by a beginner in electronics.
 

bertus

Joined Apr 5, 2008
22,270
Hello,

What kind of signal do you want to recieve ?
A "nomal" SW communication reciever has a sensitivity of about 1 μV.

Here is a link to the DX Zone, with a lot of RF information.
http://www.dxzone.com/

Greetings,
Bertus
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

What kind of signal do you want to recieve ?
A "nomal" SW communication reciever has a sensitivity of about 1 μV.

Here is a link to the DX Zone, with a lot of RF information.
http://www.dxzone.com/

Greetings,
Bertus
Hello,

Thank you for the link.

The signal which I want to receive is RF pulse with amplitude as large as noise level - one microVolt, rf frequency 7.5 MHz and duration of pulse is about 5 seconds. The signal is repeated many times, hence it's actually a train of such pulses which I will need to receive.
 

KL7AJ

Joined Nov 4, 2008
2,229
Hello,

What kind of signal do you want to recieve ?
A "nomal" SW communication reciever has a sensitivity of about 1 μV.

Here is a link to the DX Zone, with a lot of RF information.
http://www.dxzone.com/

Greetings,
Bertus

YIP....DXZone is one of my fav'rit places.


You might look at a DIRECT CONVERSION receiver for such a wide range. We built them using DDS synthesizers as a local oscillator for ionospheric research. World's cheapest spectrum analyzer!

eric
 

KL7AJ

Joined Nov 4, 2008
2,229
Hello,

Thank you for the link.

The signal which I want to receive is RF pulse with amplitude as large as noise level - one microVolt, rf frequency 7.5 MHz and duration of pulse is about 5 seconds. The signal is repeated many times, hence it's actually a train of such pulses which I will need to receive.

You got my curiosity glands going. Is this part of a HAARP experiment? If so, I have some really slick integration techniques I can share with ya.

eric
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Ahh... Okay. Well the pulse length is still long enough to use some long integration and averaging...which can really to a lot for improving your S/N ratio.

eric
Yes, there will be time averaging (to increase SNR) handled by digital part of NMR spectrometer (by its host computer actually) - it's just a way how we do it in pulsed NMR.
 

bertus

Joined Apr 5, 2008
22,270
Hello,

What is the fieldstrenght of your magnet ?
What nucleus are you trying to detect ?
For 7.5 Mhz PROTON you will need a fieldstrenght of 0.17625 Tesla.

Greetings,
Bertus

PS I am a NMR service engineer.
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

What is the fieldstrenght of your magnet ?
What nucleus are you trying to detect ?
For 7.5 Mhz PROTON you will need a fieldstrenght of 0.17625 Tesla.

Greetings,
Bertus

PS I am a NMR service engineer.
Sounds great, Bertus :)
My recent question about power amplifier also relates to this low-field NMR project.

It's for proton NMR and I have two different permanent magnets for this project - one is 0.2T (about 8.5MHz) and another 0.4T (about 17MHz). I will design 0.2T spectrometer as some of equipment which I have is limited to 0.1Hz - 15MHz band which is lower than proton NMR frequency in 0.4T field.

P.S. I had mistyped in post #3 - the frequency of the signal which I want to detect is 8.5MHz (not 7.5MHz).
 
Last edited:

bertus

Joined Apr 5, 2008
22,270
Hello,

Are you measuring the FID only or are you trying to do a fourier transform on the FID ?
The resolution of an 8.51 Mhz NMR will be low.

Greetings,
Bertus
 

KL7AJ

Joined Nov 4, 2008
2,229
Sounds great, Bertus :)
My recent question about power amplifier also relates to this low-field NMR project.

It's for proton NMR and I have two different permanent magnets for this project - one is 0.2T (about 8.5MHz) and another 0.4T (about 17MHz). I will design 0.2T spectrometer as some of equipment which I have is limited to 0.1Hz - 15MHz band which is lower than proton NMR frequency in 0.4T field.

P.S. I had mistyped in post #3 - the frequency of the signal which I want to detect is 8.5MHz (not 7.5MHz).
When I was in the UCLA plasma lab, I got to do some work on a curious device called an ICRH (Ion Cyclotron Resoance Heating) chamber. We were attempting to isolate a certain calcium isotope, which had a resonance down around 40 KHz. We needed about 100 KW of power to excite this helical coil, and we ended up strapping a whole trailer full of CROWN sound reinforcement amplifiers together to get the job done. They were they only commercial amps that came close to running at that frequency! I don't think they got any isotopes they were looking for, but it was cool hardware anyway.. :)

eric
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

Are you measuring the FID only or are you trying to do a fourier transform on the FID ?
The resolution of an 8.51 Mhz NMR will be low.

Greetings,
Bertus
Well, FT is carried out over envelope of NMR signal after detection - i.e. over signal whose line width in spectrum can be as low as one Hertz for liquid samples (when homogeneity of magnet is good enough).
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
When I was in the UCLA plasma lab, I got to do some work on a curious device called an ICRH (Ion Cyclotron Resoance Heating) chamber. We were attempting to isolate a certain calcium isotope, which had a resonance down around 40 KHz. We needed about 100 KW of power to excite this helical coil, and we ended up strapping a whole trailer full of CROWN sound reinforcement amplifiers together to get the job done. They were they only commercial amps that came close to running at that frequency! I don't think they got any isotopes they were looking for, but it was cool hardware anyway.. :)

eric
100 KW is always cool ... i mean - hot :)
 

bertus

Joined Apr 5, 2008
22,270
Hello,

Are you using the same coil for the pulse transmitter and reciever ?
In our NMR's it is done this way.

Greetings,
Bertus
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

Are you using the same coil for the pulse transmitter and reciever ?
In our NMR's it is done this way.

Greetings,
Bertus
Yes, the same coil. There were cross coil systems in early NMR (and in some of MRI machines now) but in modern (including commercial - Brucker and Varian) machines it's about one coil, as far as I know.

Recently, I came across a simple NMR spectrometer in Web at http://www.geocities.com/norberto.geo/nmrito.htm and it describes a very simple receiver (it's basicaly just a 3-stage amplifier on 2N3904). What do you think of this design?
 

bertus

Joined Apr 5, 2008
22,270
Hello,

I see there a diode switch is used to block the transmitter power to the reciever.
In our systems a similar circuit is used (hotswitch we call it).

Greetings,
Bertus

PS I am working for Bruker.
 

KL7AJ

Joined Nov 4, 2008
2,229
Hello,

I see there a diode switch is used to block the transmitter power to the reciever.
In our systems a similar circuit is used (hotswitch we call it).

Greetings,
Bertus

PS I am working for Bruker.
I imagine that would be one SERIOUS PIN diode!
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Hello,

I see there a diode switch is used to block the transmitter power to the reciever.
In our systems a similar circuit is used (hotswitch we call it).

Greetings,
Bertus

PS I am working for Bruker.
That's good.
Raggio's spectrometer also provides TDA pulse which makes Q-factor of probe's resonance circuit low when transmitter is on.
 

bertus

Joined Apr 5, 2008
22,270
Hello eric,

The diodes are PIN diodes.
At the highpower transmitters they are active switched.
(You do not want to have 1 kW into the reciever).

Greetings,
Bertus
 
Top