PC based Proton Precession Magnetometer?

[msolc]

PC-Mag

Hi Willy...
After long time experimenting, I have built PC mag: 2 coils similar as yours,hybrid fet/reed relay polarisation switch, 5xJ-fet and opamp preamp, switched cap filter, PC software with IIR filter and a kind of adapted software PLL to measure frequency.
At the end it works excellent, but I have noticed a kind of frequency dragging: when resonating cap is changed for several % it also shifts NMR frequency a little bit, cca 5nT for 10 nF on coils like yours..
Is your system also doing it?

[Willy Bayot]

Hi Willy...
After long time experimenting, I have built PC mag: 2 coils similar as yours,hybrid fet/reed relay polarisation switch, 5xJ-fet and opamp preamp, switched cap filter, PC software with IIR filter and a kind of adapted software PLL to measure frequency.
At the end it works excellent, but I have noticed a kind of frequency dragging: when resonating cap is changed for several % it also shifts NMR frequency a little bit, cca 5nT for 10 nF on coils like yours..
Is your system also doing it?

Hi,

I have never had the possibility to experiment this case as the natural earth magnetic field is not stable enough to make a comparison with a precision of a few nT, even at short intervals.
How did you notice these variations? Do you have a stable magnetic field generator as the ones used to calibrate magnetometers?

Willy

[Krzysztof]

Hi Msolc.

and a kind of adapted software PLL to measure frequency

Suggestion. Is it time PLL trouble doing here. Once I used the PLL to track the fleeing QRP amateur radio signals.
It is watched in the detector signal synchronous to some extent. Yet there was a lot of PC.
Best regards Chris.

[Krzysztof]

Hi Willy...

it is sufficient to keep the main axis of the sensor coils horizontal to get a good enough signal in any horizontal orientation

The questions are not entirely "off-topic". Is the setting of your vertical coil inline will be better? What signal is the position of the W-E ?
For several years, deal with fluxgate technology, of course, not PPM. I use the vertical component of the magnetic field.
I have a linear coil in a vertical position.
However, any deviation from the vertical is already making a significant deviation from the measured value.
I always envied the PPM system, it does not need to worry about heading the coil during the measurement, but you've made about it differently.
Best regards Chris.

[msolc]

Willy,

one version is differential using both audio channels on PC, and when sensors are close to each other that effect is clearely visible. NMR inewidth of water is some 10 nT and combined with sharp Q factor it seems to shift the frequency. Water is among narrowest NMR linewidths...alcohols are 20 or more and disel fuel is much more since it is mixture of various carbon hydrates. I tried to upload image here, but images here can be up to 19k...that is enough for nothing...I can mail it to you...

[Qiaozhi]

one version is differential using both audio channels on PC, and when sensors are close to each other that effect is clearely visible. NMR inewidth of water is some 10 nT and combined with sharp Q factor it seems to shift the frequency. Water is among narrowest NMR linewidths...alcohols are 20 or more and disel fuel is much more since it is mixture of various carbon hydrates. I tried to upload image here, but images here can be up to 19k...that is enough for nothing...I can mail it to you...

jpg files can be 1MB in size. Try that.

[msolc]

fielg graph

Thanks, Qiaozhi,
two parallel lines (red and blue) are two fields...note different distance between them...little V notches are cars passing by on cca 100 m distance...note also when field is more offseted from its actual value, it gets noisier, a big noise was caused by me moving around with tools

[Willy Bayot]

Hi Willy...

The questions are not entirely "off-topic". Is the setting of your vertical coil inline will be better? What signal is the position of the W-E ?
For several years, deal with fluxgate technology, of course, not PPM. I use the vertical component of the magnetic field.
I have a linear coil in a vertical position.
However, any deviation from the vertical is already making a significant deviation from the measured value.
I always envied the PPM system, it does not need to worry about heading the coil during the measurement, but you've made about it differently.
Best regards Chris.

Krysztof,

As you know, the PPM technology is not vector-dependent as is the fluxgate technology. It measures the total field value and not one of its three vectors.
However, the vertical vector is usually the largest at occidental latitudes. Thus, if you keep the main axis of the solenoid coils horizontal, you are sure to get it at 90° to that vector and thus, giving a good quality signal in any horizontal orientation of the coils. The measured value is however not influenced by the value of this vector, it is ONLY influenced by the TOTAL field unlike the fluxgate systems.
Our inline sensors are not made of VERTICAL coils but two horizontal coils with the same main axis and a small separation between them. Each coil has its own proton-rich fluid bottle.

Willy

[Willy Bayot]

one version is differential using both audio channels on PC, and when sensors are close to each other that effect is clearely visible. NMR inewidth of water is some 10 nT and combined with sharp Q factor it seems to shift the frequency. Water is among narrowest NMR linewidths...alcohols are 20 or more and disel fuel is much more since it is mixture of various carbon hydrates. I tried to upload image here, but images here can be up to 19k...that is enough for nothing...I can mail it to you...

Hi,

What reading rate can you get from your system?

Willy

[Willy Bayot]

Thanks, Qiaozhi,
two parallel lines (red and blue) are two fields...note different distance between them...little V notches are cars passing by on cca 100 m distance...note also when field is more offseted from its actual value, it gets noisier, a big noise was caused by me moving around with tools

HI,

I have also made a number of parallel tests like those with sensors separated of a few meters (and two distinct but synchronized single-sensor PPM systems ) but I have never observed any change of values over time. The differences between the two readings were always constant if no moving ferrous object was detected.
I also wonder if your frequency measurement system is not generating this.

Willy

Willy

[msolc]

PC-Mag

hi Willy...
usually I take 1000 period which means cca 0,5 sec of samples. This was recorded with isopropyl alcohol and polarisation was 0,5 sec...what means 1 measurament per second. Due to wide linewidth of alcohol now I use water. With water I can still use 0,5 sec polarisation because preamp is very low noise and signal is still acceptable.
Signals are visible on the picture on lower left side of the screen, having a bit uneven amplitude due to deliberate detuning of sensor coils

[msolc]

no, it was tested with gps controlled reference signal generator-certain thermal shift is visible with PC-built-in soun card (field shift 5nT over first 15 minutes), with external USB hi quality sound interface there is almost no freq shift, and those steps seems to be caused by shifting resonant capacitance.
When resonant capacitors are removed everything is well, signal level is much smaler...
Send me your mail, I'll send you my software so you can test it...my mail is [email protected]

[Willy Bayot]

no, it was tested with gps controlled reference signal generator-certain thermal shift is visible with PC-built-in soun card (field shift 5nT over first 15 minutes), with external USB hi quality sound interface there is almost no freq shift, and those steps seems to be caused by shifting resonant capacitance.
When resonant capacitors are removed everything is well, signal level is much smaler...
Send me your mail, I'll send you my software so you can test it...my mail is [email protected]

What pre-amp chip do you use?
My email address is in my personal profile [email protected]

Willy

[Willy Bayot]

hi Willy...
usually I take 1000 period which means cca 0,5 sec of samples. This was recorded with isopropyl alcohol and polarisation was 0,5 sec...what means 1 measurament per second. Due to wide linewidth of alcohol now I use water. With water I can still use 0,5 sec polarisation because preamp is very low noise and signal is still acceptable.
Signals are visible on the picture on lower left side of the screen, having a bit uneven amplitude due to deliberate detuning of sensor coils

Hi,
What are the specs of your coils?
Willy

[msolc]

coils are 8 layers of 90 windings with 1 mm wire on 50 mm tube
preamp is :
front end 5 j-fets followed by NE5534 (4nV/Hz-2, i was having those ), 4 fets are in parallel connection pre amplifier( hi input capacitance but who cares, to resonate i have to add much more) and 5. fet is current source for first 4 fets
In wiring diagram are only 2 fets, disregard component values
Fets have to be selected for ballanced IDs