Homemade proton magnetometer - real tale

[Mikediver]

I have found this forum several years ago, looking for some info on PI metal detectors. My main hobby is diving. I live on the shore of Baltic Sea in the region where second world war stopped several days after official surrender. On the sea bottom we have great number of wrecks. Many of them are very interesting diving sites.
About 12 years ago several friends asked me how we can find new wrecks? I had no access to side scanning, so the only possibility was magnetometer. I decided to built one, because commercial devices with acceptable parameters were too expensive or too heavy to be portable (we searched from everything floating - tiny boats, ribs, fishermen boats). The main problem was lack of info. After checking all scientific libraries in my region I have found the most promising book - very old (from sixties) russian book, about precession in earth magnetic field. In spite of its age, the book was written very clearly, with good theoretical background of not only precession, but also all aspects of observation of its signal. Up till today I have the most important part of this book copied and keep it as a bible! But great pity it is written in russian (I have learned russian in the middle school).
I have spent more than one year building different versions of sensor, analog part and digital readout. As the result of this work, proton magnetometer was constructed.
Parameters:
resolution 0,2 nT
sensitivity - better or the same as above
range - from 48000 nT to 52000 nT (3 ranges, every 2000 nT)
repetition rate - from 2 sec to 5 sec, (3 values)
Battery - acid accu, 12 V, 9 Ah, I = 0,8 A (average value)
sensor - diameter 75 mm, length 240 mm
volume of liquid 2 x 100 ml
sensing liquid - water
I have two basic sensors: one for land and another for sea.
I want present here as many details as it will be possible to help anybody who will dare to spend a bit of money and a lot of time to build his own mag. Using this magnetometer I have found many shipwreck (not only steel, but also wooden!) and a lot of war trash (mainly od land).
I want apologize irregular style of future messages. Now we have heavy western gale sweeping our seashore, but when the wather will improve I am going to continue my activity.
Two photos of my magnetometer one with the sea sensor and another with the land version.
Mikediver

[ApBerg]

Sounds good, like to know more !!

Ap

[Mikediver]

After reading the book I tried to observe precession signal. Knowing estimated value of magnetic field I prepared coil: 1400 turns, wire diam. 0,8 mm, coil average diameter 70 mm, length 120 mm. Inside was plastic bottle 150 ml filled with water. Before filling bottle, I removed dissolved gasses boiling water about 10 min. Coil was connected to preamplifier made with fet transistors and opamp. Gain was about 400, and noise 1 nV/sqrt(Hz). I copied ultralow noise amplifier used for gravitational antena, published in Review of Scientific Instr. (can send copy of this article by mail). Next stage was selective amplifier (Sallen-Key structure) with bandwidth 60 Hz and frequency 2100 Hz (precession freq. in my region). Total gain of this stage was 12000. All setup was powered using my car battery, 12V of course. Switching of polarization current was obtained using manual switch. Estimated value of polarizing field was 95 Gs in the center of the coil. Current was 1,2 A. To the last stage (follower) I connected earphones.
Having device completed I selected quiet place in the forest, distant not less than 1 km from any human activity. There the coil was hanged on the tree branch, 2 m over the ground to cancel the influence of magnetic material present in soil. I put coil along East - West line to be sure that amplitude of signal will be as high as possible and started to switch current on and off. Very quickly I listened lowering sound of precession. To be sure that I have real precession I mounted screwdriver to the coil with the tape. Signal totally disapeared!
This part was very hopefull, but to be honest I must add that from that moment I spent one year to measure frequency with the precision which satisfied my needs.
During this phase of experiment I had a problem because initial amplitude of signal changed from one switching to another in chaotic way. From time to time I had no signal at all, but the next cycle resulted in very loud sound! Also I learned how amplitude of signal depending on the polarization time and how long I can listen the signal after switching the current off.
Also I checked the influence of the angle between the coil axis and direction of measured magnetic field. From this experiment I realized that in practice signal was unacceptably low for angle lower than 20 deg.
The next step was obvious - digital redout and selection of mechanical components for back pack device.
Mikediver

[Alexismex]

Hello Mikediver,
Very good experiment and one year of dedication give you a good magnetometer felicitation!
just a little stupid question :
Did you experiment with different" kind of water" to see if sensibility was different ? I am very interesting
see this site:
http://www.newphys.se/fnysik/2_1/schauberger/
and many others who experiment effect of " live Water "
perhaps you can "hear" the different kind of water ....OK
Saludos from Mexico
Alexis.

[Mikediver]

sensor liquid

As the liquid, water was used for the first experiment described before. Later I tried another liquids: benzene, kerosene, hexane and some other organic compounds. Main goal was to select liquid with rather short polarization time to obtain quick repetition of measurements. Water has long time T1 and polarizing field would be long lasting to polarize efficiently hydrogen nuclei. For all other measured organic liquids T1 is shorter, but those compounds causing swelling of plastic containers. Precession time T2 is also shorter in comparison with water, but this is not so important as my time of aqusition of signal is about half second. Including start delay and switching suppression it never reached 1 sec. If we consider attenuation of the signal 2,7 times at the end of T2 period, values of T2 greater than 1 sec are acceptable. As the final solution I decided to use water. To lower T1 value I used low concentration of water dissolved paramagnetic salt with such concentration that I shorted T1 time to c.a. 1 sec with moderate shortening of T2 value. The tiny problem is the unstability of paramagnetic salt which decompose very slowly and after several years shorted T1 time increases about 50%. This is not so great problem because exchange of liquid in plastic bottle is not complicated.
Last month spent on the sea wreck hunting. Results seems to be good - two anchors with chain and something on the other end of the chain... Water clarity is not good enough to make some photos. Must be patient - water will improve in several monts after strong autumn gales.

[ApBerg]

paramagnetic salt

Hello Mike

Thanks ! But can you tell some more about the paramagnetic salt you use ?

[Negu]

What an interest project. I am wreck diver from Spain, so your project is very interesting for me. I have sent you a pm.

[Mikediver]

type of paramagnetic salt

I used several types of ferrous salts. The simplest solution was inorganic salt - ferrous chloride. The main problem is very low stability of this salt in water solution. This is result of (always present) impurities and traces of oxygen present in water. So you must use this compound as clean as possible. It is possible to buy technical grade but spectral grade purity is very expensive and not widespread. I tried to clean this compound with crystallization, but with only moderate success. Concentration used in sensor was in range of several milimol. Also tried ferrous alginate, but this organic salt was partially decomposed (it was very old!). Used ferrous sulfate, but with worse results (shortening of T2 and T1).

[Fred]

Hi Mikediver,
I have found this information really interesting.Practical results are most interesting!
I hope you will continue sharing your excellent work.
Many thanks!
regards,
Fred.

[Mikediver]

Winter storms are coming. So, I will continue. The most important thing is the sensor. Coil is divided for four sections. Every section has its length 32 mm. Outer diameter is 70 mm, inner 52 mm. distance between coils is 70 mm. Coils are wound using copper wire with 0,7 mm diameter. every section has 11 layers of wire, cartoon sheet is put every 3 layers to obtain well ordered winding and to lower the capacitance of the coil. Inside two plastic bottles are inserted. Diameter of bottles is 47 mm, volume about 110 ml. Bottles are located neck to neck. Both coils are wound in opposite direction. Inductances of coils are 22 mH and 22,3 mH. Inductance of sensor is 43,6 mH. This value is lower than the sum of two inductances as the result of negative coupling of coils.
Opposite wound coils can suppress environmental noise. In the first version, and in marine sensor, coils are wound in one direction, all sections have no gaps between them, but all other parameters are the same as described. Bottles are located bottom to bottom. This configuration is not immune to noise, but marine environment is so low noise that this fact does not increase signal to noise ratio.

[ApBerg]

Hello Mike.

Thanks for the Info, can you give some more info from the electronics you use ? Think in the past I have seen some photos from you, using the mag ?
On a Polisch dive site?
Can you give the Website where the photos are ??

Regards

[Fred]

Hello Ap ! Hope everything is fine with you.

Mike, when you say there is no gap between coils , the result is equivalent to only one single coil that fills all the lenght of the plastic tube, right?
In that case is-it necessary to have 2 bottles?

Thank you!
Regards,
Fred.

[Willy Bayot]

There are actually four coils, two pairs of two. The ones which have no gap are probably just there to make their winding easier. The important noise cancellation is coming from the distance of 70mm between the two pairs of coils.
I also think that a single (long) bottle could be used instead of two. The main problem is to find such a slender bottle. I do not think that exists as such except if you built one from a long tube and two glued caps at their ends. I think two separate bottles is the most practical solution indeed.

[Fred]

Ok, i understand now,thanks!
I suppose the coil former could be used to fill it directly with water,usind caps and o-rings for example.
As Ap said, the electronic part would be very welcome

Good luck,
Fred.

[Mikediver]

Hello Fred
The exact explanation of coil structure is such: Sensor is divided for two coils separated 7 cm one from another. Coils are wound in opposite directions. Every coil is divided for two sections. Sections side are turned from acrylic plate 3 mm thick. Two sections are for lowering of coil intrinsic capacitance. Coil capacitance is strongly dependent on the distance between turns. More compact coil - greater capacitance - lower frequency of resonance. To lower capacitance of well ordered coil where every two turns are in tight contact you must divide coil for sections and increase distance between wire layers using paper for example. This is not so important if you have one section for every coil or two. For shorter sections it is easier to wind such coil and also shorter wire layers result in lower internal capacitance.
Noise cancelling is coming from opposite winding directions for both coils. Voltage induced in one coil is nulled with the same voltage but with the opposite polarisation induced in another coil. So, to have excellent noise suppresion you must have exactly the same diameter for both coils and the same number of turns. Precession signal are summed and send by coaxial cable to the input of amplifier.