Tweet
well i have made everything but still it is not working.i have checked the cct it is working all right.i have tuned the filter with a band of 1.7 kHz to 2 kHz.the cct is working fine now.i think the sensors are also working bcz when i connect them in differential mode and place them together they r cancelling each other at a point.now the thing which i m not sure is of the orientation of the sensors.plz tell me in detail about the test conditions .how much distance from the building ang height from the ground.u told it took u almost a month to find out the exact conditions for testing.plz explain in detail as soon as possible
-
-
Re: orientation of sensors
How do you know it is not working? What do you observe on the FFT screen when polarisation current is cut short? Are there multiple FFT peaks at harmonics of 50/60Hz or is there only background noise? At what level is the noise? How did you adjust the vertical scale of Spectrum Lab? It should be set at -60dB to -10dB or 0 to 10mV to be sure to see the ascending/descending FFT peak of the precession frequency. For the first tests, the horizontal scale should be in full zoom-out position (0 to 5000Hz).
Did you carefully tune the sensors?
Did you try with a single (re-tuned) sensor first?
What is the inductance of your sensors? What capacity value did you use for tuning?
What do you mean by 'cancelling each other'?
What cable did you use to connect the sensors to the box?
Outside of building at least at 10 meters from any power lines; at 1 meter above ground; in any direction (preferably North-South) and sensor in horizontal position.
Willy -
Re: orientation of sensors
Back to the subject of orientation of sensors:
The principle of the precession is that the axis of the magnetic field generated by the polarisation must be as much as possible perpendicular to the axis of the earth magnetic field in your region, so that the precession of the protons is maximized.
I looked at a map and I see that the earth magnetic field in your region is inclined of only 30° North from the horizontal position while, in Europe it is 65° North. Thus, in your region, closer to Equator, you would get a better signal if you place your solenoidal sensors in a VERTICAL position or inclined at 60° North from a VERTICAL line.
Sorry for the first answer which could have been wrong.
What is the value of the polarization current?
How long do you keep the current ON?
You could also try to power the polarization with a 24V battery to get a higher current-->Higher signal.
WillyComment
-
Re: orientation of sensors
i know it is not working bcz it is not giving peak at desired frequency or atleast close to that frequency
when we r polarizing the sensors it is giving peak at 2170 Hz and when polarization current is cut short it is giving peak at 1850Hz and 1950Hz alternating (my desired frequency is 1890Hz).
also the sound available at Phill site is much clear but in my case it is not clear and distinguishable.
we hav not checked the peaks at multiples of 50/60 Hz.
also the peak during polarizing is quite high and stays at the same frequency but when polarizing current is cut short the peak is very low and is of almost same amplitude at 1850 and 1950 Hz.
how did u adjust the vertical scale of spectrum lab program.mine is at 0 to 1000mV.
yes i have carefully tuned the sensors.
no i have not tried with a single retuned sensor first.
we have not calculated the inductance values.but the capacitance value according to ur procedure is 570nF.
Ihave used microphone cable to connect sensors to box.
by canceling I mean when I connect tmem in seies and place them together they add the effect and vice versaComment
-
Re: orientation of sensors
I am using 12 volt battery normally used in bikes with current rating of 1.5 amp.
I am keeping it on for 5 sec.
would you plz give any especial suggestion regading the project bcz
cct is working fine
sensors are tunned (570nf).
then where may be the problem? ....Comment
-
Re: orientation of sensors
Note that the difference between 1950 and 1850 is exactly 100Hz. This suggests that you have problems with 50Hz interferences and its harmonics. There should be a powerful source of 50Hz noise around. Is there any neon lamps or any possible buried power line around your electronic circuit?
Note also that the two peaks after polarization SURROUND your local frequency with almost 50Hz differences.
Moreover, that peak at 2170Hz DURING polarization is not normal at all. It is quite POSSIBLE that your amplifier is auto-oscillating at the tuned frequency of the pass-band filter when its input is disconnected.
To adjust vertical scale:
1. Menu 'Options'
2. entry 'Audio Settings'
3. Second Tab '...Spectrum'
4. Adjust the 'Displayed Amplitude Range'
Also look at the complete horizontal scale range (Zoom-out with a right-click on the horiz.scale) before looking at the precise frequency bins in order to see all the interferences.
I would strongly advice you to test with a single sensor first (re-tune it).
You gave me the current rating of your battery but I asked the ACTUAL current that you push in the sensors. Is the microphone cable thick enough to accept this current? Did you measure the resistance of the whole front-end (sensors + cable + relay contacts)? Is the measured current corresponding to the formula V/R?
What do you mean by 'effect' in your last sentence?
WillyComment
-
Re: orientation of sensors
Back to Auto-oscillation:
To prove that you get such a problem, disconnect the polarization battery but keep the power for the relay and amplifier. Switch ON the relay (no current in the sensors), you should still see the big peak at 2170Hz, even at the scope since you say that your vertical scale is at 1V.
The total amplification gain of the whole chain is so large that there are PLENTY of good (bad!!) reasons to trigger its auto-oscillation.
If proven, check ALL the input circuit before the pre-amp stage. There should not be ANY wire longer than 1 cm and unshielded. The output wire/cable should be shielded (microphone cable) and kept well away from this input circuit.
When this potential (BIG) problem has been solved, you should try to make your tests at a complete different place. Try to go in the countryside, well away from any 50Hz power supply or line.
Instead of testing with a single sensor, you could also do the same by reconnecting the two sensors in series but NOT in opposition and keep the two of them very close to each other. In this way, you do not need to re-tune and they will act as a single sensor generating the SAME frequency.
WillyComment
-
Re: orientation of sensors
well for Auto oscillation I have
no occilations at all but some murmuring frequencies but below -70db but the circuit is checked indoor and if we connect or disconnect the soundcard the response is changed slightly so it is assumed that the noises are internal and we are getting no peak at 2170Hz
and another thing is the polarizing current is .8mA which could be a problem, the resistance of the sensors is about 7.4 each. the rest of the current of the circuit will be maild to u as soon as we measure it.
So no tell us whats wrong in thisComment
-
Re: orientation of sensors
one more thing and that is you are talking about using a single sensor do we connect the sensor with the circuit or not if yess then how to tune itComment
-
Re: orientation of sensors
A polarizing current of 0.8mA IS RIDICULOUSLY LOW. You would NEVER get any signal with that.
If your sensors are 7.4 Ohms each, then you should at least get 12/14.8 = 0.8Amp and not 0.8mA. Are you sure it is 0.8MILLIAMPS?
When the polarization current is ON, the input of the pre-amp is opened by the relay contacts. How is it possible that you get a peak at 2170Hz with polarization and not otherwise.
WillyComment
-
Re: orientation of sensors
The last advice I gave you was to keep the two sensors in series but NOT in opposition (exchange the wires of one of them) and to put them very close to each other. In this way, you should not re-tune the whole front-end but you get the effects of a single (but bigger) sensor.
WillyComment
-
Re: orientation of sensors
THe peaks that you observe surrounding the normal frequency are the proof that you get 50Hz interferences.
WillyComment
-
Re: orientation of sensors
i hav checked my cct again without connecting the sensors and also without giving input .i hav obsereved different frequencies at the output such as 653Hz,1.81kHz,2.04kHz,465Hz,1.98kHz,2kHz,952Hz,862 Hz and they r changing alternatey and the voltage is around is 4.65Vrms.i m not getting such frequncy as 2170Hz.
plz tell me in detail how can i get rid off this auto oscillation.so that it dont effect my required signal.
my polarizing current is around 0.8 amp.earlier that was a mistake in telling u.
plz provide a solution as soon as possible. i m waiting for ur replyComment
-
Re: orientation of sensors
4.65Vrms is too much again, this is almost the maximum amplitude accepted by an audio card. Are you sure it is not 4.65MilliVolts?
You say that you do not have a peak bin at 2170Hz but you also say that you observe peaks at 2.04KHz. It all depends on the height of those peaks. If they are only at -60 or 50dB, it's still OK, they are only background noise. The real signal should peak at, at least, -10dB to be usable and it should clearly grow and then shrink over a few seconds of time.
When you told me that you got a high peak at 2170Hz, how high was that peak in mV or in dB?
I can not help you much from here to find the source of any auto-oscillation but I already explained you all the precautions you should take to avoid it.
It is not possible to get a large peak at 2170Hz during polarization (as you told me earlier) and not see it as well when you disconnect the input of the pre-amp since it is exactly in the same pre-amp status.
WillyComment
-
Re: orientation of sensors
If you do not find the exact origin of the auto-oscillation, it is possible to avoid it completely if you short-circuit the pre-amp input DURING polarization. You could use a supplementary contact on the polarization switch or relay for that purpose.
WillyComment
Comment