bipolar pulsing to remove Earth Field

[mickstv]

Hi All, just wondering if anyone remembers seeing any info relating to using bipolar pulsing in PI with a Mono coil to get rid of the Earth field effect. Or if someone could point me in the right direction to start making a circuit to experiment with please.




Cheers
Mick

[Tepco]

Bipolar pulsed detectors are not made to compensate for Earth magnetic field, no need for that, field gradient is too low to be detectable anyway. Instead, useful for two other purposes: compensating ˝slow” component in ground mineralization and, of course Military, requirement for zero net magnetic flux generated by detector for demining. There are some patents on this, but i cannot point to specific one right now (relaying on crappy netbook i don’t have any data here), do some search. Well known detector built this way (bipolar pulsed monocoil) is Fisher Impulse 1280 (but in energy recovery configuration, no flyback pulse).

[Tinkerer]

Bipolar pulsed detectors are not made to compensate for Earth magnetic field, no need for that, field gradient is too low to be detectable anyway. Instead, useful for two other purposes: compensating ˝slow” component in ground mineralization and, of course Military, requirement for zero net magnetic flux generated by detector for demining. There are some patents on this, but i cannot point to specific one right now (relaying on crappy netbook i don’t have any data here), do some search. Well known detector built this way (bipolar pulsed monocoil) is Fisher Impulse 1280 (but in energy recovery configuration, no flyback pulse).

Hi Tepco,

"field gradient is too low to be detectable anyway"
This may be true for a low sensitivity detector. Once you reach higher sensitivity, the 1 or 2 micro volt signal from the coil moving through the Earth field, becomes important.

I just made a quick sensitivity test. Using only the preamp, LP and integrator. Total gain, 5500.
In reality the gain is a little less, because of the low pass filter, but even if we consider the real gain to be 2500, this gives us a signal amplitude of around 2micro volt for an aluminium cocacola can at 1.3meter from the coil.

This is using a fairly large coil of 1m x 0.5m and pulsing it with 18V at 190us.

The measured integrated signal amplitude is then about 5mV.
Divide the 5mV by the gain and you end up with an initial signal amplitude of somewhere between 1 and 2 micro volt.

You will not find this target if you have 1 or 2uV of EF noise blanking it.

Tinkerer

[Tepco]

Yes, something like that, but still Earths magnetic field is not detectable, it is relatively uniform, local variation is too small, even on equator, for detector to respond.

[Davor]

I think if you skew and twist an existing monocoil into a figure 8 shaped ... thing, you'll get many things cancelled out, including radio sources and any far field emission. No bipolar pulsing required.

[Altra]

Mick
I have been experimenting with bipolar circuits. Have not seen any great advantage.
In fact it is more complex than any gains that may be achieved. You need a bipolar power source, ie two batteries. At high voltage ratings, common N and P channel fets are hard to match. A 500v N-ch fet may have a RDSon of less than a ohm. While P-ch fets rated at 500v will have a RDSon over 4 ohms. Then you need to do a level shift for switching the second fet. etc. The plus side is you have a clean bipolar power supply for your analog circuits. No noisey dc to dc converters. More options for audio output, since you can use either + or - rail for power. There are probably other unseen advantages to be or are being developed.

mono pulse - Negative pulse, sample1, Target is positive and EF is positive
sample2, Target is near 0 and EF is positive

Invert sample2 add to sample1, efs cancel, results in 1x target

Bipolar pulse - Negative pulse, sample1, Target is positive and EF is positive.
Positive pulse, sample2, Target is negative and EF is positive.

Invert sample2 add to sample1, efs cancel, results in 2x target

Both work, because the earth field is not effected by tx pulse and remains constant and maintains the same polarity over the short sampling periods.

Tepco - There are static geomagnetic fields and local anomalies which effect metal detectors. The detector becomes a induction magnetometer when in motion.

Here is a scope shot of one I been playing with

sampling at 8us, preamp output

[Tinkerer]

Yes, something like that, but still Earths magnetic field is not detectable, it is relatively uniform, local variation is too small, even on equator, for detector to respond.

The Earth's field is roughly parallel to the earth surface at the equator, in a north south direction. The influence would be minimal. However, there are local anomalies that may distort the field and the angle may be near 90 degrees, generating maximum current in the coil.

At higher latitudes the general angle is more like 45 degrees to the earth surface.

The earth field is considered to be about 0.5 Gauss, although there is a region in south America where it is closer to 0.25 Gauss.

Tinkerer

[Carl-NC]

Earth field is easily detectable. Bipolar pulsing is a good way to cancel it.

Bipolar pulsing does not require 2 batteries, or even a bipolar power supply. It can be done with an H-bridge, or per the Fisher Impulse.

- Carl

[mickstv]

Thanks for the replies everyone. I was hoping it would be easier to use bipolar than multiple samples to remove EF but by the sounds of it, it's going to be just as complicated.






Thanks
Mick

[Tinkerer]

Experimenting with EF compensation

The Earth's field or EF signal, manifests itself most strongly, when we suddenly tilt the coil. For this we need to be in the field.

How can we experiment with this in the lab?

One way is to disconnect the TX pulse. Without the TX pulse, the coil(s) become passive receivers.
Now we sweep a ceramic magnet across the coil and observe the response on the scope.

The response should be the same on sample #1 and sample #2. With the response the same, the signal will be nulled in the difference amplifier.

If the signal shows opposite polarity, invert it first.

Tinkerer

[mickstv]

Hi Tinkerer, thanks for that. What I wanted to do was try and cancel the EF first then hopefully be left with only messing with one primary target sample and a second inverted sample for Iron mineralization/Ground balance.

I was thinking that this would enable me to have a faster repetition rate because I wouldn't need late samples for EF.



Mick

[Tinkerer]

Hi Tinkerer, thanks for that. What I wanted to do was try and cancel the EF first then hopefully be left with only messing with one primary target sample and a second inverted sample for Iron mineralization/Ground balance.

I was thinking that this would enable me to have a faster repetition rate because I wouldn't need late samples for EF.



Mick

How high do you want to go with the PPS? Is 5000 good enough? 10,000 maybe?
Have a look at the Tinkerers TEM, it shortens the decay time considerably.

Tinkerer

[Tepco]

Tepco - There are static geomagnetic fields and local anomalies which effect metal detectors. The detector becomes a induction magnetometer when in motion.

Exactly, local distortion in magnetic field (presence of some object or mineralization) can create high enough gradient and can be detected, whit detector working in “passive|” mode, TX turned off, this is now some sort of magnetometer. What i ment to say is, Earths magnetic field alone (in absence of any anomaly) will not be detectable, moving coil thru the relatively homogenous field will not induce anything in it no matter how sensitive detector is. All trickery is about eliminating ground influence, not Earth magnetic field influence itself, making GB, not canceling EF.

One more comment here is about Impulse, and my relation between love and hate whit it (hate, mostly), my opinion is that 1280 well may be great “missed opportunity” in detector design, approach can be exploited whit lot to offer in some other incarnation.

[Altra]

duplicate

[Altra]

Hello Tepco,

What ever it is, it exists and needs to be eliminated.

I agree with you on the Fisher half bridge transmitter. I built it using bipolar transistors
and fets, but could never get the delay down. The diodes that make it work also kill the signal .It works good as a continous wave transmitter.

As Carl stated, an h-bridge might make it all practical. Here is one I am going to try next.
see attached patent. The schematic has a mistake or two