Detection distance for a US nickel and quarter

[green]

Read through this thread and the links in it that go over the issues of detection distance, signal and various target conductivity.

https://www.geotech1.com/forums/show...pth-not-really

Statement at end of test reply#1 Decreasing the background noise in the electronics by 50% gives the same depth improvement as increasing the power by 10 times (or there abouts)

Been thinking decreasing noise by 50% gives the same depth improvement as increasing signal by 2(4 times the power).

[green]

This brings up the idea of the integration which adds up a signal over a period of time (the sampling periods). Looking at the Voltage at an instant of time is not the best way.

Have wondered which is best. Interested in detection distance for a quarter with King JL's detector he is working on. Think he is sampling at a delay time not integrating. Sample at start of sample time is greater than average. Integrator out=average volts in during sample time*integrator gain[integrator gain=R feedback/R in*sample time*sample rate]. Wondering if one method is better for short TC targets(1us) and maybe the other is better for long TC targets.

[green]

Remember that a US Quarter is a high conductor (longer TC) verse the Nickle as a low conductor (shorter TC).

I found that a high conductor needs a longer TX pulse to fully excite and have it 'return' a strong signal.
The typical rule of thumb is the TX pulse time should be 3-5 times the target's TC.

I found that increasing the TX pulse time from 100us to 180us noticeably increased the output signal of high conductors (Quarters) and increased the detection distance. This is part of the logic behind some of the PI detectors that do two or more TX pulse times. For example the SD2000 TX pulses are 240us then 4 pulses at 60us. Of course the Sampling timing is also later for the long TX pulse and sooner for the short pulses.The long pulse detect high conductors whereas the short pulse detect low (gold nuggets) conductors.
Try increasing the TX pulse time to over 200us and see if the Quarter's signal increases.

Another thing to look at is the higher conductor's signal is still pretty good for a longer time. What I do in my HH2 is have a longer sampling time on the second sample that drives the integrated output in the opposite polarity. This creates a larger (negative) output signal for high conductors than sampling using only the first sample.
This brings up the idea of the integration which adds up a signal over a period of time (the sampling periods). Looking at the Voltage at an instant of time is not the best way.

Thanks for the replies. Recorded constant current because I couldn't get 70us constant rate. 20us ramp to .5A, then constant .5A. Charted log Y scale, easy to compare change in detection distance with included calculated distance chart(at 400mm 1division is about 20mm change in detection distance). About 1division increase in amplitude from 1TC to 2TC on time. Not much increase after 2TC increase in on time. Test controls constant current, If current wasn't constant, increasing on time could increase peak current and on times greater than 2TC would cause a higher increase in signal.

[waltr]

Read through this thread and the links in it that go over the issues of detection distance, signal and various target conductivity.

https://www.geotech1.com/forums/showthread.php?22641-PI-Pulse-Voltage-More-Voltage-More-Depth-not-really

[waltr]

Remember that a US Quarter is a high conductor (longer TC) verse the Nickle as a low conductor (shorter TC).

I found that a high conductor needs a longer TX pulse to fully excite and have it 'return' a strong signal.
The typical rule of thumb is the TX pulse time should be 3-5 times the target's TC.

I found that increasing the TX pulse time from 100us to 180us noticeably increased the output signal of high conductors (Quarters) and increased the detection distance. This is part of the logic behind some of the PI detectors that do two or more TX pulse times. For example the SD2000 TX pulses are 240us then 4 pulses at 60us. Of course the Sampling timing is also later for the long TX pulse and sooner for the short pulses.The long pulse detect high conductors whereas the short pulse detect low (gold nuggets) conductors.
Try increasing the TX pulse time to over 200us and see if the Quarter's signal increases.

Another thing to look at is the higher conductor's signal is still pretty good for a longer time. What I do in my HH2 is have a longer sampling time on the second sample that drives the integrated output in the opposite polarity. This creates a larger (negative) output signal for high conductors than sampling using only the first sample.
This brings up the idea of the integration which adds up a signal over a period of time (the sampling periods). Looking at the Voltage at an instant of time is not the best way.

[green]

Have been thinking comparing detection distance in ground is difficult because grounds are different. Was thinking in air would be better. Detection distance is controlled by signal to noise ratio. Testing in air in a faraday cage vs next to a AM radio station would probably give different results. Thinking measuring signal at amplifier out might be another way. Coil signal=amplifier out/amplifier gain. Tested my 8inch figure8 coil and a 8inch flat spiral. With Tx_160us,cr, 1A peak the figure8 calculated a little less than 2mV@10us after Tx off, the 330uH flat spiral a little over 2mV for the nickel. N, Q_4 is calculated coil volts out at 300, 400 and 500mm distance using distance vs amplitude_15 with coin centered on coil. Target delay_1 are some measurements I made awhile back with target swinging on a pendulum at about 1meter/second. Target was lowered to coil until it was detected every time. Have read where a GPX4500 can detect a quarter at 24 inches, a Garrett ATX 19 inches. Think the peak current is higher and a larger coil than 8 inches. Hoping to learn what I need to do to detect the quarter near 20 inches. Looks like going to a 12 inch coil and doubling the peak current might get the quarter close to 19 inches.

Does 2mV at coil out 10us after Tx off with 1 A peak current make sense?
Just some thoughts, any ideas or suggestions appreciated.