PI TX Settling

[Ferric Toes]

ive been reading with interest your postings on here concerning time constants...and it seems like the larger copper and silver coins and larger gold rings have a longer tc than say low conductive items...ive been asking several people on youtube about their whites tdi pro's and larger heavier items and they all use delay of 10 uS for low and high con ductive items...surely their pulse delay should be around the 20-30 uS setting....bearing in mind people on this thread say the larger coins and rings have a longer tc.

Hi Dave,

A short delay, such as 10uS, will always give more signal on both long and short TC objects so those people on youtube are correct. You are sampling higher up the curve on both types of object, but if you then go to a longer delay, say 20uS or even 50uS, you will find that your short TC objects have disappeared but your long TC objects remain. No thin rings, but plenty of copper coins and the odd silver coin; pound coins too.

In the 1970's PI's were renowned for their ability to ignore foil from sweet wrappings and cigarette packets. That is the reason; they used delays of 20uS and upwards. An adjustable delay is still a useful feature as you can set it to minimise the sea water signal on a beach, ignore some foil and bits of pulltabs, lead shot etc. On the dry sand however, you can run 10uS, or even less, with no problem, but you have to be prepared to dig plenty of small rubbish items.

Eric.

[daverave]

ive been reading with interest your postings on here concerning time constants...and it seems like the larger copper and silver coins and larger gold rings have a longer tc than say low conductive items...ive been asking several people on youtube about their whites tdi pro's and larger heavier items and they all use delay of 10 uS for low and high con ductive items...surely their pulse delay should be around the 20-30 uS setting....bearing in mind people on this thread say the larger coins and rings have a longer tc.

[green]

I think the testing I've done suggests the flat Tx is no better than the ramp Tx. Maybe the test methods. Has anyone done tests that show different?

[baum7154]

Thanks Green,

So looking at the attachment on this 6MM Al can sidewall it appears that a 6us sample window would be pretty good.

[green]

Hello Green,

It would be interesting to see the 6mm x 6mm Al can side target in this test as well.

Regards,

Dan

Hi Dan

Lost 6x6mm can side I tested before. Tested one cut this morning as close to 6x6mm as I could get.

[baum7154]

Hello Green,

It would be interesting to see the 6mm x 6mm Al can side target in this test as well.

Regards,

Dan

[green]

Some scope pictures. Tx(160usec ramp, 1000pps) Target distance was adjusted for amplifier out=1.5 volts at 6usec delay. A shorter sample window be better for the 10x10mm can side. The copper coin has a faster decay than the US quarter in the target window with the 160usec Tx. Scope external trigger(fet gate command off).

[green]

I wonder if the simulated results are a bit different to what happens in practice. From what I have both read and observed in tests with a logarithmic front end, the response of a solid target such as a sphere or cube only settles to a single exponential
after one Tau. Prior to that, it is a sum of exponentials (the skins of an onion effect). This I believe can be simulated by having additional inductors and resistors mutually coupled to the single one.

This effect reduces for thin flat objects, such as coins or rings. A medium to thin gold ring will display a single exponential, while a coin such as a silver dollar will display the sum of exponentials effect.

Invariably PI detectors sample well with the Tau of many objects, particularly higher conductivity ones where the TX pulse is too short to excite the primary Tau. What we then we see is only the skin effect response and we wonder why a big silver dollar has much less range than a nickel.

Fundamental theory states that that for 99% excitation the TX pulse should be 5 x the target Tau although 3 x is still 95%. Likewise for maximum current and a almost flat termination of the field before switch off the TX pulse should be 5, or at least 3, time the coil TC. The easy way of improving the coil TC is to add some series resistance.

Eric.

Compared a US nickel(10usec TC) vs 1 troy oz. 99.9% copper coin(500usec TC) with the Tx ramp I've been using. Tx 160usec ramp(6250 amps/sec). Target delay(6usec), target sample(10.2usec),GEB delay(4usec), GEB sample(100usec), EF samples same as target and GEB sample. Change in integrator out with coins at 8 inches, GEB off (US nickel 5.3mv, copper coin 4.2mv) GEB on (US nickel 3.3mv, copper coin 3.5mv). With targets swinging from a pendulum the US nickel was detected less than 1/2 inch farther than the copper coin with GEB off. With GEB on they detected the same distance. Might see an improvement with a 1amp flat 160usec TX(twice the average coil current) didn't test to see how much.

[green]

Your results are coming out good. Slope of t^-1 is what it should be. You will find very slight variations in ground from different places though. What is your method of plotting the results?

Eric.

Using the bench PI circuit I've been playing with. With a DSO, record amplifier out with no target, record amplifier out with target near coil. Dump the recordings into Excel. Subtract the no target signal recording from the target signal recording. Chart the signal with Excel, linear or log scales X or Y. Add a line to the chart parallel to target signal and calculate TC or slope. Tx current is in a servo loop so I can command a flat or ramp. Normal operation, Tx(1000pps, 160usec). Modified the circuit for about 1.5pps and a jumper for 160usec or 16000usec Tx for the latest charts. Some times I use a mono coil and amplifier. Most of the time a differential amplifier with Rx(2 round coils figure eight, 1.5inch or 8 inch diameter) Tx(oval surrounding Rx). The smaller coil if the signal is low for better S/N noise. I do see very slight variations in ground, The Tx time, slope or ramp have a larger effect on the decay slope. Tried 16000usec Tx(to be greater than 5 times the target TC). Didn't know it would change the ground slope to -1.

[Ferric Toes]

Charted ground decay with flat Tx, 160usec and 16000usec on time. Finally got a decay slope of -1. Looks like if Tx time is flat and long enough the slope is -1.

Your results are coming out good. Slope of t^-1 is what it should be. You will find very slight variations in ground from different places though. What is your method of plotting the results?

Eric.

[Ferric Toes]

I was thinking after reading this thread, after 5 target TC's the eddy currents should have decayed with a flat Tx. 16000usec flat TX/5=3200usec TC. The copper coin with a TC of 500usec would have 32 TC's to decay during the flat TX. Maybe I'm missing something or doing something wrong. Does your Hocking have a different reading with stacked coins?

About the Hocking unit, the operating frequency is 250kHz so it's penetration is limited. With two coins together the reading is about the same as for one coin, whatever the conductivity.

Eric.

[green]

Charted a silver dollar(1oz. fine silver) borrowed from a friend. My silver dollars were stolen in a break in awhile back. Close to the 1oz. 99.9% pure copper coin. Tx 16000usec flat.

[green]

Charted ground decay with flat Tx, 160usec and 16000usec on time. Finally got a decay slope of -1. Looks like if Tx time is flat and long enough the slope is -1.

[Ferric Toes]

Here is the linear response to a silver ring 21mm O.D., 1mm thick and 4mm band. It showns how the amplitude peaks at 20uS delay and a TX width of 80uS. The measurements this instrument takes are that the TX pulse width increases by 4 x each delay figure. Hence at 10us delay the TX is 40uS. At 20uS delay the TX is 80uS and so on. All flat topping. The sample pulse width is equal to the delay for all measurements. This demonstrates that TX widths less than 80uS are below optimum for this ring.



Eric.

[green]

By 'below' I assume you mean for times shorter than 100uS. This is because you are in the 'skin effect', 'early time', part of the decay curve. the eddy currents have not yet had time to diffuse to the core of the coin, due to it's high conductivity. At 100uS the diffusion ceases and then we have a single exponential; hence the straight line from that point on.

For a good explanation of this process, download the paper 'Applications of Transient Electromagnetic Techniques' Technical note TN-7 from the Geonics Ltd website. Page 6 for 'Target Responses'. Although this is for geophysical prospecting, the same scaled down principles apply for PI metal detectors.

Eric.

I was thinking after reading this thread, after 5 target TC's the eddy currents should have decayed with a flat Tx. 16000usec flat TX/5=3200usec TC. The copper coin with a TC of 500usec would have 32 TC's to decay during the flat TX. Maybe I'm missing something or doing something wrong. Does your Hocking have a different reading with stacked coins?