PI TX Settling

[green]

I agree the simulation is not going to capture real-life nuances because of the over-simplified target model. But it does provide some good insight. Stage 2 of this investigation is to build a test circuit with a double-acting TX driver and 2 RX demods. I want to measure target responses for both TX cases exactly simultaneously.

Have you tried stage 2? Have you changed your opinion on constant current vs constant rate?

[green]

Could you also also do a chart of ground using the same techniques?

S/N isn't good enough with my bench detector to chart the ground sample while at constant current. Might have some ripple current that I don't see on the scope exciting the reactive component.

[Old cart]

Tried charting target decay during coil current on and off. Tx(100usec CC) about 10usec ramp before constant current(90usec). Current is in a servo loop. The loop was to control constant rate during coil on time for ferrous non ferrous indication. To noisy for looking at amplifier out during coil on constant current. Good enough for what I'm charting. Charted absolute value(target signal recording- no target signal recording). 1 troy oz. 99.9% pure copper coin. I would be interested in what ground decay would look like during the constant current on time. Would need to improve constant current loop to chart it. Maybe someone could say what to expect and how it would effect GEB. Tx(1000pps)

Could you also also do a chart of ground using the same techniques?

[green]

Tried charting target decay during coil current on and off. Tx(100usec CC) about 10usec ramp before constant current(90usec). Current is in a servo loop. The loop was to control constant rate during coil on time for ferrous non ferrous indication. To noisy for looking at amplifier out during coil on constant current. Good enough for what I'm charting. Charted absolute value(target signal recording- no target signal recording). 1 troy oz. 99.9% pure copper coin. I would be interested in what ground decay would look like during the constant current on time. Would need to improve constant current loop to chart it. Maybe someone could say what to expect and how it would effect GEB. Tx(1000pps)

[Tinkerer]

Charted the copper coin reply #47 and California ground reply #51 on three different charts(lin.-lin., lin.-log, log-log). Don't know how to determine the ground TC. The coin charts a straight line on a lin.-log chart after a 200usec delay. Ground charts a straight line on a log-log chart. Ground charted a slope of -1 with a 16000usec Tx, with Tx times 500usec and less the decay was steeper.

Thank you for the interesting charts and thank you for doing all these tests that give us a better understanding of how PI detectors work.

The following comments are more of a guess than anything else. It would be good if more knowledgeable people would give their comments.

Looking at the lin.-lin. traces, one can see that the ground decay is considerably faster than the copper. The copper decay has not yet reached 0 at 1100us.

Looking at the lin.log traces, the slope of the copper shows 435us, that seems to correspond to the decay or TC of the lin.-lin. traces.

Looking at the log-log traces, we see different behavior of the ground and the copper. This seems logical since the copper response is resistive, while the ground response is permitive.

It is interesting to see that in the lin.-lin., both traces look very similar, except for the TC, no indication of or difference between resistive target and permissive target.

Will samples taken at the beginning of the TX ON time (first 500us) show a difference?

[green]

So it would seem that the CA ground has a TC of more than 200us

Charted the copper coin reply #47 and California ground reply #51 on three different charts(lin.-lin., lin.-log, log-log). Don't know how to determine the ground TC. The coin charts a straight line on a lin.-log chart after a 200usec delay. Ground charts a straight line on a log-log chart. Ground charted a slope of -1 with a 16000usec Tx, with Tx times 500usec and less the decay was steeper.

[Tinkerer]

Wondered what effect the different flat Tx times have on ground response.

So it would seem that the CA ground has a TC of more than 200us

[green]

Charted three coins with 50, 100 and 200usec flat Tx. US nickel(10usec TC) US quarter(130usec TC after 100usec delay) 1 troy oz. 99.9% pure copper coin(500usec TC after 100usec delay). Each coin was taped to a spacer for about a 2 volt signal at 5usec delay with 200usec Tx to give repeatability when removing and placing the coins on the coil. Charted log amplitude, easier to see the difference at lower amplitudes. Charted target recording-no target recording. Rx(two 1.5 inch round, figure eight) Tx(oval surrounding Rx). Better S/N with the smaller coil. Don't know if the larger coil would chart similar. Don't know why the copper coin amplitude is closer at 5usec delay than the US quarter with the different Tx times. No difference with the US nickel(10usec TC)as expected. About 10usec ramp to flat Tx current.

Wondered what effect the different flat Tx times have on ground response.

[green]

Hi Dave,

There are a few things that if done together will improve the depth on larger objects. 1) More TX current 2) Wider TX pulse 3) Wider sample pulse 4) slower response on motion mode to prevent deeper and wider signals tuning out.

I use proportional timing on most projects now. i.e 50uS TX, 10uS delay, 10uS sample 1, 50uS delay to sample 2, 10uS sample 2, then next TX. The next switch position would be all of those previous numbers x2; The next position x3, and so on. Everything goes up in proportion except the pulse frequency which goes down in proportion. Using a constant current supply for the TX and the battery drain is the same in all settings.

If you have extra sample pulses for GEB the beauty is that there is virtually no adjustment needed in the GB setting whatever delay etc you are using.

Eric.

Charted three coins with 50, 100 and 200usec flat Tx. US nickel(10usec TC) US quarter(130usec TC after 100usec delay) 1 troy oz. 99.9% pure copper coin(500usec TC after 100usec delay). Each coin was taped to a spacer for about a 2 volt signal at 5usec delay with 200usec Tx to give repeatability when removing and placing the coins on the coil. Charted log amplitude, easier to see the difference at lower amplitudes. Charted target recording-no target recording. Rx(two 1.5 inch round, figure eight) Tx(oval surrounding Rx). Better S/N with the smaller coil. Don't know if the larger coil would chart similar. Don't know why the copper coin amplitude is closer at 5usec delay than the US quarter with the different Tx times. No difference with the US nickel(10usec TC)as expected. About 10usec ramp to flat Tx current.

[green]

Hi Dave,

There are a few things that if done together will improve the depth on larger objects. 1) More TX current 2) Wider TX pulse 3) Wider sample pulse 4) slower response on motion mode to prevent deeper and wider signals tuning out.

I use proportional timing on most projects now. i.e 50uS TX, 10uS delay, 10uS sample 1, 50uS delay to sample 2, 10uS sample 2, then next TX. The next switch position would be all of those previous numbers x2; The next position x3, and so on. Everything goes up in proportion except the pulse frequency which goes down in proportion. Using a constant current supply for the TX and the battery drain is the same in all settings.

If you have extra sample pulses for GEB the beauty is that there is virtually no adjustment needed in the GB setting whatever delay etc you are using.

Eric.

What is the timing sequence with the extra sample pulses for GEB?

[daverave]

[QUOTE=Ferric Toes;222214]Hi Dave,

There are a few things that if done together will improve the depth on larger objects. 1) More TX current 2) Wider TX pulse 3) Wider sample pulse 4) slower response on motion mode to prevent deeper and wider signals tuning out.

I use proportional timing on most projects now. i.e 50uS TX, 10uS delay, 10uS sample 1, 50uS delay to sample 2, 10uS sample 2, then next TX. The next switch position would be all of those previous numbers x2; The next position x3, and so on. Everything goes up in proportion except the pulse frequency which goes down in proportion. Using a constant current supply for the TX and the battery drain is the same in all settings.

If you have extra sample pulses for GEB the beauty is that there is virtually no adjustment needed in the GB setting whatever delay etc you are using.

Eric.

Hi Eric....I think your idea of three way switching is very good !!! can the sample delay stay at 10 uS for all three switch settings or does the delay have to double at each setting ????i have a buck voltage drop converter with presets for current and voltage and i wonder would this be ok if i use this for my constant current power supply....maybe 24v input to buck voltage converter and say 15v output.

[Qiaozhi]

Are you integrating together the samples mixed from different setup or have you multiple integrations, one for each setup?

I think Eric means that these different settings are switch-selectable, not that they run in sequence.

[Altair]

Hi Dave,

There are a few things that if done together will improve the depth on larger objects. 1) More TX current 2) Wider TX pulse 3) Wider sample pulse 4) slower response on motion mode to prevent deeper and wider signals tuning out.

I use proportional timing on most projects now. i.e 50uS TX, 10uS delay, 10uS sample 1, 50uS delay to sample 2, 10uS sample 2, then next TX. The next switch position would be all of those previous numbers x2; The next position x3, and so on. Everything goes up in proportion except the pulse frequency which goes down in proportion. Using a constant current supply for the TX and the battery drain is the same in all settings.

If you have extra sample pulses for GEB the beauty is that there is virtually no adjustment needed in the GB setting whatever delay etc you are using.

Eric.

Are you integrating together the samples mixed from different setup or have you multiple integrations, one for each setup?

[Ferric Toes]

Hi Eric....i understand now from what you said...i noticed the mirage pi uses a switchable sample pulse width for small and larger gold rings/coins...is that good for getting extra depth on larger gold/coins using wider sample setting ????? im using a very narrow sample pulse and i seem to find more low conductive items but not the bigger rings and coins..good targets in my area are now very deep and even the old large pennies are not much being found...so i just need to get my detectors better for larger deeper targets...ive seen on youtube people getting over 20" on coins using the white's tdi pro...i guess its because it uses app rox 500-600 ma giving it more tx power.

Hi Dave,

There are a few things that if done together will improve the depth on larger objects. 1) More TX current 2) Wider TX pulse 3) Wider sample pulse 4) slower response on motion mode to prevent deeper and wider signals tuning out.

I use proportional timing on most projects now. i.e 50uS TX, 10uS delay, 10uS sample 1, 50uS delay to sample 2, 10uS sample 2, then next TX. The next switch position would be all of those previous numbers x2; The next position x3, and so on. Everything goes up in proportion except the pulse frequency which goes down in proportion. Using a constant current supply for the TX and the battery drain is the same in all settings.

If you have extra sample pulses for GEB the beauty is that there is virtually no adjustment needed in the GB setting whatever delay etc you are using.

Eric.

[daverave]

hi eric

Hi Eric....i understand now from what you said...i noticed the mirage pi uses a switchable sample pulse width for small and larger gold rings/coins...is that good for getting extra depth on larger gold/coins using wider sample setting ????? im using a very narrow sample pulse and i seem to find more low conductive items but not the bigger rings and coins..good targets in my area are now very deep and even the old large pennies are not much being found...so i just need to get my detectors better for larger deeper targets...ive seen on youtube people getting over 20" on coins using the white's tdi pro...i guess its because it uses app rox 500-600 ma giving it more tx power.