Amplifier frequency response

[green]

I think the 5, 10, 20, 40mm sizes for aluminum can and pendulum arangement are good standards. Can you recommend an output level to use as a standard for an acceptable detection? This could be something like a signal to noise ratio at the final audio output jack across a standard resistor value. This would allow comparisons between commercially built detectors that have no access or difficult access to the internal circuity.
For do-it-yourself detectors a similar signal to noise ratio level could be established for the output of the receiver and/or the integrator.


Have a good day,
Chet

Hi Chet

Looks like I've been taking stupid pills lately. I'm used to measuring time between peaks to get period but there are two zero crossings for every peak or valley crossing the reason I kept getting the wrong pendulum period. Some thoughts. Maybe adjust target distance for peak target signal equals 1.5 or 2 times peak noise signal. Maybe someone could scope the audio output of some commercial detectors to see if something similar could be used. Modified my pendulum setup. Used a plastic 12oz water bottle full of water instead of a glass bottle. Ran the pendulum line thru a hook in the ceiling to an anchor point to give easy height adjustment. Coil to hook height 219cm. Pendulum height changes but I couldn't think of an easy way to move the coil to adjust target height. Including a scope picture to show signal waveforms. The direct coupled integrator out is capacitor coupled to the next gain stage. Once I get the noise better and we agree on the pendulum setup I'll try to adjust target height for the 1.5 or 2 times signal to noise level for the 5x5, 10x10, 20x20 and 40x40mm can side targets.

[Chet]

Maybe someone could scope the audio output of some commercial detectors to see if something similar could be used.

Hi Green
It will take some days but I will see if I can set something in the backyard away from some of the EMI noise. How far away from the center/coil do you release the bottle to start the pendulum swing?
Have a good day,
Chet

[green]

Hi Green
It will take some days but I will see if I can set something in the backyard away from some of the EMI noise. How far away from the center/coil do you release the bottle to start the pendulum swing?
Have a good day,
Chet

Hi Chet
I moved the bottle about 1/2 meter from the resting position.

[Monolith]

Target is best at the center of the bottle bottom, because the bottle sometimes gets a light spin.

The red trace shows the 5x5mm can-side target swinging 20mm above the coil, distance measured between the rim of the TX coil bundle and the target. I stuck the target on a piece of transparent PET foil to make it easier to handle.

The blue trace is with the bottle fixed above the center of the coil, showing the noise at the same scope setting, 100mV/200ms/div.

[Monolith]

The red trace shows the system noise at 20mV setting, probe x 10, with TX ON, RX coil disconnected and input grounded with a 39 Ohm resistor.

The blue trace shows the system noise with TX OFF. Several noise frequencies are distinguishable.

The job, now, is to find the origin of each of the noises and find a way to eliminate them.

[Monolith]

For those who are analyzing amplifiers that integrate many RX pulses that are synchronized with the TX pulses, as all Pulse induction (PI) detectors do, are utilizing a technology that is used to extract very weak signals in the presence of noise called lock-in amplifiers: https://en.wikipedia.org/wiki/Lock-in_amplifier. How long the target stays within the coil sensitive area while the coil is being swept across the ground will dictate how many RX signals are being integrated and how sensitive the target response will be. Eric Foster was one of the early adopters of this technology in his low power PI machines. He designed PI machines rated in the 3K Pulse Per Second region and calculated that with a slow sweep speed that he would integrate between 1000 to 1500 pulses depending on the actual user sweep speed and coil size. See the above web link to see how lock-in amplifiers and integrated Pulse Induction metal detector RX signals are very similar and affects the sensitivity to a variety of targets.

With modern microprocessor controlled PI machines the TX rate, delay, as well as the RX window width can be adjusted and optimized for the electrical characteristics of the primary targets sought. However, the ground conditions may not allow the optimum TX and RX parameters to be used so the ground needs to be analyzed first.

Joseph J. Rogowski

bbsailor, thank you for the reminder about Lock-IN amplifiers.
Lock-in technique is indeed the solution for extracting weak signals out of high noise.
The pictures on posts 64 and 65 could serve as examples for that. With an impossible noise background, the signal of a tiny 5mmx5mm alu can piece can still be extracted.

[green]

bbsailor, thank you for the reminder about Lock-IN amplifiers.
Lock-in technique is indeed the solution for extracting weak signals out of high noise.
The pictures on posts 64 and 65 could serve as examples for that. With an impossible noise background, the signal of a tiny 5mmx5mm alu can piece can still be extracted.

[In a nut shell, what a lock-in amplifier does is measure the amplitude Vo of a sinusoidal voltage] copied from a link describing lock in amplifiers. I'm missing where the sinusoidal voltage is in a PI decay. I also see a multiplier in the lock in amplifier circuit which I don't see in a normal PI circuit. Could someone explain what I'm missing?

[Davor]

In a way every synchronous sampling is in fact lock-in. Virtues of a real lock-in show themselves with an arbitrary signal, not the one you extract synchronously. Synchronous are like shooting fish in a barrel.

[Monolith]

[In a nut shell, what a lock-in amplifier does is measure the amplitude Vo of a sinusoidal voltage] copied from a link describing lock in amplifiers. I'm missing where the sinusoidal voltage is in a PI decay. I also see a multiplier in the lock in amplifier circuit which I don't see in a normal PI circuit. Could someone explain what I'm missing?

I understand "lock-in", as locking in to a specific spot in time, on the PI wave form. We take a sample window at a specific time, synchronized to the TX-RX cycle, whereby we integrate the signal during the window time. Then we amplify this result and then do other things to it. Therefore I call this lock-in Amplifier.

Off course there are "many ways to skin the cat" and I like to try them all and find new ways, but for now I stay within the broad sense of my definition above.

[green]

A couple experiments. The pendulum setup works good. Repeatable target height and sweep speed. Used a 1 inch spacer to set target height. Pendulum_2 noise compares a mono coil with a figure 8 coil(Tx turned off) amplifier TC .2usec. Pendulum_3 compares amplifier feedback RC time constant at .2usec and 2usec. IB coil.

[green]

More .2 and 2usec amplifier TC tests. The coil decay information gets shifted into the target sample with the 2usec TC amplifier. Would be a big problem if trying to chart target TC's. Think it might be a problem with ground balance but don't know. Other problems or no problems? Scope pictures of amplifier out, IB coil.

[Chet]

Hi Green


Thank you for more interesting charts. Is chart Penduim-3 capacitive coupled? Penduim-3 shows a great improvement on a very small 5x5mm target by using a 2usec amplifier time constant. Do you think that the amplifier feedback time constant needs to be adjusted for different size targets and different sweep speeds? Are the waveform overshoots / distortions in Penduim-4 caused by capacitive coupling? Or maybe a small change in the damping resistor might smooth them out? Where is the ferrite bead installed?


Have a good day,
Chet

[green]

Hi Green


Thank you for more interesting charts. Is chart Penduim-3 capacitive coupled? Penduim-3 shows a great improvement on a very small 5x5mm target by using a 2usec amplifier time constant. Do you think that the amplifier feedback time constant needs to be adjusted for different size targets and different sweep speeds? Are the waveform overshoots / distortions in Penduim-4 caused by capacitive coupling? Or maybe a small change in the damping resistor might smooth them out? Where is the ferrite bead installed?


Have a good day,
Chet

Hi Chet,

Pendulum_1 shows integrator out direct coupled with the last stage gain about x10 capacitor coupled. The trace in pendulum 3 is the capacitor coupled output. The direct coupled signal would go + only(opposite end of figure 8 coil as pendulum_1). I'm taking a 10usec sample after a 5usec delay so the 2usec amplifier filters and moves more of the signal into the sample period. I think taking a 1usec sample after the 5usec delay would be better with the .2usec amplifier. The 2usec amplifier moves what happens during the coil decay into the sample window which seems like might be a problem, maybe not. Pendulum_4 is the output of the amplifier, direct coupled, no capacitors. The upper traces, 2usec amplifier a ferrite bead was positioned on the coil to zero out the offset at 4usec and left there for the can target traces. The ferrite bead is detected during the 10usec sample with the 2usec amplifier but not the .2usec amplifier(can see the ferrite bead change in the first 4usec). I think my first problem is minimizing noise and then seeing if moving the signal during the coil decay into the target sample window is a problem.

[6666]

A couple experiments. The pendulum setup works good. Repeatable target height and sweep speed. Used a 1 inch spacer to set target height. Pendulum_2 noise compares a mono coil with a figure 8 coil(Tx turned off) amplifier TC .2usec. Pendulum_3 compares amplifier feedback RC time constant at .2usec and 2usec. IB coil.

Hi Green
interesting noise results thanks,
you seem to have a good setup right now to measure noise, I was wondering if you would be interested in trying the same test on another type of coil ?
Its called a folded eight.
its just an 11 inch bunch wound coil folded back on its self, can be an ellipse or round they are supposed to be noise rejecting, and still sensitive to small gold, see attached picture
the inside loop can be adjusted for various lengths and sizes etc. It was invented by Reg.
If its not something you want to experiment with thats all ok.

[Old cart]

Hi Green
interesting noise results thanks,
you seem to have a good setup right now to measure noise, I was wondering if you would be interested in trying the same test on another type of coil ?
Its called a folded eight.
its just an 11 inch bunch wound coil folded back on its self, can be an ellipse or round they are supposed to be noise rejecting, and still sensitive to small gold, see attached picture
the inside loop can be adjusted for various lengths and sizes etc. It was invented by Reg.
If its not something you want to experiment with thats all ok.

One other addition that might be very useful would be to add a photocell pointing toward the ceiling. If this is hooked to the second channel of the scope it could sense when the target blocks the light as it would when the target is directly over the coil, this would allow you see coil response delay relative to target position.