Tweet
What determines the required preamp frequency response for PI detectors?
green
green
-
-
In theory, the frequency response should be inverse proportional to frequency spectrum of AIR signal because a conventional PI is simply a wideband metal detector with monocoil. I will post some links to threads where this is explained.Originally posted by green View Post -
The preamp response needs to be faster than the fastest target you want to (or can realistically expect to) detect. If you want to detect a 10us tau target, then the preamp should have, say, a 1us tau. This translates to BW=160kHz. Other factors can dominate, such as the speed of the coil or the overvoltage recovery of the opamp.
Based on the questions you're asking, you may want to invest in a book on metal detector design basics.
Comment
-
AC analysis in Frequency domain shows that the AIR signal is proportional to frequency. However a primitive analysis with ramp function i=C.t showed that:
The conductivity generates TGT and GND signal proportional to d2i/dt2 - the second derivative of TX current.
In freqency domain that meansThe signal from conductivity is proportional to frequency squared:
At conventional PI, the self-inductance (monocoil) generates maximal possible AIR signal. This is valid also for conventional search oscillator used in BFO type metal detectors. All know what is shown in the diagram below. This is the waveform of step-down response of self-inductance or mutual inductance in time domain. Look at how the "flyback pulse" seems in Frequency domain:
For masters of PI technology, the AIR signal is the "flyback pulse" because TX coil operates as RX coil. The frequency domain shows how to suppress the harmful AIR signal or "flyback pulse" - induction balanced RX loop.
At induction balance we have very small AIR signal. Despite this, It should be suppressed to increase possible gain of preamp. That means the gain of preamp should be inverse proportional to frequency. The TGT signal from silver, gold and bronze will be amplifyed more because it appears at low frequencies in the spectrum.
Once upon a time, Mr. 6666 posted a suitable circuit diagram of preamp in the findmall forum. The circuit was designed by Carl for Hammerhead: SPICE simulation with AC analysis showed that with the error of Hammerhead circuit diagram (0.47uF capacitance connected as negative feedback in the second stage), the preamp operates well with induction balanced search head.
Remains MINELAB to patent this.
Comment
-
I'm trying to understand the above image. I'm assuming the flat top is about .6 volts. The first sample is over .5 volts. I thought sampling started at or at least close to zero volts. What is the minimum number of coil TC for the first sample. Maybe change is all that matters.Comment
-
Do not think of volts to the input of the preamp. Important is its gain and its output. Initially you can't take sample because the output of preamp is saturated. The saturation depends on gain. At gain 1000, an input amplitude more than 8 mV causes saturation (if supply voltage of preamp is 10V).Originally posted by green View Post
If we use the frequency domain to explain this, the preamp appears saturated by AIR signal because it has high gain at high frequencies.
You are right that change matters. TGT signal appears as modulation (change) of AIR signal with very small modulation index. To increase modulation index, we should suppress AIR signal. Then we can increase the gain of preamp.
http://www.geotech1.com/pages/metdet...ead/HHv1p5.pdfComment
Comment