This Waveform Thing

[Anonymous]

Funny how this waveform thing gets to you and you can

 

[Anonymous]

Eric:
Yesterday I did a bit of research into the question of whether the character of the laterite response waveform is dependent on field strength, which in theory it should be. The conclusion I came to was that for transmit field strengths less than about 100 gauss, most laterites will be excited in their linear range. Most metal detectors have fields on the order of 1 gauss near the searchcoil. Some PI units may run a bit higher than that, but I doubt any hand-held units approach 100 gauss.
It's been my observation with metal detectors-- pulse, single frequency, and multifrequency-- that when you balance out a piece of laterite or other "red stuff", it's in balance regardless of its distance from the searchcoil, unless the signal saturates the electronics, or the magnetic susceptibility pulls the inductance of the searchcoil so far that signal timing is upset. This means that there is no appreciable dependence on field strength.
You are right, that the longer the transmitter is on, the more domains align with the change in the field. The reason that stretching the timing out doesn't have a pronounced effect on the response to laterite, is because the thermal tail of the laterite response is differentiated by the receiver coil. More field change, but it takes longer.
--Dave J.

 

[Anonymous]

I was wondering about how the target time constant really relates to the transmit time/waveform shape and this answered it best. Nice practical stuff.
So I guess if you only wish to hunt for large conductive objects, a high inductance (many turns for transmit power and receiver sensitivity) is ok and maybe even good. Turn off can be slower if you can't stand the flyback. Damping of the coil is longer and thats ok too, cause the first sample is way out there at 35 us or more.
If you wish to hunt for gold flakes then a low inductance fewer turns is required, but short pulses are ok and some gain regained by the number of samples/sec, but you're going to have to get it turned off in a microsecond or less and the coil damped in about the same time, cause the leading edge of first sample is at 5 us or less. At which point you are going to get alot of interference from different problems the ground has, cause now you're looking at all sorts of little bits in the ground.
All those things in the ground PI was supposed to ignore, and did until you try to make it "sensitive" at which time you do.
Makes me glad I don't do this for a living.
JC

 

[Anonymous]

Hi JC,
That's the general idea. Bigger coils, more turns, higher current, lower rep rate, for large objects buried deep. Smaller coils, less turns, less current, higher rep rate for small shallow objects. However, you can juggle coil time constants as I did with 1mH and 50ohms circuit resistance. If you use a separate RX coil, this can have more turns than the TX as its time constant is governed by the RX input impedance and a separate damper. One important point for high sensitivity units is to keep the coil circuit capacitance low, as the self resonant frequency determines the damping and the rate at which the current can be brought to zero.
Eric.