field test unit no 001 "model T"

[Davor]

I thought of this one:

Basic theory requires that the TX switch off should be at least 5X faster than the fastest object TC that you want to detect, preferably 10X. Beyond that there is no benefit and energy is wasted.

Eric.

IMHO everything beyond 5X falls into wishful thinking category. True, I have no first hand experience yet, but simulations support this flow of thought.

If we drive it waaaay upstream to absurdity, we could produce infinitesimally short pulse spiking at few kV ... with very mild target excitation.

On the low side, we could slow down flyback pulse, at non-lethal voltages, to achieve more.

[Ferric Toes]

In the light of inevitably delayed first sample, and seeing the sample delay as 1/t low pass, wouldn't it be more energetic to excite targets with a little less spiky Tx pulse at the same peak voltage?

Not sure what you mean by a less spiky TX pulse. Are you referring to the current or the back emf at the end of the pulse? It's the current that sets up the magnetic field so looking at the current waveform tells you what the field is doing. Some PI detectors have very low resistance coil, so the current never reaches near the final value at the time of switch off, unless of course you lengthen the TX pulse to 5X the coil time constant. In the case of a 300uH 1 ohm coil this would be 1500uS. Too long for most applications and wasteful of battery power. For a 100uS TX pulse the current waveform looks decidedly spiky and is still changing rapidly at switchoff time with the potential to generate reverse eddy currents in the target. I prefer a current pulse that flat tops for a time before the switch off and an easy way of doing this is to add some series resistance in the coil circuit. Coil current is less per pulse but then you increase the repetition rate to compensate. One version of a PI that I have has a peak current of 0.4A in a coil with a winding resistance of 4 ohms and L = 300uH. This is done by adding a 27 ohm resistor to the coil circuit. Small current switches off faster and the back emf just rises to 60V, well below the avalanche voltage of the IRFD210 mosfet. Sample delay is easily achieved at 10uS and with a bit of tweaking plus a fast preamp can be reduced to 5uS.

Higher currents usually mean running in the avalanche mode which stretches switchoff time and you have amplifier recovery on top of that. Basic theory requires that the TX switch off should be at least 5X faster than the fastest object TC that you want to detect, preferably 10X. Beyond that there is no benefit and energy is wasted.

Eric.

[Davor]

The flyback, of course.

[moodz]

In the light of inevitably delayed first sample, and seeing the sample delay as 1/t low pass, wouldn't it be more energetic to excite targets with a little less spiky Tx pulse at the same peak voltage?

Are you referring to peak tx voltage or peak flyback voltage?

[Davor]

In the light of inevitably delayed first sample, and seeing the sample delay as 1/t low pass, wouldn't it be more energetic to excite targets with a little less spiky Tx pulse at the same peak voltage?

[Ferric Toes]

... and nailed down with patents ? There seems to be some aggressive moves in the industry to claim any form of sampling ground balance as proprietary.

The Poole patent filed in 1979 (GB 2041532) covered just about everything, subtraction, adding gain in a sampling channel, ratio of samples, and taking "any number of samples".

Eric.

http://worldwide.espacenet.com/publicationDetails/originalDocument?CC=GB&NR=2041532A&KC=A&FT=D&ND=1& date=19800910&DB=&locale=en_EP

[Qiaozhi]

I think you're going to like this ->

Here is an LTSpice simulation of equation 11 in John Alldred's paper:

.......... (Eq.11)

Since >> and >> , the dimensionless scale factor is typically close to unity, and can therefore be excluded from the equation. Hence we have:





Unfortunately LTSpice does not possess a library of mathematical symbols, but it is possible to achieve the same functionality by using B elements for the exponential terms and a couple of 2-dimensional polynomials for the multipliers. This has the additional benefit of allowing the individual terms to be plotted in the graph.

The graph shows how the output voltage (V) varies for a specified main sample delay width () as the target decay value () is swept from 1us to 1ms.

Initially I ran this simulation as a DC sweep, and performed a parametric sweep of . Although this gave the same result, the x-axis was labelled in volts. Then I realised that LTSpice has access to the current simulaton time via the "time" variable, and it became a simple matter of using a time domain analysis to sweep the value of .

You can readily observe how the amplitude of V is affected at the extremities of the curve. i.e. on the left-hand side is the dominant factor, but on the right-hand side has little effect. This was correctly predicted in the previous calculations.

I have attached the LTSpice file for your delectation.
If you uncomment the .step command, you can generate a set of curves for from 1us to 100us.

[Aziz]

...
..
There seems to be some aggressive moves in the industry to claim any form of sampling ground balance as proprietary.

It goes even further. They claim to own the physical facts and natural laws... *LOL*
I think, I should engage my thermo nuclear weapon for the end game ... *LOL*
Do we already have the 21-dec-2012?
Aziz

PS: Nice Moodz, you finally did something.. (except yawning of course).

[moodz]

I have that sorted, tried, and tested in the field. Works beautifully.

Eric.

... and nailed down with patents ? There seems to be some aggressive moves in the industry to claim any form of sampling ground balance as proprietary.

[Ferric Toes]

I would have thought the ground balance would be of more concern with regards to sampling.

I have that sorted, tried, and tested in the field. Works beautifully.

Eric.

[moodz]

I looked at my CRO and this is what I saw with the following setup. TX pulse 60uS, delay 20uS, sample pulse 20uS, delay to 2^nd sample 100uS (Start SP1 to start SP2), and second sample 20uS. Repetition rate 5K pps.Sine wave generator injected into RX input such that the sine wave gives pk – pk of 1V on the output of front end amplifier.. CRO connected to differential integrator output. Leaky integrator time constants 0.1uF//1M0 and input resistor 2K2. Adjusting dc offset of front end gives no change in integrator o/p with scope set to 20mV. 50Hz input give no noticeable change either. Hence 0 – 50Hz there is no pass of signal. You can of course see the ramp up during SP1 and the ramp down during SP2 but the baseline does not move. Ramps can have positive or negative values depending on the part of the cycle is being sampled. Raising the signal generator frequency one starts to see fluctuations in the baseline at about 0.5kHz. Tuning of the sig. gen. is very critical to see this but it comes and goes like a beat note on the baseline. From then on there is interference every 0.25kHz. i.e. 1.0, 1.25, 2.0, 2.5……. Severe interference occurs at multiples of the repetition frequency. This goes on right up to 200kHz although the effect seemed less at the high frequencies because you are getting integration within each sample pulse width.This is exactly what I would expect to see from such a sampling system, whose purpose is to minimise offset, EF and powerline noise. Higher noise frequencies are best dealt with by other means. Although, by employing an adjustment of the clock frequency driving the TX and sample rate many of the problems with higher frequency CW interference can be overcome because it then becomes asynchronous. I once tried a system where the clock rate was swept over a range so that the TX/sampling pulse rate was continually changing. Results were inconclusive though.I should point out that the detector I used for the above test has considerable filtering after the integrator i.e 1^st order LP immediately after and switched 1^st/2^nd LP later.I’m, not quite sure what is meant by a “flawed foundation” in describing something that has worked just fine for over 40 years. Not to say that there are not other alternatives that may do things better. The reciprocating internal combustion engine could be said to be a flawed foundation for driving the wheels of a car. It is still the majority power plant.My Magnetic Viscosity Meter uses a similar sampling and integration method but one can produce accurate graphs of the decay of ferrous and non-ferrous objects, but magnetic soil in particular.Eric.

The crystal set comparison was somewhat unfair ...of course one would expect over several decades of improvement that there would be a highly refined and performed example of the art. The high pass filtering whether performed earlier or later in the signal chain is fairly straightforward .... I would have thought the ground balance would be of more concern with regards to sampling. My reference was to the use of a damping resistance in pi circuits .... A comparative technology is radio where they stopped using resistors to match antennas a long time ago ... The use of a resistor saps signal to noise .....and adds noise to the system. It's not that the pi won't work ...it is just less optimal than it could be.

[moodz]

yawn ...

[Aziz]

BTW guys,

one can code a small C/C++ program and can do all the demodulator steps in software (simulation) and can evolve the transfer function easily. You have to make a Monte-Carlo method (varying frequency and phase of the low frequency EMI) and have to take the worst case output. Voilà, that's it.
This is what I would do, if I don't know, how to do it.
Aziz

[Davor]

Please note that the Nyquist filter has a comb transfer function. We are talking about the same thing.

[Aziz]

Thanks Eric. Can we conclude, that subtracting integrated windows is a phase-sensitive high-pass filter with typical characteristics of a HP-filter up to it's cut-off frequency (-3 dB mark) and on higher frequencies, it then becomes a comb transfer function?

The phase (time) sensitivity can be eliminated, if one takes always the worst case condition.
Aziz