Time Domain ElectroMagnetic (TDEM) R&D project

[Willy Bayot]

Hi all,


We have now successfully designed and built a Proton Magnetometer instrument (See under Magnetometer forum, the MarkIVb PPM project).
This device is now mature after 6 years of hard work and we are looking for a new challenging R&D project.


Before this project, back in 2003, we (a team of two) have designed and built the prototype of a digital PI metal detector with magnetic/non-magnetic discrimination capabilities. My aim at that time was to contradict those who claimed that this was impossible. This project produced a working prototype but my partner let me down and the project died away.


In 2004, I have built a simple DC resistivity measurement device to be used the area of our archaeological field.
This simple device worked well to show man-made underground structures like fire pits, wells and wall foundations but the operation was slow, boring and error-prone even with several operators.


In 2009, I conducted a new R&D project over the FDEM technology to replace this device for the same usage.
This technology is very delicate to setup because the receiving and measurement process is executed DURING the transmitting frequency. In spite of the bucking coil, it was very difficult to get good SNR.
The frame on which the various coils are fixed had to be both solid and light to avoid the distortion due to vibrations during the operations.
After a large number of lab tests, I have started the real field tests but these did not give satisfactory results compared to the results I got from the DC resistivity device.


Now, I want to start a new R&D project based on the TDEM technology to measure the relative conductivity of the underground soil and possibly to detect as well any metallic targets (made of magnetic material or not) buried there.
The processes used in this technology are very comparable to the ones used in my previous digital PI project. Thus, I could reuse part of my past experience on that subject.


Unlike the FDEM technology, the TDEM data capture process is executed AFTER each transmitting pulse making it easier to reduce the noise generated by the transmiting coil.


A TDEM device differs from a PI detector with several characteristics:
- Most PI systems use the same coil for transmitting pulses and listening to the induced signal while TDEM systems use separate coils.

- PI systems detect the presence of potential targets by measuring the difference of voltage decay between a pre-captured ground signal and the most current signal capture and aurally and visually signal the results in real-time based on a programmable threshold.
They measures this voltage at a single time slot of the decay slope.
TDEM systems capture and record the decaying voltage at multiple time slots (> 20) during the early decay duration and also during the late decay time. We shall see later how these results are used in post-processing.

- The results of the surveys using TDEM systems are not so much displayed in real-time but rather, are studied in the back office using a number of post-processing tasks.

- Surveys made with TDEM systems (like those made with a magnetometer) are either made under GPS control or at least, as a number of parallel grid lines over the area to be surveyed. This allows to display the results of the survey as 2D or 3D colored scale plots. These plots are then used to make the real target detection or to evaluate the regions of the surveyed area which are useful for the specific purpose.

- TDEM systems inject a larger electromagnetic energy in the ground in order to 'view' deeper ground layers and thus, use larger coils and a larger transmitting current.


The best-known (and now already old) commercial device reference of that TDEM technology is the EM61 instrument from Geonics.


The TDEM system is made of several functional modules, each applied as a separate hardware module and its associated firmware/software code.
- Pulse Transmitter module and it coil
- Receiving module and its coil(s)
- Signal Data Processing, display and storage module
- Post-processing program(s)


I want to start a discussion thread on this forum on that subject with anyone interested in that technology and wishing to share his experience in any aspect of the project.
There will be discussions and work on the subjects of mechanical structure design, electronic design and digital signal processing.
I bring to that project my background as professional data processing and telecom engineer and my long practical experience with digital signal processing using powerful ARM-based microprocessors.


Welcome to active participators and even to simple visitors!!


Willy Bayot

[Willy Bayot]

The very first design decision relates to the specs of the transmitter (coil and its power supply).

I have no precise data about the transmitting coil of the EM61-MK2 (latest EM61 model from Geonics) except that it is a rectangular Air-cored coil, 1.0 m x 0.5 m size carried at 40 cm over ground. It is fed with Uni-polar rectangular current with 25% duty cycle. It is powered from a single 12 V battery.
I do not know the capacity of the 12 V battery carried in the backpack but the specs give a total duration of 4 hours without recharging.
I assume that the frequency of the pulse generation is an entire multiple or sub-multiple of the frequency of local power lines, 50Hz in Europe or 60Hz in USA.

The EM63 from Geonics has an air-cored transmitting coil of 1 m x 1 m fed with Bipolar rectangular current.
The 12V battery weights 10 Kg and is good for 8 Hours of continuous survey work.

I have seen more specs from another such device from Ebinger the UPEX 745 P²I which is a competitor of the EM61.
The specs of its transmitting coil are : air-cored coil, 1 m x 1 m size carried at 40 cm over ground.
The pulsed current is 34 Amp with 15% duty cycle. It is powered from a battery of 24 V.
This enables us to calculate the transient power injected in the ground : 34 Amp x 24 V = 816 Watt for an equivalent continuous consumed power of 816 Watt x 15% = 122 Watt.

This is all I have gathered from the web about the transmitting coil specs of existing TDEM systems.

Any further details about those or about other types of commercial TDEM devices are welcome.

WIlly

[Willy Bayot]

A reminder of the TDEM principles described as simply as possible in layman words:
- a Uni-polar or bi-polar rectangular pulse is generated by a transmitting coil at a reasonable rate, usually 25 Hz for regions where the power lines are at 50 Hz or 30 Hz for regions at
60 Hz. The energy injected in the ground directly depends on the transient current, the duration of the pulses and the apparent surface (dimensional surface x number of turns) covered by the transmitter coil. In order to measure deeper layers of the ground, it is necessary to generate enough eddy current and magnetic field and thus, inject as much as possible energy into the ground in the limits of the physical system (battery duration, voltage loss and heat generated in the transmitter and in the coil,...)
- a separate receiver coil located in the same vertical axis and usually with the same shape as the transmitting coil picks up the secondary pulse resulting from the fast change of current when the current in the transmitting coil is cut.
After the current cut-off, a small delay is allowed to let the oscillations in the receiver coil to disappear.
Then, we must measure the voltage decay of the received pulse at multiple time slots using a fast ADC. This generates an array of exponentially decreasing voltage values.
After proper filtering by using the values captured from multiple successive pulses, we get an array of raw values.
What do we do with that?
The array of values can be categorized into several types: a few early decay time slots, a number of intermediate slots and the late decay time slots.
The early time slots give an evaluation of conductivity of shallow ground or of potential metallic targets.
The late time slots give an evaluation of conductivity of deeper ground layers.
The simplest method to process the raw results is to plot in 2D the values captured at a given time slot or at a range of successive time slots based on the survey coordinates stored together with the data. Using early time slots will give a view of the underground at shallow depths while using late time slots will give a view of deeper layers.

Optimizations:
1. If using bi-polar rectangular pulses during transmitting, the array of values given by a negative pulses are subtracted from the values given by the previous positive pulse. If the rate of pulses is selected as half the period of the local power lines, the adverse effect of the proximity of those power lines is mostly suppressed.
2. When looking for metallic targets, an approximate apparent time constant can be calculated for the whole duration of each net pulse decay slope. This very well summarizes the specific shape of the decaying slope and can thus be expressed as a single value to be plotted, it represents potential targets at any measurable depth in one single plot.
3. a second receiver coil is fixed over the primary coil in the same vertical axis but at a distance of 40 to 50 cm. This coil acts as a gradiometer to help evaluating depths of potential targets. Shallower targets generates a more sensible difference of net results between the two coils than deeper targets.

Willy

[HiiiQ]

Hi Willy. I have both TX & RX coils from EM-83 by geo-western. If you're interested. The only problem is that they're here in Bulgaria...

[Willy Bayot]

Thanks for the proposal but the coils are the least problem we have to solve.
I tried to find the specs of the EM-83 on the web but did not find anything;
Do you have them?

[Willy Bayot]

I see that many people are 'Viewing' my posts but nobody reacts.
I am rather disappointed!! I thought this type of forum was dedicated to TECHNICAL DISCUSSIONS but I see that it is just one more curiosity filling.

Ho, well!! I suppose that's real life out there!!!
Willy

[HiiiQ]

I see that many people are 'Viewing' my posts but nobody reacts.
I am rather disappointed!! I thought this type of forum was dedicated to TECHNICAL DISCUSSIONS but I see that it is just one more curiosity filling.

Ho, well!! I suppose that's real life out there!!!
Willy

Hi Willy. Treasure hunters in the US consider it a blessing if they find a button or a belt buckle from the civil war. Even if it's worth a few pennies. There is no history here and nothing worthwhile to look for. If you want, check out the borders of the former Roman empire. People here find lots of stuff....

[Davor]

Or you could simply say that you are up to building a PI machine and you'd have a full attention of the local trolls.
Many of us are aware of the TDEM and various excitation options, and we are (not) holding our breaths to see what specific solution you are up to, and perhaps give you a hand.

[Orbit]

it already exists Tomorrow I'll write more about it !http://www.youtube.com/watch?v=5AlbvVJxiP8
it works on the same principle

[Willy Bayot]

This is exactly the time we are looking for in our archaeological field. Each summer season, during three months, we systematically survey one more segment of a large vicus which was occupied by Romans between 1st century BC to 3 third BC.
That is one of the objective I have with this instrument, i.e. scan the underground to locate the existing man-made sub-structures by conductivity surveys and thus, better locate the best area to start digging.

WIlly

[Willy Bayot]

Or you could simply say that you are up to building a PI machine and you'd have a full attention of the local trolls.
Many of us are aware of the TDEM and various excitation options, and we are (not) holding our breaths to see what specific solution you are up to, and perhaps give you a hand.

I do not want to build 'another' PI machine.
This project will indeed use the same electromagnetic technology as the so-called 'PI detectors' but its objectives and operations are very different.
1. This type of instrument does not report its results in real-time while walking randomly over the ground like all the metal detectors. It stores all its net results in a file together with their precise coordinates in the field while just reporting in real-time the level of quality of its results and potential reading errors. The results are then post-processed and displayed as 2D color plots to make all necessary diagnostics about potential targets or area of interest.
2. This instrument is designed to survey the underground at different depth layers and separately report the structures/targets found at that depth layer.
3. It will use a high power pulse generator system injecting around 1000 A.sqm of energy in the ground with very short bi-polar half-cosine pulses using a large coil assembly of one sqm (1 m x 1 m).
4. Objectives of the surveys:
- Measure and plot conductance of various underground layers in order to detect any man-made soil disturbance for archaeological surveys.
- UXO detection, discrimination from clutter, depth evaluation.

Willy

[Davor]

The resident trolls do not know that.
Seek "UX1355 - FINAL REPORT" for inspiration of PI colliding with other methods

[Stefan]

Hi Willy,

Thank you heaps for the time an energy you have displayed in providing insight to proton magnetometers. I was first introduced to proton magnetometer in 1981. Things have changed quite a lot since then. Your contribution to this forum has not been lost to some of the older generation. And again thank you very much for your contribution.

Regards,

Stefan

[Willy Bayot]

The resident trolls do not know that.
Seek "UX1355 - FINAL REPORT" for inspiration of PI colliding with other methods

Thank you for the info.
I have made an overview on this document and it was quite informative.
The environment where that study was made (Highly volcanic Hawaii Islands) is quite hostile to any type of magnetic survey (magnetometers go crazy there!!!). This represents the highest limit of difficulty to detect and discriminate targets from the ground effect.
Here in Europe, we have rarely where UXO or Archaeological surveys have to be made in highly volcanic area.

I want to pursue this differential process with a system which is hopefully more sophisticated, more flexible and powerful than the commercial EM61 or EM61-MK2 or even the EM63 that they used for conducting this study.
The pulse rate, the pulse width and the pulse shape will be programmable as well as the field moment (A.sqm) using the most advanced modern power control devices existing now (MOSFET's, IGBT and SCR's) as well as high resolution ADC and fast DSP for sophisticated real-time filtering.
My previous experience with my old digital PI project and our long lasting PPM project should give SOME advantages in this particular project but such a project can not be conducted alone, it requires many types of specialized wits and that's the reason why I have started this thread.

Willy

[Willy Bayot]

This is a document reporting about the design and tests of the prototype of a High power TDEM transmitter from Zonge with the funding of SERDP (dated 2003).

Very clever use of SCR's to generate high power, narrow half-cosine pulses.
I have looked on the web and I do not see this technology re-appearing anywhere in commercial systems but that could have been integrated in high-end instruments since then (2003).

That will be the base of our first prototype as we were seduced by the principles.

Willy