LDC1000EVM - Evaluation Module for Inductance Sensors

[Teleno]

I have just received the Texas Instruments LDC1000EVM. It includes a PCB with the LDC1000 inductance measuring chip, a printed coil, a USB interface and a GUI. It costs around €33.

Quick Start Guide



The LDC1000 drives a resonant circuit and digitizes its inductance (L) and its equivalent parallel resistance (Rp).

Both parameters are affected by metal targets as follows:

- Rp decreases as the target gets closer, due to the increasing transformer-coupling between the target's resistance and the Tx resonant circuit.
- L decreases as the target gets closer. The mutual inductance increases with proximity ad gets subtracted from the Tx inductance (L_total= L_tx + L_target - 2M).



The manual says it's possible to determine metal composition, though it does not state how. I guess each metal has its own dRp/dL.

http://www.ti.com/lit/ml/slyb212/slyb212.pdf

I've been experimenting with a gold nugget and an iron nail. dRp/dL was much larger for iron than for gold. Experiments with ferrite provided the smallest dRp/dL, sort of a baseline to tell mineralisation from metal.

I'm also developing a Pinpointer based on these principles.

[WM6]

Experiments with ferrite provided the smallest dRp/dL, sort of a baseline to tell mineralisation from metal.

Nice experiments.

Do you have idea how to use this module as Soil magnetic susceptibility meter?

[Teleno]

Nice experiments.

Do you have idea how to use this module as Soil magnetic susceptibility meter?

It's not the design objective of my pinpoiner ...

But I suppose the changes in L value should provide sufficient information for your purposes.

[Teleno]

The slope as you move the target is very good at discriminating metals and these from ground.

[Teleno]

Got a question for the professionals.

Why are there no field metal detectors based on measuring variations in L and Power of a tank circuit? The possibilities for target discrimination are so good!

[Carl-NC]

That's how many pinpointers work, like Garrett, Minelab, old White's Bullseye, etc. The oscillator works much like a BFO, with all the same limitations. So you could also ask, "Why are there no BFO detectors any more?"

[Teleno]

That's how many pinpointers work, like Garrett, Minelab, old White's Bullseye, etc. The oscillator works much like a BFO, with all the same limitations. So you could also ask, "Why are there no BFO detectors any more?"

The oscillator-based pinpointers I know of rely either on a change in amplitude/power (MD20, Nupi) or in frequency (BFO). None combines the two effects in a 2D plot for discrimination.

Look at the brown and yellow lines in the graph, they prove it' s possible to discriminate even in highly mineralized conditions. Which present oscillator-based pinpointer does this?

[bklein]

I had a thread on this device here a few years back. I was encouraged with first experiments in that it seemed feasible to make a discriminating pinpointer with it. I was not very happy using TI's software tools though - they showed the potential but I was always having to reset baselines etc. to see the effects. Did you develop your own test software? I was going to use Total Phase devices and software but never got to it - but I did buy the Beagle and Aardvark and used them a bit on other projects. Recently I was looking at haptic motors and TI's controller chip - and trying to understand how driver files are written for the Arduino for it. That's where I stopped. Please post your progress in detail so I can catch up and share what I find too. I made notes of what I had been seeing but can't make sense out of them now :-) My goal is too see if it will be able to differentiate between a gold ring and a pull tab that share the same ID on my CTX.

[Teleno]

Did you develop your own test software?

No, I use the comma-separated capture from the GUI in a log file, extract the (Rp, L) pairs from the file and then pass them to gnuplot for display.

I move the targets above the coil up and down and sideways to acquire the points.

My goal is too see if it will be able to differentiate between a gold ring and a pull tab that share the same ID on my CTX.

The pull tab yould be the light-blue line, the gold ring the red line. Their angles relative to the center point (where lines converge) differ by a whopping 20 degrees.

[Teleno]

Discrimination between Au and Tungsten. True 24kt gold ingot vs. gold-plated tungsten ingot.

[Carl-NC]

The oscillator-based pinpointers I know of rely either on a change in amplitude/power (MD20, Nupi) or in frequency (BFO). None combines the two effects in a 2D plot for discrimination.

Look at the brown and yellow lines in the graph, they prove it' s possible to discriminate even in highly mineralized conditions. Which present oscillator-based pinpointer does this?

None that I know of, just saying they use the same basis. The old off-resonance designs (AH Pro) did this, and they were sold as true discriminators. But they only worked to about 4-5" deep.

The answer to the "why doesn't anyone use this method" question is most likely because it doesn't work as well as traditional IB methods. At least, that's what history suggests.

[bklein]

Teleno, did you play with osc. frequency? I read a tech support comment suggesting 5khz was good for discrimination but I thought higher would be better.
Are you using the eval pcb coil or one of your own design?
I'd like to know your exact setup (register settings) to duplicate your results. Last time I did this I had taken lousy notes and could not see the discrimination well enough to get excited.

[Teleno]

Teleno, did you play with osc. frequency? I read a tech support comment suggesting 5khz was good for discrimination but I thought higher would be better.
Are you using the eval pcb coil or one of your own design?
I'd like to know your exact setup (register settings) to duplicate your results. Last time I did this I had taken lousy notes and could not see the discrimination well enough to get excited.

My setup is the default coil and these values:

Sensor freq. 5KHz
Sensor cap: 100pF
Amplitude: 4V
Response time: 6144 cycles
Rp_MAX: 3926.9991KOhms
Rp_MIN: 1.796KOhms
LDCK: 6MHz

You can import the attached file in the GUI (first unzip)