Controls on a Pi

[bklein]

I?m wondering about the controls on a Pi and have some questions:
1) anyone use the pulse delay control? Why doesn?t it allow you to set less than the default design minimum? I mean wouldn?t that be more interesting than the later half of knob travel that really kills sensitivity?
2) would a push button to boost TX source voltage be worthwhile? Maybe a boost converter to a big cap or super-cap?
3) manual variance of sample 1 vs 2 mix. What would it do?
4) variable slew or adjustable lowpass filter in preamp area?
5) in the MPP threads there is mention of a low gain inverting input 1st op amp buying some response time. Will this apply to the surf designs as well?

[dbanner]

Interesting, and how about changing the sample delay with each and every pulse or with every few pulses(accomplished by microprocessor), creating a data profile of the target for discrimination processing later down the road. The information can then be compared to the data profiles of known targets, providing comparative analysis to determine what the target might be. Some storage would be required and some small latency unavoidable, but hey. Finally, pulse discrimination for the 21st century.
This sort of approach can be done in conjunction with multiple period tx pulses, yielding even more information about the target.
Easy to enunciate, difficult to do!

[waltr]

1- yes, I have and use the sample delay control. On my design it does go less than the minimum for any particular coil I am using, ~6usec min. It would be up to you to adjust internal timing circuit or the microprocessor control for what the minimum delay would be.
2- Not so simple. For proper target saturation the magnetic field produced by the coil should become steady meaning the coil current should become steady. This is determined by maximum coil current, the L/R time constant and the coil ON pulse length. Pulse length is best to be about 3 to 5 times L/R. With a microprocessor controlling pulses and switch to higher Voltage or lower series R. this is not to hard to do.
3- Depends on many factors but on a simple two simple PI not much.
4- Probably will not be helpful since the preamp needs to be FAST to accurately amplify the fast decay of the coil plus target.
5- Yes, a two stage pre-amp should improve performance of any PI detector. There are several designs and lots of discussion about this through out the forum here.

I use the Three sample method (see thread for details: https://www.geotech1.com/forums/show...ake-on-the-HH2 ) that will do GEB and has the bonus side effect to discriminating high conductors verse low conductors. So I have a panel pot for GEB sample pulse width.
Another useful control could be a fine adjustment of the pulse period to cancel some types of interference particularly AC mains. The White's TDI has this adjustment.

dbanner,
Microprocessor control and sampling can do all kinds of signal analysis.
I am currently working with a PIC32 that only pulses the Coil and does 2usec sampling during Coil on and for 150usec after coil off. The ADC samples are saved to arrays for analysis.
Upon detection of a target the TC of the decay curve is calculated which is helpful to ID a target. One thing I have done is run the pre-amp output into a Logarithmic converter. The output then is a straight line of which the slope is the target TC (thanks to Green, 666 and others for this idea) with a fairly wide dynamic range. Coil on and other analysis is planned. This is still very much a work in progress and still too early to post anything but it does look promising.

I have worked on concept only so far to vary Coil power (change series R) and coil pulse width. This idea comes from Minelab SD2000 detector (see thread and patents) and some other PI detectors and patents. Then instead of multiple analog channels all is done with ADC sampling and processing. In theory analysis of the curves during Coil on and coil off at different pulse widths should allow discrimination. How much and how accurate is yet to be determined.

The hardest part is figuring out what analysis to do and how to actually code it then what information to present to the user.

[dbanner]

Next up is the wobble coil. What is the wobble coil you ask?
Well this may seem far out there, but if one can dream this **** up then I suppose anything might be possible.
The wobble coil is one which physically oscillates about its own centre axis at a decent rate(visualize a wobbly top, only that the coil don't rotate, it wobbles). Bombarding the target from slightly different planes. The effect of this is that the target is being manipulated in three dimensions.
Now even more data can be harvested.
Wobble coil technology.

[Teleno]

Next up is the wobble coil. What is the wobble coil you ask?
Well this may seem far out there, but if one can dream this **** up then I suppose anything might be possible.
The wobble coil is one which physically oscillates about its own centre axis at a decent rate(visualize a wobbly top, only that the coil don't rotate, it wobbles). Bombarding the target from slightly different planes. The effect of this is that the target is being manipulated in three dimensions.
Now even more data can be harvested.
Wobble coil technology.

Interesring, but the same effect can be achieved by usig several coils facing the ground at different angles.

In any case when you sweep a coil over a target you're doing exactly that.

[bklein]

Just saw something about targeted wireless charging but didn’t read it unfortunately. Maybe there’s something there to accomplish the “wobble”. Might have been on EDN or Electronic Design blog.

[bklein]

What if you just blast with stationary tx coil but compare responses with two out of phase RX coils (figure 8 orientation).
Your micro can then have fun with the two RX signals for EMI cancellation, ground effect, and target orientation...

[dbanner]

The wobble coil thing was meant to be wild speculation, not a serious suggestion. Maybe has some other use by employing Doppler effect in sonar for range and 3d rendering. Who knows.
The wobble effect, think of a see saw, but moving around the circumference of a mono coil. Or a rotating coin slowing down just before it stops, it wobbles. This dynamic instability(phenomenon) must have some practical use, albeit not in metal detector.

[godigit1]

The Wobbly coil.
No Really do you think a coil like thus could have benefits?

[Ferric Toes]

Interesring, but the same effect can be achieved by usig several coils facing the ground at different angles.

Following on from this idea, how about a rectangular coil that rotates around its long axis parallel to the ground. The coil would be driven by a small motor way up the shaft and connected to the coil axle by cable. The difficulty would be in commutating the TX/RX to the coil, but if this could be solved and the coil embedded in a ceramic former with sharp edges, you could cut the grass while detecting.

Eric.

[dbanner]

Yes, actually I do.
A target couldn't easily escape detection simply by its orientation in the ground. This would improve detection depth for small flat targets whose orientation to the coil greatly affects the range of detection. Think tiny gold nuggets of indeterminate shape and random orientation to coil.
The wobble effect would allow the coil to "see" the target from different perspectives.
This is well known in star trek science as the Wobble effect.

[Teleno]

Yes, actually I do.
A target couldn't easily escape detection simply by its orientation in the ground. This would improve detection depth for small flat targets whose orientation to the coil greatly affects the range of detection. Think tiny gold nuggets of indeterminate shape and random orientation to coil.
The wobble effect would allow the coil to "see" the target from different perspectives.
This is well known in star trek science as the Wobble effect.

A true wobbler wobbles everything, from the coil to the shaft through the housing, the batteries and his hat. No hidden treasure can take such pressure, the targets dig themselves out of the ground.

I think you opened a Pandora box.

[Qiaozhi]

Following on from this idea, how about a rectangular coil that rotates around its long axis parallel to the ground. The coil would be driven by a small motor way up the shaft and connected to the coil axle by cable. The difficulty would be in commutating the TX/RX to the coil, but if this could be solved and the coil embedded in a ceramic former with sharp edges, you could cut the grass while detecting.

Eric.

Ideal for detecting in pasture land, as you'd be able to see where you've been.

A true wobbler wobbles everything, from the coil to the shaft through the housing, the batteries and his hat.

According some of the reports on the Chinese version of Fisher's T2, the shaft already provides that facility.

[dbanner]

Ideal for detecting in pasture land, as you'd be able to see where you've been.


According some of the reports on the Chinese version of Fisher's T2, the shaft already provides that facility.

No doubt, there was American Chinaman aboard the bridge of starship Enterprise who turned out to be spy. He gave top secret Wobble technology to the Chinese.
Now all metal detectors from China incorporate this technology.
Actually, he was one of the best space ship drivers ever.R.I.P

[dbanner]

Following on from this idea, how about a rectangular coil that rotates around its long axis parallel to the ground. The coil would be driven by a small motor way up the shaft and connected to the coil axle by cable. The difficulty would be in commutating the TX/RX to the coil, but if this could be solved and the coil embedded in a ceramic former with sharp edges, you could cut the grass while detecting.

Eric.

Actually, that sort of arrangement would be ideal for extracting feric chloride from water after purification process. A rotating magnetic field around a cylindrical water tank. Albeit while cutting grass.
This thread has now gone bonkers thanks to you guys. I take no responsibility for this.