IGSL

[Davor]

Could you elaborate a bit on ground noise specifics? It is still new for me and I don't want to dismiss a valuable concept just because I don't understand it.

[simonbaker]

Could you elaborate a bit on ground noise specifics? It is still new for me and I don't want to dismiss a valuable concept just because I don't understand it.

"ground noise" is just my terminology -- maybe there is an official name for it.

I just mean, if you wave a MD coil over soil that has no targets, you will still pick up signal due to variations in ground conductivity, mineralization, hot rocks, etc. That is the "fog" people talk about. If you bury a target deep enough, no matter how much gain or S/N you have in your MD, your target signal will still be buried in the "fog" and it doesn't matter if you have a super-gain, zero noise circuit and zero EMI -- you'll still just see the fog. I believe there is always some level of fog at all phase angles.

Basically, background detected signal that can't be separated from the target.

I don't know if anyone has measured some data for it.

-SB

[Davor]

OK, I got it. So basically it is the noise out of various sources, including differences in mineralisation, that is aggravated by the proximity of ground.

[zoomix]

VIDEO IGSL coil TGSL

Hi all,

My IGSL is over and here's a video of a test in the grass.

http://youtu.be/mjy5ZpFoZ2U


There was an error on my front! normally this is "accept NF" and "reject Ferrous" ....

The two sensitivities is almost at max (ca chatters)
The coil was "nulling" to a value of AC 4mV on RX and filled in with epoxy.
Discrimination ferrous and nonferrous seems correct (seen on the oscilloscope and compared with images in the folder of Don B.)

GB are the potentiometer on the front and 50% are what looks good for potential adjustments.


The igsl works well until 20/25 cm

I always used to make the transition from coil and my tests of the target high enough so the sounds are limits

I would like a little more power / heights ... but I'm really excited to try that in the fields. The discimination of aluminum works.

If you have any advice or tests for enhanced depth, I'm interested.

I would like to increase volume of HP .. I can replace the two 330R by 220R on the BD140 ?

[tras20]

Hello how this connection should be?

http://www.nuggets.at/DETECH-30x38-c...-kHz-SuchSpule

[simonbaker]

OK, I got it. So basically it is the noise out of various sources, including differences in mineralisation, that is aggravated by the proximity of ground.

Specifically I mean phase modulation of the TX/RX signal due to anything in the ground as the search head moves across it. So it is theoretically "repeatable" noise, because it is spatially determined. It is identical to trash, except it is the natural material of the earth (but trash can be included as well). So it is not a time-varying noise source, such as EMI or circuit noise. Much of it can be discriminated out (especially by ground-balance circuit), but what is left over is the "noise" I'm referring to. I call it the "ultimate" noise, because no matter how much you increase the TX magnetic field, it will not help bring the target signal out; the noise gets louder too. In contrast, increasing the TX field should help overcome EMI and circuit noise.

That's all, nothing too profound. But if we could know the typical "ground noise" where we hunt, we could theoretically know when to stop trying to improve the S/N of our MD circuit, because it won't help any more.

-SB

[Davor]

OK, I got it. So you are actually describing some baseline response of the ground itself. True, for such response increasing Tx power would increase such phenomena proportionally, and the real advance would come if a system is capable of suppressing ground response, just like PI does by delaying Rx sampling.

I think there is a way to notch out most of the ground response by Tayloe filtering. I'll have to check about this first.

The surefire approach is by employing differential coils and a 4 quadrant Rx. This would take care of the EMI as well.

I believe VLF can regain ground lost by PI with these improvements. That is - if PI does not get improved as well.

[zoomix]

Problème instabilité

Hi all,

Here I make some settings (inverted RX connection) and the detector is 50cent euro + 25cm 30cm in my garden

I was in a plowed field with short grass. I did the usual settings and I found a small room. After a few minutes the detector becomes unstable donations .... I give a 2 euro coin on the floor and redid the settings.

The detectors preserves its stability 5-10 minutes and then starts to become unstable.

Sensitivity has no effect on the instability .. it would rather the ground balance and discrimination. I detected mainly in the fields

I saw that some people have increased the value of the capacitors C8 and C9 ... Someone on a board?

I saw that sold RFI filter? to put on the cable as


Thank you for your advice because I'm almost at the end of this project and I want it to be stable ...

[Davor]

This happens to me when batteries are run down. For some reason the oscillator (Tx) has some stability issues if set too close to the oscillation limit. It is the cleanest sinewave there, but the rig stability is better if you turn the trimmer a notch to the lower resistances.
This problem was more pronounced before I corrected a problem in audio circuitry, and had much to do with weak batteries.

What board do you use?

[zoomix]

Instabilité

Hi Davor,

Thank you for your answer. I used the PCB Eduardo IVONIC corrected.
Prior to the test, I measured my battery à11.7 volts. I think it is still good because the regulator 78xx voltage must be entered in Min 2Volts than the output voltage.

My component is controlled but I do some corrections / welds with IC placed on the PCB. Maybe a replacement IC would be interesting to test.

You talk down the value of the threshold trimmers audio?

Can you give me the post with your audio correction?

[Davor]

On my PCB a polarity of a large capacitor was shown wrong, and I didn't notice it, so in reality my audio was transferred to the power rail. It is much better with the correctly working capacitor.
When you adjust the oscillator, you'll notice a spot where oscillator just starts, and if you check the waveform you'll see that it is the spot with cleanest sinewave possible. Well, turn a little bit more so that the amplitude of oscillation just reaches its maximum. It is not as clean waveform as before, but its stability is much better.
Regarding my threshold controls mods, please check the schematics from a few posts before. They can be joined to a single stereo potentiometer, and I suggest you to join the Disc Sens controls to a single potentiometer as well if the Fe channel is also corrected, as per the schematics. Unfortunately there is a mistake on the original one.

Regarding stability you'll be better off if you keep the LF347 in gain blocks, because it has lower offset than the possible replacements. For example LM324 will have offset of about -15mV and it will ruin your sensitivity. TL064 that I use instead of LF347 has offset up to +10mV and you can't predict where, and it has somewhat worse noise, but it uses less power. It is just a compromise. You can play with replacing them. You can always revert to the original setup.

Are you maybe interested in promoting the proportional volume for the target response?

[Davor]

I'm about to investigate a better frontend for VLFs in general, but with high hopes to improve my IGSL. My guess is that it will be some kind of add-on that will be able to accommodate some wider range of Rx coils and run near the theoretical noise limit. There are a few implementations that are on my mind, some of which are using discrete BJTs to reach low noise at low cost.
Anyway, I just wonder if there is any interest here about such an add on, and if there are some wishes to try to meet. My kinda goal is to make it non-resonant, and balanced, with or without a center tap signal ground connection, with some additional high pass filtering to avoid piking up hum from the mains etc. Such frontend would also promote unshielded coils usage.

So, how about that?

[golfnut]

stability.

try wrapping a wire around the GEB and Disc pot wires and grounding it at one end.

In earlier posts, the gain of one of the amps following the det/mix were a little too high and could take off. Try less gain there with a resistor change.

S

[golfnut]

The main difference of IGSL, IDX etc to modern commercial dets - is they only detect and use half of the incoming signal.
Half of the Rx signal is discarded.


A good jump forward towards an average current machine is to demodulate both halves of the incoming signal - which doubles the Rx signal power for little or no extra noise.

The verator 3 on here has such a scheme or the Whites Patent 4783630 - sheet5
S

[simonbaker]

I'm about to investigate a better frontend for VLFs in general, but with high hopes to improve my IGSL. My guess is that it will be some kind of add-on that will be able to accommodate some wider range of Rx coils and run near the theoretical noise limit. There are a few implementations that are on my mind, some of which are using discrete BJTs to reach low noise at low cost.
Anyway, I just wonder if there is any interest here about such an add on, and if there are some wishes to try to meet. My kinda goal is to make it non-resonant, and balanced, with or without a center tap signal ground connection, with some additional high pass filtering to avoid piking up hum from the mains etc. Such frontend would also promote unshielded coils usage.

So, how about that?

Always interested in new ideas and circuits to try, especially if don't add too much complexity. We know there are newer MD designs that have more complexity.

-SB