GoldScan IV

[Max]

Hi

I use aluminium foil. This is my coil:
[ATTACH]6156[/ATTACH]

Now the problem is the signal became wider than unshield coil.you can see in the picture.The black is the unshield coil single.
the red is the shield coil.GS4 can't work well use this shield coil. I think the shield add too many capacitance in the coil.
[ATTACH]6157[/ATTACH]

thanks

Hi,
you're right. It's not good. But you don't require heavy shield on it.
Sometimes AL-foil is not good for your purpose... then you can choose different stuff... e.g. mylar (plastic+aluminium , very thin --> low conductive layer), graphite, copper mesh tape, copper tape, lead tape, nickel varnish etc...

Most of these things are sold for shielding RF purposes...

Depends on what you have there available, and what can find/buy. The graphite is easy... you can find varnish both in spray cans or even made yourself (you MUST have chemical skills to do it... but it's possible with right stuff).

Sure, if don't solve that problem you'll get less or no sens to lowest conductive metals, e.g. small gold things.

Kind regards,
Max

[zzy]

Hi,
put at GND.

Kind regards,
Max

Hi

I use aluminium foil. This is my coil:


Now the problem is the signal became wider than unshield coil.you can see in the picture.The black is the unshield coil single.
the red is the shield coil.GS4 can't work well use this shield coil. I think the shield add too many capacitance in the coil.


thanks

[Max]

Hi
The shield connect to where? GND or -12V?

Hi,
put at GND.

Kind regards,
Max

[zzy]

Hi
The shield connect to where? GND or -12V?

[Max]

Hi,

I finish the GS4, Very good sensitive,I can clearly detect a 20 euro cent at 30cm use a 8'' coil. I modify it a non-motion machine.

Now, the GS4 is not very stable. I should adjust it ofen.

Does the coil for GS4 need shielding?My coil is not shield.

The other disadvantage is the battery is 14.4V. This battery is not easy to buy. I think maby my battery makes it unstable.

Hi,
as I told you! detect 1eur at 30cm with 8'' with no efforts at all! Same for the 20cents...

With lot of efforts you can gain more depth even to extremely low conductive targets...

The coil is a critical piece of it: you need shield for use in the real field. Use mylar, aluminium foil or graphite depending on what you can obtain there for cheap.

The stability issue is related to mosfet mainly... but battery kind has a big, if not enormous, effect on device stability... like with other PIs too BTW.

You can use NiMh of large capacity to reach 14.4v, other solutions are possible but a bit more complex like a DC-DC coverter from 12V to 14.4V... then run over a gel SLA battery of 12V nominal.

Remember also that NiMh drop rate is really important issue... so if you plan using them you need at least 1800mAh ones... otherwise you'll get many troubles from cells voltage drop.

Another, simpler, option could be using a pack of 3 batteries of 6V 0.8-1.2Ah in series to gain 18V... then using a linear shunt regulator like LM317 to get the 14.4v stabilized.

Kind regards,
Max

[zzy]

Extreme good sensitive!!!

Hi,

I finish the GS4, Very good sensitive,I can clearly detect a 20 euro cent at 30cm use a 8'' coil. I modify it a non-motion machine.

Now, the GS4 is not very stable. I should adjust it ofen.

Does the coil for GS4 need shielding?My coil is not shield.

The other disadvantage is the battery is 14.4V. This battery is not easy to buy. I think maby my battery makes it unstable.

[zzy]

Hi Max,
I found this machine is hard to adjust.It's very unstable.
Do you know how to operate this machine?First,i turn the "delay" knob.Second,"GND" knob and then "SENSE" knob.
Are this right?
The "GND" knob is design for what?
I try to change damping resistor,no effect.still have noise.

[Max]

Hi Max,

I use 2SK435, the pin is differ to j113. Does the jfet's "D" and "S" pin can interchange? Some people said the the jfet's "D" and "S" pin are same.

Hi,
your fets are good replacement; you can swap D and S cause it's bidirectional device...

Probably problem is elsewhere.

Kind regards,
Max

[zzy]

Hi,
the 15uS there are good. In the pictures I see ringing in last part of signal evenlope... you'll better avoid that, selecting appropriate value for the damping resistor.
The fact you get low sens could be related to some other mistake in pcb maybe... cause about timing it seems good.

I will try to follow signal path and see if I'll get good response e.g. at integrator outputs when nearing metallic stuff.

This could help insulate the fault.

PS: what about the jfets ? did you use j113 or different ?

Kind regards,
Max

Hi Max,

I use 2SK435, the pin is differ to j113. Does the jfet's "D" and "S" pin can interchange? Some people said the the jfet's "D" and "S" pin are same.

[Max]

Hi Max,
This is my measure of the delay time.Sorry my bad camera.
I measure the Q5 to Q8
[ATTACH]5854[/ATTACH]
[ATTACH]5855[/ATTACH]
Q5 about 75uS delay
[ATTACH]5856[/ATTACH]
Q6 about 60uS delay
[ATTACH]5857[/ATTACH]
Q7 about 200uS delay
[ATTACH]5858[/ATTACH]
Q8 about 15uS delay

I change the 1n5818 to 1n4148. no effect.

Hi,
the 15uS there are good. In the pictures I see ringing in last part of signal evenlope... you'll better avoid that, selecting appropriate value for the damping resistor.
The fact you get low sens could be related to some other mistake in pcb maybe... cause about timing it seems good.

I will try to follow signal path and see if I'll get good response e.g. at integrator outputs when nearing metallic stuff.

This could help insulate the fault.

PS: what about the jfets ? did you use j113 or different ?

Kind regards,
Max

[zzy]

Hi Max,
This is my measure of the delay time.Sorry my bad camera.
I measure the Q5 to Q8


Q5 about 75uS delay

Q6 about 60uS delay

Q7 about 200uS delay

Q8 about 15uS delay

I change the 1n5818 to 1n4148. no effect.

[Max]

Hi Max

I use coil: 436uH, resistance 2.2ohm Diameter is 20cm.

I alse find in all metal,some metal can't detect at all.

Do you konw the channel1 and channel2 sample at what?

I find the channel1 sample point is too late.Is the problem there?

I can't find diode BAT85,I use a 2n5818 instead.

I also can't find diode OA85 and 1N94. Do you konw which can instead of them?

Hi,
let me check about coils data... yes seems about right... resistance is a bit low.

about diodes: you don't need germanium, use small shottky diodes instead , don't use the (probably you meant) 1n5818... try to find out something bat48 or bat86 if you can... or bat54, even better...much better bat54 as replacement! If you cannot find them all just use common 1n4148 or 1N914 etc common silicon diodes there, not best choice but better than using 1N5818 there, the 1n5818 are fast rectifiers 1A... no good at timing section.

About timing... first delay is extremely important: you can go down in the 15uS range for it, check with scope for it, measure from gate/mosfet switchoff.

If is higher you'll miss low conductive stuff... e.g. gold foils... very thin mylar or aluminium etc but still detect larger things like thick rings.

Don't need germanium, instead of OA stuff you could use another shottky barrier there, like bat85, bat 48 etc small signal type.
If you can find... OA90 will work nice there... but any shottky small signal diode will do.

PS: things to look at are
1. offset of NE5534, tune the trimpot there
2. the damping resistor influence... too overdamping of the signal could give lower detection

Kind regards,
Max

[zzy]

Hi,
you probably have some mistake there... check very carefully... cause device is hell sensitive, you can belive me...I've made it long time ago... you can get 1eur at 30cm pretty easy with a 20cm coil , without any effort at all I mean.

Actually device can be tuned so sensitive that you'll never belive till you see that but that requires much efforts expecially on coils.

If you saturate the integrators it will require much time to recover threshold... but in these cases you'll use the reset to make things go quickly.

What about coil you use ? Inductance ? Resistance ?

Kind regards,
Max

Hi Max

I use coil: 436uH, resistance 2.2ohm Diameter is 20cm.

I alse find in all metal,some metal can't detect at all.

Do you konw the channel1 and channel2 sample at what?

I find the channel1 sample point is too late.Is the problem there?

I can't find diode BAT85,I use a 2n5818 instead.

I also can't find diode OA85 and 1N94. Do you konw which can instead of them?

[Max]

Hi Max,

Sorry my bad observe.The fact is this:
I remove the two resistors. GS4 becomes a non-motion machine,
but it's beating frequency increases gradually.Only you push the "Reset" button,the beat'll stop 3 second. Then it's beating frequency increases gradually.

Hi,
you probably have some mistake there... check very carefully... cause device is hell sensitive, you can belive me...I've made it long time ago... you can get 1eur at 30cm pretty easy with a 20cm coil , without any effort at all I mean.

Actually device can be tuned so sensitive that you'll never belive till you see that but that requires much efforts expecially on coils.

If you saturate the integrators it will require much time to recover threshold... but in these cases you'll use the reset to make things go quickly.

What about coil you use ? Inductance ? Resistance ?

Kind regards,
Max

[zzy]

Hi Max,
Just like you said,i remove the two resistors. GS4 becomes a non-motion machine.Very thankful!
And i find that when a metal object near the coil, the beating frequency will decrease, when more near the coil a little,the beating frequency increases rapidly,and then turn very sharp tone,when i remove the metal, the sharp tone will delay about 15 second to the original beating frequency.


I test it in all metal mode. Detection is poor, I use a 20cm diameter coil,
20 euro cent is at 10cm.

Hi Max,

Sorry my bad observe.The fact is this:
I remove the two resistors. GS4 becomes a non-motion machine,
but it's beating frequency increases gradually.Only you push the "Reset" button,the beat'll stop 3 second. Then it's beating frequency increases gradually.