TESORO GOLDEN SABRE

[simonbaker]

"...Instead, I turn Ground Balance until iron goes away...."

Very wrong! You should turn GEB until ferrite rod goes ALMOST away - not iron. All this performed in ALL METAL mode.
Better than ferrite rod would be thing made from small pieces of rust, black sand, hot rock and very tiny and small pieces of ferrite, mixed up togather and formed in some 5x5cm shape...somebody already suggested and explained this at very first pages of this thread...search for it.
Than turn GEB untill that "thing" goes away...and at the same time switch to Disc mode (Disc pot to "1"), and try to detect small and thin silver coin. If it is detectable than leave that setup/preset that way...if not than loose GEB a bit just to start detection on thin silver and thin foil...and that is all!
Than switch back to All Metal mode and test again on ferrite mixture....If detection is appearing and it is strong than your coil is bad - make another one !

Thank you (Thor) for taking time away from hammer to answer... but still confused!

I find that iron phase is greater than silver; is that different from everyone else? If it were before foil or nickel, I could use discriminator to eliminate, but it it is on other end of range -- right next to silver, only further. If I discriminate iron, everything else gone too, do you see what I mean?

Maybe I'm not careful enough turning GB pot, but... it seems if I can detect iron in All Metal mode, it comes through in Discrimination mode too! I will try harder...

Note: All Metal mode JFet gate signal is very triangular, not square. Is that correct? Maybe the problem is there. Also other JFet gate signals have rounded leading edge -- OK or not? I am like dog with a bone, not giving up...

Good experiments with Chinese MD, very interesting and entertaining, better than Harry Potter...you are the craziest I've ever seen but game aint over yet...Aziz and I taking lessons...

Cheers

-SB

[Qiaozhi]

Thank you (Thor) for taking time away from hammer to answer... but still confused!

I find that iron phase is greater than silver; is that different from everyone else? If it were before foil or nickel, I could use discriminator to eliminate, but it it is on other end of range -- right next to silver, only further. If I discriminate iron, everything else gone too, do you see what I mean?

Maybe I'm not careful enough turning GB pot, but... it seems if I can detect iron in All Metal mode, it comes through in Discrimination mode too! I will try harder...

Note: All Metal mode JFet gate signal is very triangular, not square. Is that correct? Maybe the problem is there. Also other JFet gate signals have rounded leading edge -- OK or not? I am like dog with a bone, not giving up...

Good experiments with Chinese MD, very interesting and entertaining, better than Harry Potter...you are the craziest I've ever seen but game aint over yet...Aziz and I taking lessons...

Cheers

-SB

Your coil is not adjusted correctly. Unfortunately, if you've already sealed it, you will have to make another one.

[Qiaozhi]

Ha,ha,ha! You are spying my work??? Perhaps you installed some bugs in my workshop??? Damn!

Yes - I have installed a lot of spyware on your computer.

[Max]

Another method is to set the ground balance control to mid-position, and adjust the coils to achieve minimum voltage, while monitoring the output of the sample gate in the ground channel. This will automatically give you the correct phase shift between TX and RX. In this case, you only need a single-channel scope, or a meter.

Hi,
that is interesting really. I expected that, cause geb channel samples at zero crossing of received signal, thus producing a minimum voltage at output of sampling gate when TX/RX phase at coil is centered on right value.

I tested that before, it works really, but had troubles with multimeter also that time... using scope I can see a distorted sinus on top of a mean value that increase/decrease when moving coil one way or another and stay at minimum voltage peak to peak at what I think is the right TX-RX phase.

In this case, also the residual phase on air is just around 200° like when measuring original coils.

Probably, I would say, this method is much more accurate cause keep track of non-linearities at oscillator and other factors that a simple "put TX/RX at 200°" will be.

That's sure not a big deal rotating the geb trimpot to compensate small inconsistencies later but that one will do using ferrite rod... having the ability of understand the "right" position where silver is still detected and ferrite almost not, and it's not so intuitive as nulling the output signal of geb sampling gate.

That's why , I think, this method is even better for those who are confused about all this stuff.

The problem is that, you compensate ferrite rod very good...and iron is not detected in disc mode (at least soft iron)... but you lose a lot of sensitivity to silver.

I found that totally nulled geb signal there means decreasing silver detection of 50% !

So, I usually retune the trimpot there to get again about full silver detection when ferrite is ALMOST nulled (means that sound at a couple of cms from overlaps)... is that right ? Who knows ?

In that position the gate output is not at zero but positive... 25-30mV.

With red bricks it works nice... but when on hot soil... red sand and the like wow ... get noise and have to reduce sens to 6 also.

This is expecially on larger coils... on 8'' or 9x8'' is not a big issue, works about good.

What do you think about ?

Kind regards,
Max

[ivconic]

Oh never mind my pc - nothing really important in it. Only dozen of trojans and few of S.Layout files...
What i suspect is that there are some buggy cameras in my workshop, behind my back!?
Yes i agree, God only knows what is phase shift at your coil Simon...???!
Make another one.
When phase shifted much out of proper angle, everything is possiblle. I had not experimented that way; to see all the possible side effects...
Usually it goes this way: ferrite,iron,thin silver,thin al foil..etc..etc.


Amtech if you want file, here it is... but seems you missunderstood me? I never traced and made copy of this nonsenced 3009..i only used its pcb shape and dimensions to draw TGSL...so if you looking for original 3009 trace...ha, i dont have it. Who need that crap anyway?

[Max]

Thank you (Thor) for taking time away from hammer to answer... but still confused!

I find that iron phase is greater than silver; is that different from everyone else? If it were before foil or nickel, I could use discriminator to eliminate, but it it is on other end of range -- right next to silver, only further. If I discriminate iron, everything else gone too, do you see what I mean?

Maybe I'm not careful enough turning GB pot, but... it seems if I can detect iron in All Metal mode, it comes through in Discrimination mode too! I will try harder...

Note: All Metal mode JFet gate signal is very triangular, not square. Is that correct? Maybe the problem is there. Also other JFet gate signals have rounded leading edge -- OK or not? I am like dog with a bone, not giving up...

Good experiments with Chinese MD, very interesting and entertaining, better than Harry Potter...you are the craziest I've ever seen but game aint over yet...Aziz and I taking lessons...

Cheers

-SB

Hi,
Qiaozhi is right: when coil is made right and you are around right phase between TX/RX you'll not get iron detection in disc mode, just all metal you'll detect it.

The geb trimpot is not for nulling iron: you MUST get it nulled by discriminator circuit when coil is made right.

The trimpot serves for what indicated : ground balance tuning

If you use a small ferrite rod you'll see that, when coil is made right way, you can null its effects at about middle position.

Problem is that interfere also with e.g. silver detection... so I experienced detuning a bit the trimpot to gain good silver detection again...but still ignoring almost the ferrite.

Iron is always rejected by good coil+discriminator circuit.

ALSO, NOTICE: when I wrote "IRON" I mean soft, pure iron... like in iron nails and the like... not steel alloys.

Sometimes people confuse soft iron with steel stuff... steels kinds have sometimes very different magnetic behaviour than soft iron things.
That's why you'll detect that thing.

An easy try: find a real iron nail stuff...5-7cm long you'll be sure it's just iron. Then find a 5cent (eur cent) coin.

They try to detect them in disc mode: on soft iron , if you made detector/coil good you'll get no signal at all. But you'll still detect the 5cents.

Why ? People think that 5cents are made of copper alloy... it isn't. It's just covered by that stuff, inside they are made of steel alloy core really detectable by MDs...

You can find other countries coins made of that metal alloy and all are detected easy, made actually of steel.

Kind regards,
Max

[Qiaozhi]

Hi,
that is interesting really. I expected that, cause geb channel samples at zero crossing of received signal, thus producing a minimum voltage at output of sampling gate when TX/RX phase at coil is centered on right value.

I tested that before, it works really, but had troubles with multimeter also that time... using scope I can see a distorted sinus on top of a mean value that increase/decrease when moving coil one way or another and stay at minimum voltage peak to peak at what I think is the right TX-RX phase.

In this case, also the residual phase on air is just around 200° like when measuring original coils.

Probably, I would say, this method is much more accurate cause keep track of non-linearities at oscillator and other factors that a simple "put TX/RX at 200°" will be.

That's sure not a big deal rotating the geb trimpot to compensate small inconsistencies later but that one will do using ferrite rod... having the ability of understand the "right" position where silver is still detected and ferrite almost not, and it's not so intuitive as nulling the output signal of geb sampling gate.

That's why , I think, this method is even better for those who are confused about all this stuff.

The problem is that, you compensate ferrite rod very good...and iron is not detected in disc mode (at least soft iron)... but you lose a lot of sensitivity to silver.

I found that totally nulled geb signal there means decreasing silver detection of 50% !

So, I usually retune the trimpot there to get again about full silver detection when ferrite is ALMOST nulled (means that sound at a couple of cms from overlaps)... is that right ? Who knows ?

In that position the gate output is not at zero but positive... 25-30mV.

With red bricks it works nice... but when on hot soil... red sand and the like wow ... get noise and have to reduce sens to 6 also.

This is expecially on larger coils... on 8'' or 9x8'' is not a big issue, works about good.

What do you think about ?

Kind regards,
Max

For real soil conditions the ground balance needs to be set slightly negative, somewhere between 0 and -5 degrees, but nulling the coil using the output of the ground channel sample gate is a good starting point. Then (as Max says) slightly adjust the GEB pot away from the ideal position.
George Payne has explained it quite well in this article ->
http://www.jb-ms.com/Baron/gb.htm

[Qiaozhi]

Another question for the experts:

R40 100K is mysterious to me. It connects output of comparators U106 (after diode D12) back to non-inverting input of comparator U106b (ground bal channel).

It causes clipping of the comparators output to near ground (actually near the non-inverting input which is the output of LM308 ).

I see similar thing on Amigo II circuit. What is the purpose? Why clip to LM308 output instead just ground for example?

Regards,

-SB

I think this feedback resistor provides some hysteresis for the comparator. i.e. it stops any spurious oscillations.
The diode/resistor combination forms a precision clamp circuit, to keep the output at a fixed value.

[tiktak]

Hi

Another method is to set the ground balance control to mid-position, and adjust the coils to achieve minimum voltage, while monitoring the output of the sample gate in the ground channel. This will automatically give you the correct phase shift between TX and RX. In this case, you only need a single-channel scope, or a meter.

Hi,
You mean pin7 lf353?

And Ivica you are rare bird but we are same blud.Smash it with the hammer!
)))
I hate those chinees too.But they are everywhere!People buy this crap for a lot of money.Strange?
And those seben are everywhere.

[Qiaozhi]

Hi,
You mean pin7 lf353?

And Ivica you are rare bird but we are same blud.Smash it with the hammer!
)))
I hate those chinees too.But they are everywhere!People buy this crap for a lot of money.Strange?
And those seben are everywhere.

No - I mean the connection between TR5 (source pin), C12 and R28.
This is the output of the ground channel sample gate.

[Max]

For real soil conditions the ground balance needs to be set slightly negative, somewhere between 0 and -5 degrees, but nulling the coil using the output of the ground channel sample gate is a good starting point. Then (as Max says) slightly adjust the GEB pot away from the ideal position.
George Payne has explained it quite well in this article ->
http://www.jb-ms.com/Baron/gb.htm

Hi,
yes, I read the article a number of times... It's really illuminating stuff...

What is not clear to me till now is about the small detune at geb trimpot required to get more sens to silver items.

When at initial point, trimpot at middle as explained, and find phase looking at minimum amplitude at gate output, I get exactly 50 kohm there at trimpot.

When I do the detune I have instead something maybe 56 or 57Kohm , the trimpot screw is rotated to the left... I get a right shift of sample gate signal and so (cause it samples on positive edge , at zero crossing, of the signal) I get a slightly positive voltage as mean value... the 25-30mV I told you above.

Now... do you think it's something -5° when I do so ?

Kind regards,
Max

[Qiaozhi]

When I do the detune I have instead something maybe 56 or 57Kohm , the trimpot screw is rotated to the left... I get a right shift of sample gate signal and so (cause it samples on positive edge , at zero crossing, of the signal) I get a slightly positive voltage as mean value... the 25-30mV I told you above.

Now... do you think it's something -5° when I do so ?

Kind regards,
Max

I think this is correct. Since all targets cause a phase-shift to the left, then moving the pulse to the right is causing it to sample slightly negative.
How much does the sample pulse move when you've made the adjustment? Is it about 1us?

[amtech2005]

TESORO GOLDEN SABRE

Oh never mind my pc - nothing really important in it. Only dozen of trojans and few of S.Layout files...
What i suspect is that there are some buggy cameras in my workshop, behind my back!?
Yes i agree, God only knows what is phase shift at your coil Simon...???!
Make another one.
When phase shifted much out of proper angle, everything is possiblle. I had not experimented that way; to see all the possible side effects...
Usually it goes this way: ferrite,iron,thin silver,thin al foil..etc..etc.


Amtech if you want file, here it is... but seems you missunderstood me? I never traced and made copy of this nonsenced 3009..i only used its pcb shape and dimensions to draw TGSL...so if you looking for original 3009 trace...ha, i dont have it. Who need that crap anyway?

Thank-You Ivconic for the sprint file. I always am very interested in what american model's the chinese copy. I wanted to know if the md3009 was the fisher 1265 or 1266 ??? If we find and identify , identical circuits or clones and the performance / depth is not good , then the chinese must be making very bad and cheap quality search-coils . I have not seen anything new or great from china . They are not going to make good money on garbage . A company called Famous Trails based in california sells 3 or 4 chinese detectors , but are way over-priced for the quality and performance . The pictures I posted below , they are made in china and sold by a big u.s. distributor , they work good , but have a very high breakdown rate . Is it the quality of parts , or the programmable chips do not hold the memory very long ??? ....Maybe someday soon we will see a real "super detector" from China and very inexpensive , but it has not happened .......... not yet anyway !!!!................Eugene

[Qiaozhi]

I hate those chinees too.But they are everywhere!People buy this crap for a lot of money.Strange?
And those seben are everywhere.

If you stop to think for a moment ... you are also copying other people's work.

[simonbaker]

I think this feedback resistor provides some hysteresis for the comparator. i.e. it stops any spurious oscillations.
The diode/resistor combination forms a precision clamp circuit, to keep the output at a fixed value.

I thought that too at first, but:

A) Why only hysteresis on one channel, not other?

B) Output impedance (low) of LM308 shunts the feedback, has virtually no effect.

"Clamping" (clipping) -- yes, but important? Wouldn't V+ be good too, just driver for audio?

I have another idea what they were originally for, but too far out to say now (probably wrong too. ).

Thanks, cheers!

-SB