TESORO GOLDEN SABRE

[Max]

I fixed some stuff making troubles, found others.
Found how I can measure phase shift correctly, it was the scope doing that.
It was keeping the signal steady and sharp. Swiched off the god damn option. Gets tuff to measure but at least works. Glad I found that finally.
Strange thing is that when reading calibrating signal everything looks Okay, correct dual channel mode.

I did a quick bench test, after all the optimum values which I was able to achieve correct phase of the signals maintaining low amplitude was ~13mV (2xDiv, 5mV/Div). Going lower gets hard to achieve correct phase.
Do you think is good enough, and how good this"good" it is.

Measuring without scope seems hard, w/o seeing what actually is going on, though many people do that.
And my guess is adjusted coil with scope to certail level, lets say 5mV does't actually mean that the multimeter will read 5mv, perhaps it will show
the trivial 0.0xxx I got two (2) unaccurate unrealiable multimeters, so I did't use them for adjustments. I'm glad I can avoid that.
If it works use it

Adjusting i do like this:
Connect 1 channel of the scope to Rx, using just the cap 15nF
Second channel is detached, any try to use the second channel - bang!, wrong readouts like
I first experienced hardness to maintain low voltage level
After I get close to the 5-10mV I switch on second channel, bang - the voltage goes crazy unaccurate, but who cares.
Thus I try to keep voltage minimum rate like it was, while adjusting shifting.
That's all. Hard it will be to keep this when using the epoxy, but I am still doing tests,
I don't care about that right now. Still work lays ahead.


Regards

Now seems good.

Test discrimination... it will probably reject soft iron now.

Kind regards,
Max

[eclipse]

I fixed some stuff making troubles, found others.
Found how I can measure phase shift correctly, it was the scope doing that.
It was keeping the signal steady and sharp. Swiched off the god damn option. Gets tuff to measure but at least works. Glad I found that finally.
Strange thing is that when reading calibrating signal everything looks Okay, correct dual channel mode.

I did a quick bench test, after all the optimum values which I was able to achieve correct phase of the signals maintaining low amplitude was ~13mV (2xDiv, 5mV/Div). Going lower gets hard to achieve correct phase.
Do you think is good enough, and how good this"good" it is.

Measuring without scope seems hard, w/o seeing what actually is going on, though many people do that.
And my guess is adjusted coil with scope to certail level, lets say 5mV does't actually mean that the multimeter will read 5mv, perhaps it will show
the trivial 0.0xxx I got two (2) unaccurate unrealiable multimeters, so I did't use them for adjustments. I'm glad I can avoid that.
If it works use it

Adjusting i do like this:
Connect 1 channel of the scope to Rx, using just the cap 15nF
Second channel is detached, any try to use the second channel - bang!, wrong readouts like
I first experienced hardness to maintain low voltage level
After I get close to the 5-10mV I switch on second channel, bang - the voltage goes crazy unaccurate, but who cares.
Thus I try to keep voltage minimum rate like it was, while adjusting shifting.
That's all. Hard it will be to keep this when using the epoxy, but I am still doing tests,
I don't care about that right now. Still work lays ahead.


Regards

[ivconic]

"30 cm" is not problem to achieve. I achieved 44cm on 1e coin with latest 2-3 TGSL's. And 48cm on Cu coin 2.5cm diamm.
All the geag is in pretty precise coil balancing and nothing else.
I will suggest here something which (as experienced and selfrespecting enthusiasts shouldnt!) can provoke you to laugh. But try to balance coils without any instrument, using just 1e coin and observing air "depths".
Do not use scope or milivoltmeter, just bare eyes and ears.
Move coils slowly and test with coin, move again a bit, again....and when you achieve best "depths" than measure residual voltage on RX...
Recently ( yesterday) i discovered that and was pretty disapointed - using that method i balanced coils much better than in previous cases when coils were nulled using instruments!!!??
At 200mV AC range on my multimeter usually i tried to achieve 0.00 or 0.01. Than i knew coil was balanced properly.
But when nulled without instruments and achieved best possible air "depths", later i disconnected RX from pcb and connected to 200mV ac and guess what i measured??? Ha,ha,ha ... 0.05 !?!?!?

This is obvious proof that multimeter is not accurate at all! All the time i balanced my coils slightly ofset!?
That's why my coils "choked" iron to much! I always wandered what is the problem! Sheeeeesh! Now i have 0.05 as reference for the future.

Troy Shadow X3 coil is cause of all my problems. Last year i wrongly assumed that right residual voltage should be same as i measured on that coil.
At the time i measured that coil and result was 0.00.
Troy also works in range 18-19kHz. Damn!
So..do not try to null coil to reach "zero" on lying multimeter. It is also hard to null with scope, but main benefit using scope is cose you can cleary see the moment when phase is switched.
For a start; use eyes and ears only and after that check results with scope - you will be amazed!
Cheers!

[Max]

You didn't know Max and Ivconic have secret schematic for 30 cm?

23 cm not bad in real world, but 30 cm is standard here so we're disappointed if not get it.

Are you using schematic with "threshold" pot, or fixed resistors on final LM358? Threshold pot acts a little like sensitivity control, must be played with also.

Where is your sensitivity pot set? Is it at max sensitivity, or must you reduce it because of noise? In other words, is noise the problem, or you just don't hear any target over 23 cm at max sensitivity?



Do that mean changing leads or moving coils for nulling? Are your coils glued already? I forget, what size are your coils?

What frequency is your oscillator?

Usual checklist: faraday cage on coils has gap, insulated from each other?

Can you post pictures of coil and cables and any shields (if not sealed up)?

Your tests will help us all tune our MDs.

Cheers,

-SB

Hi,
if use phase reversal he will get about phase required... swap leads! We talk of this many times before.

We have no special setup: the circuit is that of schematic, the board could vary (this could change S/N ratio of course) but is always about the same... just extra mods you can use or not... like transistor audio or monostable... is your choice.

Happens that when it works the right way you hear the cracks and broken audio of e.g. far coins... very weak audio on the 16ohm speaker: if speaker is NOT GOOD you'll simply get NO AUDIO for that.

That could be a reason of failure in detecting stuff at 30+ cm ... easy solution: use an headphone set, good quality (100-150ohm is good for TGS too using impedance matching resistor, or you can use large headphones with e.g. 32-36 ohm per channel... like stereophonic cheap -large- ones)

You'll hear the cracks and sounds much better and unlike when there's a bad speaker there... that will mask anything.

Kind regards,
Max

[ivconic]

"...Sensitivity pot act like.. well actually doesn't act a bit...."

Once i had same situation. After 2-3 days nerving and torturing myself to find a problem i almost gave up. Just by lucky coincidence i found out small crack on pcb - one trace was broken, almost impossible to spot with eyes. Use VU scale ommmeter and fast "run" over pcb and traces...you will find simillar problem - i am sure.
Pretty often, turning Sense pot will not show any changes if coil is mismatched much also. Personally i have doubts that this is your problem. I rather beleive that you have broken trace.

There is not any "secret" schematic of TGSL. I posted my working versions here descently.
Sometimes i hurry much to post here some version without detailed checks, so in few cases i posted schematics and pcb's with mistakes - totally unintentionally, sorry if you had problems with those.

[Max]

Simonbaker,
Sensitivity pot act like.. well actually doesn't act a bit.
I can't see any difference when rotating the pot, already checked it's connected and working, but does not seems to react.
--
I Am using the schematic w/o the threshold, with fixed level.
Actually my PCB is based on Ivconic's board posted in "gift-pack"..
except the missing 1K resistor, the up-side down transistor in oscilllator and
the thresold part everything looks the same as the schematic.
--
The frequency of the oscillator is about 14.6-14.7 Khz
I am using several different set of Tx and Rx coils
One one set didn't have shield on them, the other have shield and the gap.
--
Important question for me.
How your phase shift appears when moving coils?
Does it "goes" all around the 360 degrees?
Because mine jumps a bit with about 30 degrees +/- and
nothing more I could do to change it.
That's measuring Tx (leads connected to GND and Tx connector 1)
and Rx (leads connected to Rx connector 1/2 and also when measuring Pin7 on LF353 - both cases same problem. Wrong totally unadjustable phase)
I have the feeling that these are scope problems. I doubt observation of JR. like me could help you guys, perhaps avoiding stupid mistakes only
--
Check out the photos:
The smaller signal is Rx, it moves only between the 2 positions as on the photos. Haven't seen more than 90 degrees shifting yet. Weird stuff.

Cheers

Hi,
the problem is at coil. The phase shift is no good as I see in pictures. If you swap RX leads you get 180° reversal appears : you must keep coil in same orientation like you have already and just reverse the leads.

If you have 30° you'll get 210° and so on.

It's obvious you get small phase shift when normally move the coils... but when you're around nulling your phase shift will change much, not few tens of degrees.

You have to match Qiaozhi's picture to get coil about in right phase shift.

BTW phase shift is not a severe limitation to depth in all-metal mode... so you'll get good detection even with not perfect value... the fact you get detection in an indication that the MD works, but then to get hi performances you have to check many things like preamp and fets.

I remember that one time I reversed gate leads at fets (cause all we use different ones and I confused BF245 with J107) but STILL I get detection using a good (already made) coil... around 15cm for a large coin!

So coil is really important and when other stuff is ok (like right connected fets! in my case) you get very good response... 30cm etc

Kind regards,
Max

[eclipse]

Simonbaker,
Sensitivity pot act like.. well actually doesn't act a bit.
I can't see any difference when rotating the pot, already checked it's connected and working, but does not seems to react.
--
I Am using the schematic w/o the threshold, with fixed level.
Actually my PCB is based on Ivconic's board posted in "gift-pack"..
except the missing 1K resistor, the up-side down transistor in oscilllator and
the thresold part everything looks the same as the schematic.
--
The frequency of the oscillator is about 14.6-14.7 Khz
I am using several different set of Tx and Rx coils
One one set didn't have shield on them, the other have shield and the gap.
--
Important question for me.
How your phase shift appears when moving coils?
Does it "goes" all around the 360 degrees?
Because mine jumps a bit with about 30 degrees +/- and
nothing more I could do to change it.
That's measuring Tx (leads connected to GND and Tx connector 1)
and Rx (leads connected to Rx connector 1/2 and also when measuring Pin7 on LF353 - both cases same problem. Wrong totally unadjustable phase)
I have the feeling that these are scope problems. I doubt observation of JR. like me could help you guys, perhaps avoiding stupid mistakes only
--
Check out the photos:
The smaller signal is Rx, it moves only between the 2 positions as on the photos. Haven't seen more than 90 degrees shifting yet. Weird stuff.

Cheers

[simonbaker]

Bad news for me I tryed it and didn't work.
I can't affect phase shift no matter what I do.
Moving Rx and Tx gives me fixed shifting on 30 and 60 degrees (or whatever - it's far from the 200 degrees).
Wrong phase - no disc, bad stability also on tested stones (mineralized rock).
Bad news for depth lover like myself - nulled coil to around 2-3mV gives me depth for a huge coin (1 eu) at ~23 cm, really poor result.

Perhaps I was wrong changing stuff first, I needed to make one standard TGS with standard components and board. Now even if I am close to heaven just have no clue on how to reach it. Maybe PCB error?
Any volunteers for checking the board (sprint 5) which I used?
I'm deeply stucked in the middle of nowhere

You didn't know Max and Ivconic have secret schematic for 30 cm?

23 cm not bad in real world, but 30 cm is standard here so we're disappointed if not get it.

Are you using schematic with "threshold" pot, or fixed resistors on final LM358? Threshold pot acts a little like sensitivity control, must be played with also.

Where is your sensitivity pot set? Is it at max sensitivity, or must you reduce it because of noise? In other words, is noise the problem, or you just don't hear any target over 23 cm at max sensitivity?

Moving Rx and Tx gives me fixed shifting on 30 and 60 degrees (or whatever - it's far from the 200 degrees).

Do that mean changing leads or moving coils for nulling? Are your coils glued already? I forget, what size are your coils?

What frequency is your oscillator?

Usual checklist: faraday cage on coils has gap, insulated from each other?

Can you post pictures of coil and cables and any shields (if not sealed up)?

Your tests will help us all tune our MDs.

Cheers,

-SB

[amtech2005]

TESORO GOLDEN SABRE

Hello , looking at the latest schematic posted by aft_72005. A few questions related to the chatter or noise. 1. Does anybody see any problems with the 12 volt supply going directly to the LM358 . 2. And is the foil shielding of the search coil connected to ground . 3. Is the noise present only in europe 50 cycle ac. If the noise is not from anything I listed then it must be in the PCB circuit placement of parts or layout/ ground plane. aft_72005 where did you find schematic you just posted ?? , what date did you find it on ??????...........Eugene

[Max]

A question for Max, Ivconic, Eclipse, anyone with good working TGSL:

When you turn sensitivity pot to most sensitive and test outdoors, what do you hear? Is it quiet? Or is there chatter?

If quiet, would it be worth trying to put a jumper across R37 (i.e. remove it) to allow reference all way to ground?

Right now, min sensitivity reference is 8 v / 2441 = 3.3 mV.

If your target signal is 4 mV, you only hear .7 mV top part. If less than 3.3 mV, never hear. So lose some sensitivity.


But if noise is approx 3.3 mV, nothing to do.

So it depends on noise. Maybe your PCBs are very low noise. Have you tested?

Cheers,

-SB

Hi,
it's difficault to answer that question... it's not always the same.

I talk about my experience. When I carry outdoor my TGS far from interferences I get it absolute quite, no chatters at sens to 10. This happens most of the times.

Sometimes, always outdoor, I switch on an get occasional chatters: have to say that in these places there are RF interferences, so are not ideal places to search with it.

That I reported is for coil far from soil.

Things that happens when my TGS is set to max sens on soil are similar: in much places I can run it quiet, just very rare, occasional chatters (who knows reason ??? maybe just internal instability) BUT other times well... I simply cannot lower the coil to soil... cause I get plenty of noises , chatters, whatever!

At first seems I've done bad geb tuning... but chatters stay on also with coil NOT IN MOVEMENT over soil... seems like radiated rf from soil enters the TGS amplification chain and show very strange interferences there!

Infact, these situations happen in that places I said above, with RF noise around from big power transmitters at few Kms.

In that last scenario is impossible stay at 10 of sens... have to reduce sometimes to 7 or even less.

On some very worse soil it happens also that noise is triggered by ground (too many iron minerals and imperfect geb I think) so have to put sens at just 6.

I don't tested the remove of R37 so far... but will try and report about... sure it couldn't be used in the worse situations... but when quiet it could be tested.

Really... when I set sens to 10 and get silent at home (not frequent situation , cause too electrical noise around) I can "detect" a switch commutation from a room 15meters far: TGS is a real beast of sensitivity when noise around is very low. Seems an LRL!

Kind regards,
Max

corection

Hi,
C7 on the picture is 10pf ,not 100pf.

[eclipse]

Hi,
if you get 30° you can reverse leads at RX and get 180+30 = 210° that's about what you need to work good.

Kind regards,
Max

Bad news for me I tryed it and didn't work.
I can't affect phase shift no matter what I do.
Moving Rx and Tx gives me fixed shifting on 30 and 60 degrees (or whatever - it's far from the 200 degrees).
Wrong phase - no disc, bad stability also on tested stones (mineralized rock).
Bad news for depth lover like myself - nulled coil to around 2-3mV gives me depth for a huge coin (1 eu) at ~23 cm, really poor result.

Perhaps I was wrong changing stuff first, I needed to make one standard TGS with standard components and board. Now even if I am close to heaven just have no clue on how to reach it. Maybe PCB error?
Any volunteers for checking the board (sprint 5) which I used?
I'm deeply stucked in the middle of nowhere

[simonbaker]

sensitivity

A question for Max, Ivconic, Eclipse, anyone with good working TGSL:

When you turn sensitivity pot to most sensitive and test outdoors, what do you hear? Is it quiet? Or is there chatter?

If quiet, would it be worth trying to put a jumper across R37 (i.e. remove it) to allow reference all way to ground?

Right now, min sensitivity reference is 8 v / 2441 = 3.3 mV.

If your target signal is 4 mV, you only hear .7 mV top part. If less than 3.3 mV, never hear. So lose some sensitivity.


But if noise is approx 3.3 mV, nothing to do.

So it depends on noise. Maybe your PCBs are very low noise. Have you tested?

Cheers,

-SB

[simonbaker]

This circuit uploaded in this thread at the past
And again .

Yes, good to post again, I had not seen this variant with audio switch and no threshold pot together.

-SB

[simonbaker]

The hysteresis setup with the TGSL is correct. It should be connected to the non-inverting input. Please read the attached document.

Yes, I agree, it is correct for hysteresis. (And good reference, thanks.) Reason I care is because we can't be sure of schematic for TGS, I try to investigate areas that look unusual.

With hysteresis, some things look funny:

1) only on one channel.

2) unknown amount. Referencing your document, it is classified as "non-inverting hysteresis, dual supplies" design. Resisitor R1 in the document is the output impedance of the LM308 stage -- did Tesoro designers measure that? Is it a reliable way to set hysteresis level?

3) How did Tesoro decide how much hysteresis to have? Hysteresis will create another threshold on the ground channel side in addition to the sensitivity control. If too much, hurts sensitivity. Other side not have it at all, different sensitivity?

Just want to understand it fully in case I work on mod that would affect it.

I appreciate your explanations and patience!

Regards,

-SB