TESORO GOLDEN SABRE

[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

[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

[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

[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

[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]

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]

"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!

[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

[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

[simonbaker]

"...Sensitivity pot act like.. well actually doesn't act a bit...."
There is not any "secret" schematic of TGSL

That was joke of course -- secret is experience and knowledge.

Regards,

-SB

[eclipse]

I'm gonna slow the things now with this board by making second for testing purposes, with copper layer much bolder, this one is ~35um which seems bad when you start desoldering components. First and last time I'm buying such crap, from now only 70um and up.

I rushed the things too much, I need to detail study and scope all. Also in my mind rotates the idea of first making TGS like all of you guys, I mean no innovations. Why? As I told before

"I fixed some stuff making troubles, found others."

I'm risking here to become annoing with questions, while I can avoid some of
them if I make same like your boards. All those new IC, perhaps some of them don't work as good as expected. Up to now I have the feeling that LT1057 is
not good somehow, though is improved replacement of LF353.

A tip: with adjusted coil almost all scope points are different (I'm mainly looking at those from Max, pg 57 ).

From the observations I noticed that PN2222 - Fairchild (plastic)is much better than 2N2222 - Thomson (metal) - ( I switched the one connected to 4024 with 1K, which was reversed ). Could be single unit, problem or it's old ( damn it really looks old).. but still we don't get enough for what we're paying.. always got to look for deffects.

Cheers.

[Max]

I'm gonna slow the things now with this board by making second for testing purposes, with copper layer much bolder, this one is ~35um which seems bad when you start desoldering components. First and last time I'm buying such crap, from now only 70um and up.

I rushed the things too much, I need to detail study and scope all. Also in my mind rotates the idea of first making TGS like all of you guys, I mean no innovations. Why? As I told before



I'm risking here to become annoing with questions, while I can avoid some of
them if I make same like your boards. All those new IC, perhaps some of them don't work as good as expected. Up to now I have the feeling that LT1057 is
not good somehow, though is improved replacement of LF353.

A tip: with adjusted coil almost all scope points are different (I'm mainly looking at those from Max, pg 57 ).

From the observations I noticed that PN2222 - Fairchild (plastic)is much better than 2N2222 - Thomson (metal) - ( I switched the one connected to 4024 with 1K, which was reversed ). Could be single unit, problem or it's old ( damn it really looks old).. but still we don't get enough for what we're paying.. always got to look for deffects.

Cheers.

Hi,
had trouble too with 35um... also the problems Ivconic's explained : small scratches on the surface made them tricky to debug... almost invisible break in the tracks!

I found that many of these defects originate in bad storage procedures at dealers warehouse...

In some of these presensitized UV boards I use I noticed a lot of scratches on the adesive film that covers the UV sensitive varnish layer: that are the troubles... the scratches propagate even under the protective film and then is very easy the varnish will result broken cause of that.

When you etch the board seems everything ok...but it's isn't. I found that checking all them is too time wasting... so I throwed away them all remaining and bought new from another dealer with improved protective film and more thickness too.

The 35um are too sensitive to storage problems and scratches on the varnish... thus lead easy to circuit failures really hard to find later: this is expecially when you etch boards in very fast solution... that dissolve anything in 6-8 minutes!

Kind regards,
Max

[amtech2005]

TESORO GOLDEN SABRE

If the schematics we use are close to accurate , and the parts we use are new , or within tolerance . 1. Then noise or chatter is caused by poor PCB quality or Oscillations caused by improper pcb layout or ground circuit traces not right . 2. If there was a way to do a signal to noise test without the search coil hooked up but using only a resistive rx and tx load on the Golden sabre light and then apply power and find out where the self oscillations or external interference is causing our Project Problems , that would be great !!..........I am very optimistic that the Tesoro Golden Sabre Light problem's will be solved very soon.......................Eugene

[aft_72005]

Max

If you'll not remove that short circuit at coil your MD will never work any good.

Hi Max
You are right , but I didn’t have short short circuit on coil.
The faraday shields at the transmitter . And receiver coils open in the region
Of gap .
Please more explain .
Best regards.

[ivconic]

If the schematics we use are close to accurate , and the parts we use are new , or within tolerance . 1. Then noise or chatter is caused by poor PCB quality or Oscillations caused by improper pcb layout or ground circuit traces not right . 2. If there was a way to do a signal to noise test without the search coil hooked up but using only a resistive rx and tx load on the Golden sabre light and then apply power and find out where the self oscillations or external interference is causing our Project Problems , that would be great !!..........I am very optimistic that the Tesoro Golden Sabre Light problem's will be solved very soon.......................Eugene

Not to worry much if chatters appear only when Sense is highly adjusted. Cose i saw same behavior on original Vaquero and Tejon, same story.
Somehow i managed to reduce those chatters by evolving pcb. Last one drawn for 1265 enclosure seems to "produce" least possible chatters. I made one device for my own needs. Oftenly go outdoor and search for items. On most of the soils no chatters at all when Sense on "9" and sometimes on "10". Also coil is very important. I made very small coil for some hard and pretty slope terrains, 18cm diamm. coil. So i noticed high stabillity with that coil.
Most of the time i work with maximum sense, using that coil. That coil air tests are modest comparing 27cm coil, 1e maximum at 26cm ...etc...etc.. but rock stabille on very hard soil, as showed to me already.
So actually, i dont think that we have problem here with schematic accuracy. It is more like the method of making detector and coil. Every step is important. Every "small" thing is actually very important.
Badly nulled coil can also cause often chatters and crack in audio at detection borders...
However, fact is that TGS audio is not state of the art, and need major improvement.
Regards...