TESORO GOLDEN SABRE

[Mullihaka]

Unwanted oscillation

Hi Andy,
seems you have a little problem there : the 700Khz you read are probably from self oscillations at that op amp.

The fact you read 130mA consumption is not so good... I think you've it damaged already.

My suggestion is , apart cleaning very well pcb as you can, desolder it... then remount the socket; try to add 100nF caps just under the ic socket at power rails and other lead to gnd.

That way you'll probably solve that, but have to use a brand new ic.

Other things that could be involved are pcb tracks... too near can be a problem and sometimes op amps use special ring guards for that at inputs (yes, the amplification factor plays a major role in that but not only)... but if you used any of the pcb designs here you'll probably need just that filtering at power rails.

Kind regards,
Max

I use hte PWB design attached in the first message of this thread. The bipolar capacitors are constructed of two 10u electrolytics in series, minus poles connected together.

The current consumption is 60mA. Is that correct?

There are 20mV ripple at 700kHz on the +8V rail, riding on 20mV ripple at the transmitter's frequency. How clean did you get that?

I have tried bypassing the 8V regulators in/output with 100n and the same for the PNP transistor's emitter of the negative voltage generator. No help.

I disconnected the 470n cap from pin 1 of the amp, that didn't help. Disconneting the 100k resistor from pin 3 did nelp a bit. I have a feeling, that the power rails should be cleaned before trying anything else. I can find ripple everywhere in the device, just can't find the source. Could the ripple be born somewhere else and it jsut get's amplified by the amp in question?

Andy
I have tried bypassing the

[metaldetector107]

Mayday mayday mayday

Dear Ivconic????
i am sorry but i am realy confused
i know you dont like repeating question
but this time please dont
could u tel me the thread page of your coil wiring
if u cant remeber please trouble yourself and make another one for me
you it is realy good favor;;
beacuse i know i made mistake like till your last reply to me i was connected both
shield to PCB but still any slight movement sounds
abou the cable I am using a PI metal detector cable called P-158
the cable has two pairs of wire and each pair has its own shield
many many thanks
by the way iam not lazy i have downloaded all 137 pages of TGSL
abd reading them daily but still i am hopless with you and MAX and other gentlments

[simonbaker]

Hi,
yes, clear, that's my experience too: the gnd is due to coil internal connection not as reported at schematic level in wrong diagrams.

Ivconic's one is right; if someone use e.g. double gnd connection (at pcb level and at coil too) it will generate a gnd loop... with increase of noise impact due to the current that will flow in that closed loop in the cable.

That will strongly affect circuit performances cause of noise signal flowing there and thus propagation of it to other parts of circuit, cause of the very high amplification factors in the chain.

Then , also, the cable will show more sensitive to microphonic also, like you described, so will expose user to false signals at every, even small, vibration.

Kind regards,
Max

Thanks all for clarifying.

That ground in coil head is hard to understand theoretically, but if it works, it works. You would think at first it would ruin the "differential" effect of the RX leads and create more differential noise. Also, you would think it would create a large ground loop if you follow ground from oscillator to RX input. But if it works....

Regards,

-SB

-SB

[Max]

Hi Simon,

It's the inverting input of the preamp opamp that is connected to the screen. Some back-engineered Tesoro schematics show a ground connection in the receive circuit (as in your post) but this is incorrect. The connection to ground (0V, screen) is actually made in the search head.
Ivconic's diagram shows the correct wiring for the coil.

Hi,
yes, clear, that's my experience too: the gnd is due to coil internal connection not as reported at schematic level in wrong diagrams.

Ivconic's one is right; if someone use e.g. double gnd connection (at pcb level and at coil too) it will generate a gnd loop... with increase of noise impact due to the current that will flow in that closed loop in the cable.

That will strongly affect circuit performances cause of noise signal flowing there and thus propagation of it to other parts of circuit, cause of the very high amplification factors in the chain.

Then , also, the cable will show more sensitive to microphonic also, like you described, so will expose user to false signals at every, even small, vibration.

Kind regards,
Max

[Qiaozhi]

Are these pictures you mean Ivconic?

(Grounding of RX lead through coil shields is curious.)

Can you confirm which of the RX inputs is the "grounded" wire? The one to the op amp inverting (-) or non-inverting (+) input?

Regards,

-SB

Hi Simon,

It's the inverting input of the preamp opamp that is connected to the screen. Some back-engineered Tesoro schematics show a ground connection in the receive circuit (as in your post) but this is incorrect. The connection to ground (0V, screen) is actually made in the search head.
Ivconic's diagram shows the correct wiring for the coil.

[simonbaker]

Are these pictures you mean Ivconic?

(Grounding of RX lead through coil shields is curious.)

Can you confirm which of the RX inputs is the "grounded" wire? The one to the op amp inverting (-) or non-inverting (+) input?

Regards,

-SB

[ivconic]

Yes; presuming coil was balanced properly, than only cable left to be suspected. If coil was balanced well and cable is the right one....than only wirements left.
Wirements at TGS is a bit tricky. You can wire it up on few different ways, but only one guarantee good stabillity and no falses.
In search head both cages and both coils ends are wired togather and further connected to cable shield, but also another isolated wire is wired to cable shield in coil. "Hot" ends of both coils are wired separatelly.So... you will have 3 separatelly isolated wires and cable shield arround them. That is situation in search head. On other end - pcb (through connectors) you will connect cable shield to pcb gnd and TX "hot" wire to TX output. Also RX "hot" wire and that separated wire (connected to shield in search head but separated in cable) to RX inputs. Do not make any connection between RX inputs and gnd on pcb! So...RX "gnd" is connected ONLY in search head to common search head gnd (both AL cages and both ends on cable shield + one separated cable wire)....
Sheeeshh...to much confusing here!!!
That's why is better to search back pages for sketch i posted. One picture can replace 1000 words!

In the past, during all those experiments with coil and cable wirements i tested almost all possiblle combinations and few different types of cable.
Some of those do work, but only this method i just explained, provide excellent stabillity and absolute immunity.
Undesrtand...?

[Qiaozhi]

Sir Max Ivco Quzi and all
please help my device works ok TX freq is 14.7.
RX6.5mh TX6mh after sheilding which is exactly what Ivconic
said so the only problem now i have
just very very little movement the device gives signal
i have tried many many ways like nulling from 0mv to 40mv
nothing happned the stand is plastic tube the coil very stiff
and what else
please help..........
Max by the way when i read your responce to Simon about his drwing
10minutes laughing
thanks all

OK - from your description it is probably one of two things:

1. As Ivconic says, make sure you're using the right type of screened cable to connect the coil to the control box. Certain types can become microphonic, and produce noise when moved.

2. It sounds from your description of nulling the coil, that you have not understood the importance of getting the correct initial phase-shift between TX and RX. My suspicion is that this is most likely the problem. If you set up the coil correctly, and the noise is still there, then try the other suggestions that have already been provided.

[simonbaker]

Hi,
that's not a solution !

Never decrease that sensitivity.... work always at 110% like russian reactors!

Returning serious...

Decreasing sensitivity is good when e.g. on soil... one have too much troubles at max sens then reduce... it.

But if device is quiet then you move it and sounds.... there's something wrong I think.

I think he has a mistake on either mechanical stability of the assembly... or electrostatic related "ghost" (means that pcb reads some static and thus give a false signal... if box is not at a potential like in the circuit... it will not act as faraday cage... but like the plate of a capacitor!)

Kind regards,
Max

Not as solution, as diagnostic. I want to know how high the movement signal is. So turn sensitivity to make less sensitive, does movement signal stop at some point, where?

From his other posts it seems he/she has a super-sensitive detector, 40 cm???

I agree sounds like mechanical. I would like to see photos - is his cable free to swing? Any suspicious construction? Too late I'm sure to see coil, but what kind of shielding? Is shielding loose? Is coil housing flexible?

What happens if put detector on table (no movement), and wave hands near coil? Wave hands near control box?

What kind of motion makes signal? Side-to-side? Up and down? Slow, fast?

Where is he/she testing? Outside? Inside? Near power cords?

Regards,

-SB

[ivconic]

Sir Max Ivco Quzi and all
please help my device works ok TX freq is 14.7.
RX6.5mh TX6mh after sheilding which is exactly what Ivconic
said so the only problem now i have
just very very little movement the device gives signal
i have tried many many ways like nulling from 0mv to 40mv
nothing happned the stand is plastic tube the coil very stiff
and what else
please help..........
Max by the way when i read your responce to Simon about his drwing
10minutes laughing
thanks all

Don't "sir" me! "Sir" on my language means cheese!!!

For example... Cheese Alec Guinness!!! Or...Cheese John Cleese.....Booooahahahahaha!

Ok...let's try to solve your problem, right?
The way you described your problem; looks like you made wrong wirements in coil, at cable ends and possibly at pcb.
Look back at TGS thread, somewhere on some page i posted sketches of one and only proper wirements.
Also pay attention on cable you using to connect coil with device. MUST be proper cable. Also already explained well.
Dont be lazy, search back the pages. You have all already explained very well.
Next; if PROPER cable wired as posted - PROPERLY, than you will be able to adjust phase with GEB trimmer and achieve perfect stabillity on cable movements, choking,twisting etc...etc... same as with search head. Adjust GEB trimmer at the very precise end on ferrite rod rejection (to hear just small cracks on it) and voila; you adjusted all other things also. If you can not achieve this than i am 100% sure you wired up wrongly cable-coil-pcb..
Cheers!

[Max]

Hi 107 - I try to entertain...

If you turn the sensitivity control to decrease the sensitivity, is there a point where the movement does not give a signal?

Can you take some photos of your metal detector and show us what each part looks like?

Cheers,

-SB

Hi,
that's not a solution !

Never decrease that sensitivity.... work always at 110% like russian reactors!

Returning serious...

Decreasing sensitivity is good when e.g. on soil... one have too much troubles at max sens then reduce... it.

But if device is quiet then you move it and sounds.... there's something wrong I think.

I think he has a mistake on either mechanical stability of the assembly... or electrostatic related "ghost" (means that pcb reads some static and thus give a false signal... if box is not at a potential like in the circuit... it will not act as faraday cage... but like the plate of a capacitor!)

Kind regards,
Max

[simonbaker]

Hi,
so I wasn't mistaken there... good things I see !

BTW the CC coil setup is really critical to adj even on tgs... as you noticed too. The fact is, though there's no step-up or harmonic tuning to find, the phasing and nulling are complicated by such fixed design... where coils don't move but you have to use trick to find the good tuning.

I the best s trying untill a correct method will be elaborated... so far I never saw any detailed description around about that... just simple works that describe e.g. the turns ratio for nulling... but phase issue is much more complicate than a simple calculation of turns... all we know that I think.

The fact is... that I think e.g. Tesoro will actually move internal coil a bit... e.g. in spider coils cause that way is easier than having the troubles described... then it's just a little movement so don't compromise the e.g. detection "bubble" they say...

Kind regards,
Max

Yes, easier for Tesoro, they can afford much wire to play with and get it right finally. Then maybe extra single loop adjustment will do it - we have seen that on xray pictures. Certainly not recommended for first-time builders though.

Cheers!

-SB

[Max]

Sir Max Ivco Quzi and all
please help my device works ok TX freq is 14.7.
RX6.5mh TX6mh after sheilding which is exactly what Ivconic
said so the only problem now i have
just very very little movement the device gives signal
i have tried many many ways like nulling from 0mv to 40mv
nothing happned the stand is plastic tube the coil very stiff
and what else
please help..........
Max by the way when i read your responce to Simon about his drwing
10minutes laughing
thanks all

Hi,
nice that you enjoy little humor there... just was lloking at that mess of wires... and had the illumination... that simon is an artist...

Now, coming back to the problem... if pole is stiff... and all very solid contruction... could be some wire also.

When you move the md there are long wires e.g. in the control box that move ?
Sometimes happens that long wires move inside control box and you get a false signal.

Other things I could not imagine... if anything is fixed properly and no vibrations of e.g. metal parts near coils there wouldn't be any false signal due to movements only.

Anyway, if the problem is just when you move it it's probably something mechanically related... the movement make something move and you get the signal.

Otherwise could be electrostatic issue also... but you used a metallic box for circuit right ? Have you connected the metal of control box to gnd (be really careful you don't make short circuits e.g. with sockets connections when try that) ? If you mantain box potential all the same at gnd level your control box will act as a faraday cage for circuit... thus removing any electrostatic related trouble at pcb.

Kind regards,
Max

[simonbaker]

Sir Max Ivco Quzi and all
please help my device works ok TX freq is 14.7.
RX6.5mh TX6mh after sheilding which is exactly what Ivconic
said so the only problem now i have
just very very little movement the device gives signal
i have tried many many ways like nulling from 0mv to 40mv
nothing happned the stand is plastic tube the coil very stiff
and what else
please help..........
Max by the way when i read your responce to Simon about his drwing
10minutes laughing
thanks all

Hi 107 - I try to entertain...

If you turn the sensitivity control to decrease the sensitivity, is there a point where the movement does not give a signal?

Can you take some photos of your metal detector and show us what each part looks like?

Cheers,

-SB

[metaldetector107]

crack

Sir Max Ivco Quzi and all
please help my device works ok TX freq is 14.7.
RX6.5mh TX6mh after sheilding which is exactly what Ivconic
said so the only problem now i have
just very very little movement the device gives signal
i have tried many many ways like nulling from 0mv to 40mv
nothing happned the stand is plastic tube the coil very stiff
and what else
please help..........
Max by the way when i read your responce to Simon about his drwing
10minutes laughing
thanks all