TESORO GOLDEN SABRE

[Dennis the Mennis]

Hi DtM!

Love those great graphs you get with your USB scope! I have a configuration question: can you configure the graphs to have 10 dots per square instead of 12 (for easy estimate of values)? And can you configure it to show the time scale on the horizontal axis? It would be great also if description of plot can be put on graph so it can be understood without surrounding text. Just some ideas. I hope to get a similar tool some day.

Hi Simon ,

I will try the configuration of the horizontal axis when my laptop is back from the ICT department again ...
The time scale was ( in yellow , see my other pictures ) printed under the grid untill I had a software update for my USB scope due to problems with external triggering , got this problem back ...
These were the first pictures I've posted after the update and I didn't notice the missing scale , good observation.

Yes , there is a possibility to add text to the bottom of the graphs but it isn't displayed , or at least not when I save it as a GIF file , didn't try the other 'save as' options yet .

Ok, now lets look at your data.

1. I think your sync pulse looks pretty normal and you should be able to discriminate targets. When I compare these new graphs to habitbraker, they quite similar in shift and location, don't you think so?

So my question is: what targets do you use for your discrimination test, and what results do you get again? (My best test target is US Nickel. TGS does not discriminate silver or even copper really).

It looks similar to the one of Habitbraker so I must be close ...
I didn't spend that much time at the DISC mode in the past because I was a little focused on the ' All Metal ' mode : Why do I get good detection and some time later absolute worse detection . ( In my case really unpredictable noise , some more postings about that later for anyones benefit)
And I had a defect in that part of the circuit so it wouldn't work ever ...
I will do some more experimenting with different materials.

2. I have seen a similar "spike" on my RX null signal at times. I did not consider it might be due to capacitor C9, but that could make sense due to saturation of the LF353 op amp. (The good news, as you said, is that the spike does not seem to hurt the detection. Anything very periodic at the TX frequency should not hurt.)

However, the "spikes" on the sync pulse can be definitely related to capacitor C9, because when the LF353 op amp saturates (DISC op amp, pins 1,2,3), it causes a "ringing" output at pin 1 which then can make a messy sync pulse. (Surprisingly, the messy sync pulse does not seem to cause much trouble).

So lowering value of C9 until there is no "ringing" in output at pin 1 should make the sync pulse look nicer, I believe. I think a "nicer" sync pulse may give you a little better DISC range. I looked at your video of the "wiper signal" and there is a small ringing "blip" that goes near zero at one end of the DISC pot - it may be making the spike in your DISC sync pulse signal at the very max pot setting -- but probably goes away most of the time, so no big worry.

3. I would like to see some more graphs of your LF353 output at pin 1 (compared to TX signal on other trace) for different values of C9. What value of C9 makes the ringing completely disappear? Maybe go down to 50pf, 40pf, 30pf, etc.

Actually, it looked pretty good at 68pf from your graphs, I would choose that value to keep maybe. It would be interesting to see your sync pulse at max and min DISC pot setting with C9 = 68pf.

I removed C9 and connected an IC socket as a kind of brigde were I can easily place or remove caps .
I will post some more graphs with different values for C9

4. I think your blob test made the point I wanted to test (even with your genuine Rolex watch involved). I need to think about it some more and maybe do my own tests carefully -- when the blob is really close to the coil maybe that causes some other effect entirely. So requires careful interpretation. But I think it shows that the significant "null signal phase" is not really what we see on our scopes, and so we shouldn't go crazy trying to set a particular value.

Thanks for all the work making the graphs and videos!

Cheers,

-SB

Please let me know if want to do some other 'blob' testing.

And you're welcome , yes it takes a lot time making all the graphs and videos ...
Just like you answering all the posts , I should also thank you Simon . ( and all the others with a helping hand )

I hope that the graphs and videos can contribute to anyone having problems or trying to troubleshoot and don't know what to expect , just like me at some points.

I hope also to bundle someday all info but that is a lot of work



kind regards ,

Dennis the Mennis

[Qiaozhi]

did silver sabre had a pinpoint swith? in original silver sabre micro max has. i was wondering..maybe there is a way to put a pinpoint switch i tgsl? it's wery good to find an object wery fast..i have a original ss micromax pcb without wires( they were wery mest up) and i'm still looking for good original schematic and coil parameters. i heart that bandido is wery similar but has no pinpoint

anyway..

best regards

I have a Tesoro Bandido.
It has a switch to select Discrim, All Metal or Retune. The All Metal mode is non-motion and used for pinpointing, and the Retune setting restores the threshold in All Metal mode.

[lunamay]

did silver sabre had a pinpoint swith? in original silver sabre micro max has. i was wondering..maybe there is a way to put a pinpoint switch i tgsl? it's wery good to find an object wery fast..i have a original ss micromax pcb without wires( they were wery mest up) and i'm still looking for good original schematic and coil parameters. i heart that bandido is wery similar but has no pinpoint

anyway..

best regards

[Dennis the Mennis]

http://imageshack.us/photo/my-images/824/tesoro2.png/

Hello.
This psb tgsl it is right?
It works with these prices of materials that it has?
version tgsl 3 it is better?
I request any it knows it helps me.
I thank

Hi Fmnotes ,

The picture you've posted is the Tesoro Golden Sabre 'original ' not the TGSLight
This version does work however.

If you want to build the TGSL :

http://www.geotech1.com/forums/showthread.php?t=15710

or you can buy one of Silverdogs PCB's : The TGSL EDU

http://www.geotech1.com/forums/showthread.php?t=17727

As far as I understand this is almost the same version with some small changes.

I hope I understood your message well

kind regards ,

Dennis the Mennis

[satdaveuk]

Thanks

Thanks Simon
I didnt have that information, very greatful.

Warm regards

Dave

[simonbaker]

Hi
I have ordered all the components, can someone give me the pcb size please? save messing about.
Many thanks
Dave

Have you gotten Ivconic's info from this thread? http://www.geotech1.com/forums/showthread.php?t=15710

You can print out the PCB layout and measure, etc.

-SB

[satdaveuk]

pcb size

Hi
I have ordered all the components, can someone give me the pcb size please? save messing about.
Many thanks
Dave

[satdaveuk]

O.K.
I already made two G.Sabre's and two S.Sabre's so far.Wanted to know how those are working.
Also made few coils for those. Already posted some of experiences with coils.
I want to "establish" one and only good schematic and from now one to make only that.
Much experiments so far. After dozen different projects i choose Tesoro Golden Sabre to be
my favorite detector to build. Why? Simply, all real functions that prospector need on
field G.Sabre already comprehend. Discrimination is so good, it can fully discriminate
foil, iron.....what man can wish more?
Depth is great! Maybe i had luck to calculate and make quite proper coil for it, but my
experiences are so good...much better than some posted previously on this forum. Already
made few identical coils, just to check if i can make those as standard....It was not lucky
guess or coincidence, yes i got it for good!
Also noticed that G.Sabre is for real much better than S.Sabre. Greater depth even 30%.
Solution with 4024 turns to be more effective than 7660...? Why, i dont know. It seems that
all stages in device are better synchronized if dependable on main oscillator frequency.
So i rearranged schematic and cut out "notch" part, so pcb would be as smallest as possiblle.
I also rechecked schematic and previous pcb...? Except those 2 resistors(4k7) in chopper and
one missvalued capacitor(15pF) i dont see any other mistake. 4148 diode at LM393 output
("Sense" stage) is not reversed??? I think somebody claimed here that it is reversed.
I never tried to reverse it and see....Made my detectors according to posted pcb and those are
working good.
So here is Golden Sabre "light" schematic. I hope it is accurate. Checked few times....
Next step is to draw small, one-sided pcb for it....Why one-sided? I prefer one-sided pcbs.
No conditions for metalized holes in pcb's so i mostly do one-sided..
Also most of my pcb's i draw manually with thin marker.It is much easier to draw single-sided.
But there is another problem! Huge problem! Mountain problem!
I do have DXP,Eagle,ExpressPCB,SprintLayout....but still do not have enough paraxis and knowledge
to draw good pcb with those..Already learned to input schematics.But when comes to pcb i
really do not have nervs to torture myself with those software. Already tried hundred times and
at the end, my final pcb's sucks! DXP is very strange! Making its own nets and connections from
time to time. Eagle is even more stupid. ExpressPCB is simple and stupid, almost like drawing
manually....To many parameters to input and to be considered. Lost eneormous time even for
simple designs. At the end final draw is not what i wanted - not perfect.Many lacks.
That's way i decided to give up of those and let more conversant people to do the job.
I fully recomend this project to everybody interested.
No need for "high phylosophy", no need for "high tech" projects.At the end all those are very
close when comes to performances. This schematic associate all merits which one good detector
must have...

Thanks so much
So the gold light is the final, the accurate ont to work on, and the overview and pcb layout are also correct?
If so im going to build one, thats if i can get the parts of course.
I shall take note of your coil designs
Many thanks again
Dave

[simonbaker]

Hi Simon ,

These are my settings for the DISC sync pulse vs Tx :

Yes , I was confused while posting the DISC vs Rx , I've had in mind that the Rx was sampled , but as I understand that is not for sure (?) .

DISC potmeter 0% :

[ATTACH]15260[/ATTACH]

and

[ATTACH]15261[/ATTACH]

DISC potmeter at 50% :

[ATTACH]15262[/ATTACH]

and

[ATTACH]15263[/ATTACH]

DISC potmeter at 100% :

[ATTACH]15264[/ATTACH]

and

[ATTACH]15266[/ATTACH]

Also here a little spike as described in previous post : most likely cause C9 ( in my case anyway) .

It seems that the shift of Habitbrakers pcb is a little more shifted to the right and if it works for him I'll have to 'tune' my Rx a little so I can get some discrimination .




video DISC sync pulse vs Tx : http://youtu.be/opZh96JHWFc?hd=1




video DISC potmeterwiper vs Tx : http://youtu.be/0-JAuOxsQdg?hd=1


kind regards ,

Dennis the Mennis


ps

Simon ,

You could have been right during the ' Blob test ' ... most likely I've forgot my 24 carat Golden Rolex .

Another Blobtest ?

Hi DtM!

Love those great graphs you get with your USB scope! I have a configuration question: can you configure the graphs to have 10 dots per square instead of 12 (for easy estimate of values)? And can you configure it to show the time scale on the horizontal axis? It would be great also if description of plot can be put on graph so it can be understood without surrounding text. Just some ideas. I hope to get a similar tool some day.

Ok, now lets look at your data.

1. I think your sync pulse looks pretty normal and you should be able to discriminate targets. When I compare these new graphs to habitbraker, they quite similar in shift and location, don't you think so?

So my question is: what targets do you use for your discrimination test, and what results do you get again? (My best test target is US Nickel. TGS does not discriminate silver or even copper really).

2. I have seen a similar "spike" on my RX null signal at times. I did not consider it might be due to capacitor C9, but that could make sense due to saturation of the LF353 op amp. (The good news, as you said, is that the spike does not seem to hurt the detection. Anything very periodic at the TX frequency should not hurt.)

However, the "spikes" on the sync pulse can be definitely related to capacitor C9, because when the LF353 op amp saturates (DISC op amp, pins 1,2,3), it causes a "ringing" output at pin 1 which then can make a messy sync pulse. (Surprisingly, the messy sync pulse does not seem to cause much trouble).

So lowering value of C9 until there is no "ringing" in output at pin 1 should make the sync pulse look nicer, I believe. I think a "nicer" sync pulse may give you a little better DISC range. I looked at your video of the "wiper signal" and there is a small ringing "blip" that goes near zero at one end of the DISC pot - it may be making the spike in your DISC sync pulse signal at the very max pot setting -- but probably goes away most of the time, so no big worry.

3. I would like to see some more graphs of your LF353 output at pin 1 (compared to TX signal on other trace) for different values of C9. What value of C9 makes the ringing completely disappear? Maybe go down to 50pf, 40pf, 30pf, etc.

Actually, it looked pretty good at 68pf from your graphs, I would choose that value to keep maybe. It would be interesting to see your sync pulse at max and min DISC pot setting with C9 = 68pf.

4. I think your blob test made the point I wanted to test (even with your genuine Rolex watch involved). I need to think about it some more and maybe do my own tests carefully -- when the blob is really close to the coil maybe that causes some other effect entirely. So requires careful interpretation. But I think it shows that the significant "null signal phase" is not really what we see on our scopes, and so we shouldn't go crazy trying to set a particular value.

Thanks for all the work making the graphs and videos!

Cheers,

-SB

[fmnotes]

http://imageshack.us/photo/my-images/824/tesoro2.png/

Hello.
This psb tgsl it is right?
It works with these prices of materials that it has?
version tgsl 3 it is better?
I request any it knows it helps me.
I thank

[Dennis the Mennis]

Thanks!

Your sync pulse seems to shift about 80 deg or so, which is fine. I notice that at both ends of the pot range it breaks up, which is normal for TGSL design, same as I observed with dfbowers model that he sent me (which works very well).

Hopefully Dennis the Mennis can post equivalent video/photos so we can compare his sync pulse to his Tx signal and confirm if his DISC circuit is similar.

Regards,

-SB

Hi Simon ,

These are my settings for the DISC sync pulse vs Tx :

Yes , I was confused while posting the DISC vs Rx , I've had in mind that the Rx was sampled , but as I understand that is not for sure (?) .

DISC potmeter 0% :



and



DISC potmeter at 50% :



and



DISC potmeter at 100% :



and



Also here a little spike as described in previous post : most likely cause C9 ( in my case anyway) .

It seems that the shift of Habitbrakers pcb is a little more shifted to the right and if it works for him I'll have to 'tune' my Rx a little so I can get some discrimination .

It would be helpful to do the same video of the DISC sync pulse compared to the Tx signal (because we all have different Rx null signals, so difficult to compare).

Cheers,

-SB

video DISC sync pulse vs Tx : http://youtu.be/opZh96JHWFc?hd=1

Also, I'd like to see a video of the signal at the "wiper" (center) tap of the DISC pot as you rotate the pot, compared to the Tx signal.

Also, I'd like to see the the signal at the output of LF353 pin 1 (connected to one side of the DISC pot) compared to the Tx signal -- just a photo, since it should not change as you rotate the DISC pot.

Cheers,

-SB

video DISC potmeterwiper vs Tx : http://youtu.be/0-JAuOxsQdg?hd=1


kind regards ,

Dennis the Mennis


ps

Simon ,

You could have been right during the ' Blob test ' ... most likely I've forgot my 24 carat Golden Rolex .

Another Blobtest ?

[Dennis the Mennis]

http://www.geotech1.com/forums/showpost.php?p=124941&postcount=4429

Hallo Danny,

Think you have not a good receive signal, tray to adjust it with the 'receive cap' .

Make a 'Receive Cap' test circuit and see how the receive signal changes...

Hi Ap,

Perhaps a little late ,

I've tried to adjust C9 , the cap. before input LF353 U101b pin 2, to get rid of those spikes on Rx signal , see next image ( Rx output LF353 pin 7 vs Tx )



and there are some improvements on the output of the LF353 pin 1 :
Next picture I've installed 180 pF for C9 : output LF353 pin7 vs output LF353 pin1

It seems that I accidentally posted an animated GIF , looks good , I'll keep it in mind ...




100 pF for C9 : output LF353 pin7 vs output LF353 pin1




and 68 pF for C9 : output LF353 pin7 vs output LF353 pin1



Allthough there could be an improvement in signal ( the spikes become smaller and / or shifts a little ) for my pcb , I can't get a better or worse detection.


The DISC sync pulse stays the same no matter the value of the cap ,BUT only at the beginning and/ or the end of the DISC shift there could be a spike that is influenced by these caps !

The spikes like Habitbraker posted in his last post :

http://www.geotech1.com/forums/showpost.php?p=127648&postcount=4575

It seems that this part of the circuit is pretty forgiving ...

Originally Posted by simonbaker
Quite excellent DtM! Nice way to show your new scope.

I'm interested also in the LF353 output that drives the sync pulse, what does it look like?

I get those little spikes on the RX signal also. I suspect it is caused by the negative power rail circuitry, but not sure

.

It looks like that the spikes on the Rx signal , only visible with very low residual voltage, are 'generated' internally in the LF 353 , at least they influence each other pretty good .

In the posted images I've set the rulers on the spikes : the spikes appear on the exact same places .
Also when nulling the coils , both spikes stay on the same place despite the shift.

This was just a small observation , perhaps it was told a dozen times in these threads.

kind regards ,

Dennis the Mennis

[simonbaker]

You got it. In your asked sequence. Both traces 5V/div

Thanks!

Your sync pulse seems to shift about 80 deg or so, which is fine. I notice that at both ends of the pot range it breaks up, which is normal for TGSL design, same as I observed with dfbowers model that he sent me (which works very well).

Hopefully Dennis the Mennis can post equivalent video/photos so we can compare his sync pulse to his Tx signal and confirm if his DISC circuit is similar.

Regards,

-SB

[habitbraker]

It would be useful to see three photos of the Tx signal (J1-1) compared to the DISC sync pulse (LM393 U102 pin 1) :

1. With DISC pot at minimum setting.
2. With DISC pot at middle setting.
3. With DISC pot at maximum setting.

Thanks if possible!

-SB

You got it. In your asked sequence. Both traces 5V/div

[simonbaker]

Laptop back in the air , can connect USB scope again , time for a small test:

I did a 'Blobtest' with the ferrite rods (10cm long) , it didn't matter if I used one or two ferrite cores .

The signals shown are Tx ( green) vs Rx output 7 LF353.

I started with the unshielded coils :

Kept the ferrite rod +/- 30 cm above the coils and lowered them right in between the overlap of the coils.

http://youtu.be/ZAQPJhNk9f8?hd=1 ( causing a little shift to the right )

The next test is with shielded coils and same as described as above.

http://youtu.be/8iG90U8sCdg?hd=1 ( causing a little shift to the left )


And one other question for anyone willing to answer:

I'm still struggling with my detector in DISC mode , I've found some defects and repaired them.

Is this the correct DISC shift ?

Shown : Rx output LF353 vs DISC sample puls / output 1 LM393

http://youtu.be/Q_uSpD_dUrY?hd=1

kind regards,

Danny

Dennis the Mennis

I just took a closer look at your "blob" test with the shielded coil.

I looks like your null signal is around 50 deg from your TX signal with no target.

When the ferrite blob is near, the null phase changes to somewhere between 30 - 35 deg, which is similar to the null phase of the unshielded coil.

That is the point I was trying to make -- that the shield is actually somehow moving your basic null signal phase to 50 deg, but really, as far as ferrite is concerned, the null is still back around 30 deg, because the ferrite is amplifying the "magnetic null" signal (which is the one that the ground mostly affects). So we shouldn't really worry too much about the null phase we normally see.

Now there is something I can't understand about your "unshielded blob test". As you bring the blob close, the null phase hardly changes, which is what I expect, because the unshielded coils have more of a pure "magnetically coupled" null signal. But suddenly, the null shifts to the right in a big jump. I'm not sure why that happens, unless it is because the signal overloads the circuit and the symmetry is corrupted, shifting the zero crossing. Or could your hand have also had some metal, like a ring or watch?

In any case, it is interesting to note that ferrite does shift the null signal of our coils, especially shielded ones, and to me that means that there is a magnetic and non-magnetic component to the coil coupling. And that means that we cannot assume that ground will amplitude-modulate the null signal that we see; rather, it will modulate the magnetic component of the null signal.

This is just theorizing on my part; more experiments might indicate a different theory.

Regards,

-SB