TESORO GOLDEN SABRE

[akademiker]

it is a TGSL build according to http://www.geotech1.com/forums/showthread.php?t=15710

I tried to null the coils while measuring Pin 7 of LF353 and I couldnt get lower than an amplitude of 1,6V (Yes its not millivolts, but i guess its ok cuz it gets amplified, right?)

The syncs both show a nice sawblade and change their position a little. But when i turn the disc pot in the middle position the signal looks very strange and then is almost zero.

[simonbaker]

Shields have both a gap and dont touch each other except at the position connected to the cable.
What exactly do you mean with "Sync pulse" ? Which pins should i measure?
Also i dont know in which position my disc and gb pots are at minimum as i use external pots, what signal should be at the middle part of the pot when at minimum?

Thanks for pix.

If you can make them a little smaller for viewing in future (720 horizontal prob good).

Which schematic did you build? Your first photo of TX signal looks very low voltage and does not have the "bump" like most TGSL. It is more like TGS schematic that has a JFet in the oscillator circuit.

Your second photo - is that LF353 pin 7? If so, it means your coils are not nulled at all. Try nulling them until signal is minimum, then push them together just a hair more until voltage rises a little from minimum (maybe 30%).

If you look at typical TGSL schematic, DISC sync pulse is output of U102a pin 1 (LM393 chip), and GB sync pulse is output U102b pin 7 (same LM393 chip).

Regards,

-SB

[akademiker]

Shields have both a gap and dont touch each other except at the position connected to the cable.
What exactly do you mean with "Sync pulse" ? Which pins should i measure?
Also i dont know in which position my disc and gb pots are at minimum as i use external pots, what signal should be at the middle part of the pot when at minimum?

[simonbaker]

Hi guys, I have some problems with my TGSL circuit, most likely the coil, but I dont know for sure.

My coils:
When nulled they are 26cm in diameter.
TX=5,77mH with 20,5nF for C2 at about 14,45kHz
RX=6,22mH with 15,1nF for RXC should be around 16,42kHz
Shielded with alu tape.

It detects a 4,5cm 1oz silver coin at around 10cm and 1e at around 8cm in AllMetal-Mode, very poor.

I replayed C9 with a resistor 100k, because the signal at U101b pin 1 was a very crippled sinus. (Tried with different ceramic caps, different qualities and different values like 68pf and 47pf)

Nulling is very strange. I disconnect RX and measure both ends with my scope while running TX. The sinus gets smaller and larger but I cant get past 7mV.

I also switched RX/TX starts and ends in many different ways and only one combination gave detection.

Unfortunatly only one channel and the external trigger are working at the moment, so i cant measure phase shift.

Could the porblem be in the circuit?

With external trigger on TX signal, I think you should be able to estimate phase shift of other signals, no?

Can you take pictures of scope? Put external trigger on TX signal (J1-1), then show separate pics of TX signal, RX signal at LF353 pin 7, Sync pulses at JFet gates. Make sure you show the left edge of the signal, and also tell us where zero volts is on your scope screen. Don't readjust horizontal position between pictures so we can estimate the phase.

Show the sync pulses with the DISC and GB pots at minimum. Use DISC mode, not ALL METAL. Then rotate the pots and make sure both sync pulses move back and forth a little. DISC sync pulse should move around 80 to 90 deg, GB pulse around 30 to 45 deg.

Eventually, you will want to put your scope on very slow sweep (.05 Seconds / div) and look at your target pulse at output of LM358 and LM308 chips.

Post photos of your PCB and coils. Shields on coils can keep null large, not necessarily a problem. Double-check your shields have the "gap" in them and do not touch each other to make a continuous loop anywhere.

You can also look at troubleshooting thread for ideas and asking more questions.

http://www.geotech1.com/forums/showthread.php?t=16122

Regards,

-SB

[akademiker]

Hi guys, I have some problems with my TGSL circuit, most likely the coil, but I dont know for sure.

My coils:
When nulled they are 26cm in diameter.
TX=5,77mH with 20,5nF for C2 at about 14,45kHz
RX=6,22mH with 15,1nF for RXC should be around 16,42kHz
Shielded with alu tape.

It detects a 4,5cm 1oz silver coin at around 10cm and 1e at around 8cm in AllMetal-Mode, very poor.

I replayed C9 with a resistor 100k, because the signal at U101b pin 1 was a very crippled sinus. (Tried with different ceramic caps, different qualities and different values like 68pf and 47pf)

Nulling is very strange. I disconnect RX and measure both ends with my scope while running TX. The sinus gets smaller and larger but I cant get past 7mV.

I also switched RX/TX starts and ends in many different ways and only one combination gave detection.

Unfortunatly only one channel and the external trigger are working at the moment, so i cant measure phase shift.

Could the porblem be in the circuit?

[simonbaker]

C9 modification

TGSL DISC control circuit has well-known behavior with jagged waveform at LF353 (U101b) output pin 1, sometimes causing break-up of sync pulse at low DISC pot settings.

This may not be a problem and is OK. But it seems bad.

A simple way to improve it may be to reduce the value of capacitor C9 (at inverting input pin 2 of LF353 U101b) from 100p to around 50p to 60p. You want a value that gives maximum output voltage without the jagged waveform. It depends on your TX oscillator voltage.

This should be simpler than the "SB mod" that I tried in TGSL Experiments thread, and no need to change PCB.

-SB

[simonbaker]

curious

R54 (1M) and C24 (.022) are a curious part of the TGSL circuit.

Anyone have ideas on exactly why it is there? Has anyone tried removing it and merging C24 into C25 by using a .047u cap for C25?

Even though it looks like a pull-up, the real pull-up is done by U105 output through R40 (100k) I think.

An early TGSL circuit had 4.7K resistors separating U106 outputs from C24 -- then the circuit might have had a purpose in creating a slower shutoff of pulse.

I think Molzar checked original TGS and no 4.7K resistors though.

On another note: those extra resistors can be useful during tuning -- because they slightly isolate the LM393 outputs, and then you can check each LM393 output separately and see if they trigger at the same input threshold -- if not, you can consider replacing the LM393 for one with equal input thresholds.

-SB

[simonbaker]

Analysis with 150 ohm resistor shows that when ear pieces are connected in series, they will sound about 3 dB more loud instead of parallel. That means in each ear piece flows about twice more current.

Ok, thanks for advice.

I'm actually not sure if there is a shield on my phones, but question was the same.

They are loud suckers without the resistor. I always put them above the ear to protect hearing, but I'll try resistor.

Interesting about "anti-phase", I hadn't thought about that -- probably wouldn't bother me -- but maybe nicer if target sounds like it is in front of you! (next idea -- using stereo effect to indicate location of target in ground -- some day)

Regards,

-SB

[mikebg]

However in this circuit, the serial connection can sound even more loud because of better impedance matching of 70 ohm load with 150 ohm signal source. Signal source operates as if it has impedance 150 ohm as shown.

Analysis with 150 ohm resistor shows that when ear pieces are connected in series, they will sound about 3 dB more loud instead of parallel. That means in each ear piece flows about twice more current.

[Qiaozhi]

What does everyone think about wiring the headphone jack for TGSL for stereo headphones?

Do you wire phones in series and ignore shield connection, or wire in parallel and ground shield connection, or other?

I'm using a cheap set with about 35 ohms per ear piece.

I'm tempted to wire in series, ignoring shield, to reduce power drain -- they're plenty loud that way.

-SB

Most people use headphones with a built-in volume control for the Tesoros. You can get these cheap from a metal detector shop.

[mikebg]

What does everyone think about wiring the headphone jack for TGSL for stereo headphones?

Do you wire phones in series and ignore shield connection, or wire in parallel and ground shield connection, or other?

I'm using a cheap set with about 35 ohms per ear piece.

I'm tempted to wire in series, ignoring shield, to reduce power drain -- they're plenty loud that way.

-SB

SB, Energy (and ear :-) saving is done by 150 ohm resistor as shown below. Save your ears; don't use phones without this resistor!
The 35 ohm resistance per ear piece is inherent not only for cheaper, but for all stereo headphones. If your phones have shielded cable, they are not cheap. The cheap phones use simple pair wire for each piece, even not enough twisted pair.
The serial connection of earpices produces antiphase sound pressure for ears. In this case your brain can not imagine where is the sound source. If this not create an uncomfortable feeling for you, use it.
If where no resistor 150 ohm, sound power (pressure) at seral connection decreases twice, but because of the logarithmic human hearing, the loudness (volume) decreases only by 3 dB.
However in this circuit, the serial connection can sound even more loud because of better impedance matching of 70 ohm load with 150 ohm signal source. Signal source operates as if it has impedance 150 ohm as shown.

[simonbaker]

wiring headphones

What does everyone think about wiring the headphone jack for TGSL for stereo headphones?

Do you wire phones in series and ignore shield connection, or wire in parallel and ground shield connection, or other?

I'm using a cheap set with about 35 ohms per ear piece.

I'm tempted to wire in series, ignoring shield, to reduce power drain -- they're plenty loud that way.

-SB

[Janotaking]

Im building Surfmaster too, what detector do you use? What is your favourite?

[dfbowers]

Yeah, the scope is nice
Im working on SandBanks right now. I need some help. Can you help? It would be great. See the Sandbanks PI post I've posted my questions there, some are already answered but theres still some there. Like the switcher, I don't find the switcher I only found 100k up switchers.

Stay cool men. Thanks a lot.

Have not built Sandbanks yet. HH and HH SMT, Stuart PI and Stuart Microcontroller.

[Janotaking]

It's often said that most things we learn in life are learned while we are trying to do something else.. So did you want to learn to use a scope as well?

Yeah, the scope is nice
Im working on SandBanks right now. I need some help. Can you help? It would be great. See the Sandbanks PI post I've posted my questions there, some are already answered but theres still some there. Like the switcher, I don't find the switcher I only found 100k up switchers.

Stay cool men. Thanks a lot.