IGSL

[Davor]

...And his latest declaration: "...You can't imagine my expectations... (with vomiting smiley)" is confusing me much refering his real intentions here!!!???
Logical question would be: if you don't have good expectations at all; than what is your reason to be here and deal with IGSL??
Cheers!

Ah, fortunately you got it wrong, smiley is not vomiting but drooling

I'm checking my mailbox every day in expectation of a PCB done by your neighbour from Grocka. Be sure that all my findings will find their way into the rig.

I admit the latest findings in IGSL were a mild let down to me as I obviously pushed some people's wrong buttons, however, I am pretty convinced that channel gain stage mods and replacing trimmers with a stereo pot will be a real advancement.

I'll continue with my spice analysis mostly because - no one else did it.

So my expectations are for real, and very true.

[satdaveuk]

I built two TGSL projects and using Tesoro manufactured coils or home brew using my synthetic graphite mix, get know chatter problems at all unless near electrical pylons etc.
Open fields works like a dream.
Be nice if they went the extra mile, but that applies to all detectors.
The chatter alot are experiencing are due to the gain mods which gave better air testing and poor shielding on coils.
Machine works best in the real world as Ivonics spec sheet using high tolerence components ie checking every component for specification before installing.
Making this project with out of the bag components reduces performance by at least 15% as proved by myself last year.
But on the other hand im pleased people like Simon are still trying to improve the design, because thats what this forum is all about.

Regards

[ivconic]

Lol
On a serious note ive been also following the findings on here and wonder how this project ever worked as well as it does, as its been taken down to a near useless peace of junk from china, going by the findings on this thread.
But after saying all that the truth is in the eating and once the new revised project is up and running, lets see how good it is in the real world, im open minded but still all ears and very interested.
Dont take it personal Ivonics your still the man who made the most popular project on here todate .

Regards

I am not. Don't worry. I don't have ego problem. Besides, this has nothing to do with me. This is original Tesoro design used as "Lego".
Already aproved in real life. Not perfect, but still very descent on trashy soils.
Davor attracted my attention with his pretty sane and interesting start here on Geotech.
I think that guy really can improve IGSL, no joking at all.
But i think he lost his way and he is paying attention on wrong aspects at IGSL.
So by this i only want to direct him in more proper direction.
Also i am looking forward to see his improvements in action.
I do have 2 unfinished IGSL's here in my workshop. Was lazy to complete them.
So after Davor's appearance here i decided to wait his solutions and apply those on my 2 IGSL's. But in his last posts i don't see any improvements.
And his latest declaration: "...You can't imagine my expectations... (with vomiting smiley)" is confusing me much refering his real intentions here!!!???
Logical question would be: if you don't have good expectations at all; than what is your reason to be here and deal with IGSL??
Cheers!

[ivconic]

Again, after so many same repeatings in the past, i will give here a main hint:

TGS/TGSL/IGSL chatters are coming from COIL! Less likely from machine itself.

Not being able, skilled and conversant enough to make proper coil: doesn't give you a credit to suspect intial design!


In case of IGSL ver2, if original Minelab coil is used; than chatter can appear due not well optimized circuitry, bad components, higher tolerances etc...etc...
I gave here general IGSL schematic, few of them. This is diy'ers forum, meaning that all what has been posted should be considered as generally with certain tolerances. Exact "wining" combination in optimizing it is your to discover, but without being so much "smart" and tendable to criticize.

This is an old story. Story on chatters. Last comparators are tendable to selfoscillate. Hysteresis! Pay attention on it, instead being so smart!
Some things are not explained well in Wikipedia, trust me!

[satdaveuk]

Lol
On a serious note ive been also following the findings on here and wonder how this project ever worked as well as it does, as its been taken down to a near useless peace of junk from china, going by the findings on this thread.
But after saying all that the truth is in the eating and once the new revised project is up and running, lets see how good it is in the real world, im open minded but still all ears and very interested.
Dont take it personal Ivonics your still the man who made the most popular project on here todate .

Regards

[ivconic]

Now...this i will say generally, for everybody to read and understand.

Homebrew (or diy) metal detector is worth of positive analyze and positive critics (or negative) only when:

* it is made by yourself,
* experimented with it in workshop,
* tested, checked, "tortured" on many ways,
* taken outdoor on as much as possible different soils,
* tested outdoor in real conditions, with real existing targets in soil,
* aproved by least dozen days spent on real soil and under real life conditions

Otherwise, sitting in front of pc and playing with various spice simulators, reading and misusing various Wikipedia articles (),
playing "smart", using discrete tech terms, calculations, formulas and not having a minute of practical experiences .....
well, all this will make of you only "part-time forum expert" and nothing else.

[ivconic]

You can't imagine my expectations
....
Anyway, I found where this whole oversized capacitor business comes from.
....

You don't need to be such Sherlock; all you have to do is to read TGS thread from a start.
Nobody here reinvented a wheel; IGSL (and TGSL) are completely based on original Tesoro Golden Sabre.
Many things you are discovering now; have been explained much more before.
There is quite reason for each of all the little details to be done as those are done.
Playing with spice simulators doesn't give you complete picture.
I do follow all your posts with huge attention. Many of your points are negotiable, many are not.
Before you continue; please read whole TGS thread very slowly.
Regards!

P.S.
I would give you much more credit if you show here anything of your works. Words mean nothing without material work.

[ivconic]

"..no sanity..." !?

Thanks SB!

All those years here... and i haven't seen NONE of your work so far!?? Zero!
But i see tons of words instead...

[simonbaker]

my analysis shows that time constant with 100k and 22n+22n is far too much. Actually absurdly slow.

I don't know... the time constant is 4.4 milliseconds. But consider -- the LM308 output pulse is typically near ground level for weak signals; I'm not sure the upper limit for large signals, maybe V+ without the diode; but with the diode, it is around .7?

Now it depends where you set the threshold voltage of the LM358 on the inverting input. If you set it around .7 (maximum output of LM308 with diode), then it would take a really long time for the capacitor to charge to that level because of the exponential curve. The lower you set the threshold, the faster the caps charge to the threshold. If you set it at .37 X (.7 - V-), it should take about one TC (4.4 ms) to reach the threshold and trigger the audio tone. So you can easily choose from 4 ms to much longer for your chatter pulse rejection.

If you take away the diode, the same logic applies at a different range, maybe V+ for max signal level.

The thing that strikes me about this is that weak signals would seem to have more delay reaching the threshold than strong signals, because of the exponential charging curve. That seems unfortunate. However, that is offset by the fact that the weakest signal will be near ground level at least, and if we set the threshold low enough, we can get a decent ramp up before the exponential curve hurts too much. But I still don't like it -- if anything, we want to favor weak signals as much as possible. So I'm all for improving this kooky chatter filter if possible.

Anyway, I found where this whole oversized capacitor business comes from. In a King Cobra schematic I found solutions that were copied from there, or a common ancestor of them both, which make much more sense there than in TGSL/IGSL. Even the meaningless 100k resistors at channel amplifiers inputs makes sense with bjt opamps as used in King Cobra. When you see it - you'll find it obvious. Red circles show the two capacitors, and they both are 22nF, same as in TGSL/IGSL. However, a resistor that also determines the time constant here is 2k2, which is ~45 times less than 100k in TGSL/IGSL. Besides, the C48 in King Cobra is active only in a smaller portion of the ramp because a zener diode takes care of that.

Having 100k in TGSL/IGSL calls for 470p to maintain the same time constant. Yeah, I think it may work just fine, and it would improve somewhat signal shape for small signals where level shifter kicks in too fast.

That R58 2k2 resistor looks more like it may be part of a Schmitt trigger circuit to me, which also cuts chatter in a different way. The R64 470K resistor with the 22nF cap I think is the time delay chatter filter similar to our TGSL, and it has even longer time constant. But the actual time delay rejection will depend on the relative levels of the threshold and starting voltage of the ramp.

That is a nice find by the way; it is interesting to discover the geneology of metal detectors; your point is well taken that often circuits are copied with no sanity check behind them.

-SB

[Davor]

You can't imagine my expectations

You are right about most of the things, however, my analysis shows that time constant with 100k and 22n+22n is far too much. Actually absurdly slow.

Anyway, I found where this whole oversized capacitor business comes from. In a King Cobra schematic I found solutions that were copied from there, or a common ancestor of them both, which make much more sense there than in TGSL/IGSL. Even the meaningless 100k resistors at channel amplifiers inputs makes sense with bjt opamps as used in King Cobra. When you see it - you'll find it obvious. Red circles show the two capacitors, and they both are 22nF, same as in TGSL/IGSL. However, a resistor that also determines the time constant here is 2k2, which is ~45 times less than 100k in TGSL/IGSL. Besides, the C48 in King Cobra is active only in a smaller portion of the ramp because a zener diode takes care of that.

Having 100k in TGSL/IGSL calls for 470p to maintain the same time constant. Yeah, I think it may work just fine, and it would improve somewhat signal shape for small signals where level shifter kicks in too fast.

[simonbaker]

That is the exact pot I was thinking about, however, in a 0Ω position a capacitor parallel to a 100kΩ resistor does exactly nothing. Delay is associated primarily with a capacitor next to a LM358 non-inverting input. The remaining capacitor next to a 1MΩ does exactly nothing. The very 1MΩ does exactly nothing as well. You may associate delay function with anti-chatter, but only in case of a binary response, but you may also omit it and anti-chatter with disc sens pot, and gain some real control over it.

I think it is a waste not to use this pot for volume control. For the existing rigs I'd suggest extracting the trimmers and wiring a linear stereo potentiometer of 100kΩ instead. At position 0Ω you'll have a pure binary response anyway, so everyone is happy.

I don't see it that way.

Referring to the TGSL circuit, C25 (.022uF) and C24 (.022uF) directly affect the rise time of the pulse coming from the comparator outputs. I think these LM393 comparators are "open collector" outputs, so when both triggered high, their output voltages must be pulled up by the signal coming through that 100K resistor -- but because of the two capacitors, it forms an integrator which slowly ramps up the voltage as the caps charge.

In quiescent mode, the non-inverting input of the LM358 is held at approx V- (negative rail) +.7 because the LM393s pull the voltage down. When a target signal arrives, both LM393s go "high" open-collector, which starts the ramping up of the voltage on the caps from the current through the 100K resistor.

Because the audio will not sound until the LM358 non-inverting input exceeds the inverting input, there is a time delay while the voltage on the non-inverting input is ramping up. Short pulses will never make through at all.

In fact, this is one of the weaknesses of the chatter filter -- noise riding on a weak target signal can break it up into small pieces that never get past the time delay and you hear nothing -- which may have been the intent of the designers, to make a clean signal, or none at all, to market to the casual hunter.

We (obsessed fans) may actually prefer to hear the complete noisy signal, so with experience, we can learn to interpret different degrees of chatter and derive more information about deep targets at the noise threshold.

So I'm all for experimenting with designs that eliminate the chatter filter. I did try it briefly, but couldn't get satisfying results. However, it is on my list for further experimentation.

I agree the extra RC network of C24 and R54 seems extraneous -- I think C24 and C25 could be consolidated into one capacitor and R54 eliminated.

It would be useful to build a prototype where you can switch in and out the chatter filter and see what you like best. In any case, we'll be interested in your impressions when you build the circuit and use it in the field.

Regards,

-SB

[ivconic]

Throwing components away from IGSL just like that; will lead us to ultra-turbo-super-stealth metal detector!


Something like this:




(I need jokes right now, because Nole is loosing against Rafa on most stupid way it can be!!! )

[golfnut]

You would sample the Tx signal on AtoD 1
and Rx signal on AtoD 2

Then you would have target phase information.



S

[Davor]

Hi SB,

you don't need the near DC frequency range. Direct sampling at the pre-amp output and direct digital demodulation. That's it.
Cheers,
Aziz

Yes, but unless you supply the oscillator signal as well you lose discrimination (no phase information). It can work well with a stereo input. The high side of this approach is completely avoiding 1/f noise. If processing is done properly, say using a decimating digital filter to gain S/N, you'd be able to obtain very sharp responses for very low signals.

Baseband is troublesome because all soundcards are high pass cut at~20Hz, and you inevitably lose DC information.

It is all in signal autocorrelation and crosscorrelation properties.

[Davor]

Again, I think the original use of the pot was a chatter filter too, since the voltage level determines how short a pulse from the comparators is rejected -- assuming you have the capacitors in there. Unless you're talking about a different pot.

-SB

That is the exact pot I was thinking about, however, in a 0Ω position a capacitor parallel to a 100kΩ resistor does exactly nothing. Delay is associated primarily with a capacitor next to a LM358 non-inverting input. The remaining capacitor next to a 1MΩ does exactly nothing. The very 1MΩ does exactly nothing as well. You may associate delay function with anti-chatter, but only in case of a binary response, but you may also omit it and anti-chatter with disc sens pot, and gain some real control over it.

I think it is a waste not to use this pot for volume control. For the existing rigs I'd suggest extracting the trimmers and wiring a linear stereo potentiometer of 100kΩ instead. At position 0Ω you'll have a pure binary response anyway, so everyone is happy.