GEB on IGSL/IDX

[simonbaker]

Yes, it is noise limited
It seem as if the GEB channel is completely misunderstood. In reality a GEB channel is an all metal channel, so it is maximum metal zero ground, and that's why tone signal is extracted from it. Furthermore, sensitivity of GEB channel is maximum at exactly 90° from ground phase, and funny that you mentioned it, small gold is found right about that angle. So GEB is about as optimal as it gets.

You took the words right out of my mouth.

On the other hand, Disc. channel isn't. Let me explain.

A switching detector is in effect an analogue multiplier. It multiplies a square wave with an incoming signal. Such multiplication produces some high frequency components, but more importantly a DC signal that is related to the incoming signal level, and the phase. With only a slight approximation, we may say that the output of such multiplication after filtering is cosine related to the RMS value of the incoming signal. Cosine means full value for 0°, zero for 90° phase difference, and vanning values in between.

Let's assume for simplicity that we have a gold response at 80° from Disc sample phase. cos(80°)=0.174 and when converted to dB against the value in a GEB channel (0.985) it is -15dB. It means that you'll get a positive response on small gold only for 15dB++ in GEB channel.

Shortly, Disc. channel is in considerably more trouble than GEB.

Agreed, in theory.

My suggestion ... go for a dual discrimination just as IGSL does, and overlap the discrimination channels. You'll hit two flies with a single blow:
- you effectively get a very useful third "tone" indication for Al foil, modern nickel rich coins and small gold
- by overlapping, the disc. channel vector is skewed a bit further, so that 80° from the above example becomes 60-70°, and GEB to Disc. ratio becomes -4.7 to -8.7dB, hence your depth increases for at least 10%

I already overblown the gain in Disc. channels in my IGSL to compensate for the Disc. to GEB ratio, and as a consequence my IGSL is neither losing sensitivity at low levels, nor exact discrimination indication in full span from low to high levels in the overlapped region. I just love it.

I'm not sure that is correct that you are not losing sensitivity. (But I'm not as familiar with the IGSL. My understanding is that the IGSL is two TGSLs, but one TGSL beeps when the metal is below its DISC setting and the other TGSL beeps when the metal is above its DISC setting. Both TGSLs have a GEB channel to suppress ground, although in Davor's mod they use the same GEB channel.)

Since both DISC channels are not "optimal" for gold (only the GEB channel is optimal for gold by coincidence, as Davor also illustrated), you are not really as sensitive to gold as you could be if you set the IGSL as Davor indicated (overlapping on gold). Now Davor says he compensated by increasing the gain of his DISC channels. However, this also increases the noise, so he has not gained S/N in the DISC channels, only adjusted the threshold similar to the GEB channel.

The GEB channel should still have a superior S/N for gold, because the sync pulse is optimal. So a single GEB channel would be better for hunting gold in remote areas where there is no trash to worry about.

ALL METAL mode should be good also, but the IGSL may have the same triangular sync pulse as the TGSL for ALL METAL mode, I'm not sure. If not, then ALL METAL mode would make a good IGSL gold hunter (remote areas), assuming ALL METAL mode has the same sync pulse as the GEB (cancel ferrite).

This is the idea, although in practice the slight disadvantage of the DISC channel for gold may be hardly an issue; probably not.

One other thing puzzles me - my past observations noted that the DISC channel seemed to give a larger signal for most metals than the GEB channel. Although I don't have any gold handy, I believe I tested foil and observed the same thing. However, our discussion here suggests that the GEB channel should have a stronger signal to gold/foil. Maybe my observations were mistaken, I'll check again.

-SB

[Davor]

One other thing puzzles me - my past observations noted that the DISC channel seemed to give a larger signal for most metals than the GEB channel. Although I don't have any gold handy, I believe I tested foil and observed the same thing. However, our discussion here suggests that the GEB channel should have a stronger signal to gold/foil. Maybe my observations were mistaken, I'll check again.

What you saw was most probably ground response. It is not nulled in disc channel, just motion compensated, and it must have been big. So in effect ... You did! You did tee a putty tat!

OK, copper is on the appreciated side for the disc channel, and what you saw was exactly that. Small gold and foil are more appreciated in a GEB channel.

There may be a reasonably easy way to disable excessive channels and port only GEB through, so that you can listen to anything that is nearly metallic in the ground. It would maintain full benefit of nulled ground and/or ferrites, and it can be motion compensated too for the benefit of your ears. I admit I'm partial for the IGSL solution, I love it actually, but this mod can be done. Once you finally find something, you may engage discrimination to find out what that may be. It may be done as the only function of some incredibly simple rig, but I'm too fond of discrimination to even start thinking any further about it.

[simonbaker]

What you saw was most probably ground response. It is not nulled in disc channel, just motion compensated, and it must have been big. So in effect ... You did! You did tee a putty tat!

OK, copper is on the appreciated side for the disc channel, and what you saw was exactly that. Small gold and foil are more appreciated in a GEB channel.

There may be a reasonably easy way to disable excessive channels and port only GEB through, so that you can listen to anything that is nearly metallic in the ground. It would maintain full benefit of nulled ground and/or ferrites, and it can be motion compensated too for the benefit of your ears. I admit I'm partial for the IGSL solution, I love it actually, but this mod can be done. Once you finally find something, you may engage discrimination to find out what that may be. It may be done as the only function of some incredibly simple rig, but I'm too fond of discrimination to even start thinking any further about it.

Good thoughts. I was probably air-testing targets, but I suspect I wasn't attentive to both channels, so probably the channels respond the way you describe, Sylvester.

I have several interests for designs:

1. Super-sensitive, super-deep detector at any cost.

2. Best discriminating, responding detector, depth not as important.

3. Simplest, cheapest and pretty good MD that's easiest to build.

The single channel design falls into the last category.

It would seem for the TGSL/IGSL, ALL METAL mode is pretty close to the "GEB only" mode you mention -- I think the TGSL does shift the DISC channel a little bit to maybe pick up coins better, but by dumping the resistor and capacitor (R24, C11 TGSL) in the AM circuit it would probably give you a nice "gold hunter" in AM mode, if you preferred that. I'm tempted, since I don't use the AM mode much now for coin hunting.

-SB

[Davor]

IMHO the first point is a myth. By dumping a shovel-full of exotic and expensive components into a steam punk casing and a few dashes of glitter here and there, not to mention lots of marketing gee-gah and doohickeys ... will not grant you a super duper machine. Mere quantity of bells, whistles, and bucks does not work that way. Too bad many people fall under impression that it does.

Intelligent design and a clever interface make a big difference. In case of a cleverly optimised design and an interface that you feel natural with - that's it.

I've noticed someone on this forum speaking fondly of his beloved Tejon, and how he misses her ... that's it.

[simonbaker]

IMHO the first point is a myth. By dumping a shovel-full of exotic and expensive components into a steam punk casing and a few dashes of glitter here and there, not to mention lots of marketing gee-gah and doohickeys ... will not grant you a super duper machine. Mere quantity of bells, whistles, and bucks does not work that way. Too bad many people fall under impression that it does.

Intelligent design and a clever interface make a big difference. In case of a cleverly optimised design and an interface that you feel natural with - that's it.

I've noticed someone on this forum speaking fondly of his beloved Tejon, and how he misses her ... that's it.

Agreed -- either the design performs or not. Who mentioned doohickeys? It's just one design goal.

-SB

[golfnut]

2. Best discriminating, responding detector,

After what qhahzi said yesterday about how Fe and non Fe signals appear to the comparator - I finally get it. phew.

From that knowldge, I was thinking of my case in the Iron rich fields. plaques of cast iron the size of your hand scattered everywhere.

Think of the integration of these targets on the disc Ch .... you get a v large -ve signal which holds the comparator off. A small positve going non Fe signal could be in there but you cant flip the comparator due to the elephant in the integrator giving massive -ve charge. 9 parts -ve and 1 part positive is 8 parts -ve - So Iron all the way no hear your real target.


Were doomed - ,, hang on - the Disc Ch could be uni polar - So you have no massive -ve signal from Iron . Iron would not register,

So a small positive signal could still set tone off. Currently the Dixc Rx Ch would have a belly full of -ve sig on it and no chance to hear the gold.

Sure theres reasons why Im wrong - but cant see too many from here.

Not certain how to go about it though - thats for later..

There was a guy on here saying his IDX PP was rubbish - Swap your Rx lines over mate..

Mine was the same - its really good now, also Iron has quietened right down as it would and its a joy when built right!! Not hugely deep - but very usable.



S

[Davor]

...which holds the comparator off...

That and a lazy anti chatter.

I suggest two steps here. First remove the capacitor in a second stage channel amplification, and second, put an anti-parallel diode with the existing one(s) in that very same stage. That makes for snappy recovery from iron saturation.

Furthermore in TGSL/IGSL reduce the antichatter capacitors to ~1/5 current values. These will result in far more agile rig, which is a bit "nervous" but even with large chunks of iron you'll get at least a short beep in presence of non-iron.

[simonbaker]

After what qhahzi said yesterday about how Fe and non Fe signals appear to the comparator - I finally get it. phew.

From that knowldge, I was thinking of my case in the Iron rich fields. plaques of cast iron the size of your hand scattered everywhere.

Think of the integration of these targets on the disc Ch .... you get a v large -ve signal which holds the comparator off. A small positve going non Fe signal could be in there but you cant flip the comparator due to the elephant in the integrator giving massive -ve charge. 9 parts -ve and 1 part positive is 8 parts -ve - So Iron all the way no hear your real target.


Were doomed - ,, hang on - the Disc Ch could be uni polar - So you have no massive -ve signal from Iron . Iron would not register,

So a small positive signal could still set tone off. Currently the Dixc Rx Ch would have a belly full of -ve sig on it and no chance to hear the gold.

Sure theres reasons why Im wrong - but cant see too many from here.

If the DISC pot is set to "exactly" cancel the iron, then you should not get a belly-full of -ve signal, right? Or do I not understand what you mean by the "ve" signal?

So if you are in iron-trash area, don't set your DISC pot too high, which would make the Fe response go negative -- try to find a DISC setting where the Fe response is exactly zero. Of course, you will detect other metals now, such as nickel, etc.

Also, I agree with Davor -- if you make a really "snappy" (fast response) MD, then small gold/non-Fe targets on large-sized iron may still be detected due to the motion effect, even if the DISC pot is not set to zero the iron.

-SB

[golfnut]

If the DISC pot is set to "exactly" cancel the iron, then you should not get a belly-full of -ve signal, right? Or do I not understand what you mean by the "ve" signal?
-SB

With disc set to cancel Iron it accepts real target phases.




with Fe larger than target mass - you get a -ve going result larger than small +ve going - so on average you get a -Ve signal holding the comparator off - no tone -- But you miss the real target.

Like -240mV of Fe sig and +80mV of real target together gives net result of -160mV and miss target.

Same for any machine like this - not knocking IDX - just learning.

S

[golfnut]

struggled with pic attach!!

[simonbaker]

With disc set to cancel Iron it accepts real target phases.




with Fe larger than target mass - you get a -ve going result larger than small +ve going - so on average you get a -Ve signal holding the comparator off - no tone -- But you miss the real target.

Like -240mV of Fe sig and +80mV of real target together gives net result of -160mV and miss target.

Same for any machine like this - not knocking IDX - just learning.

S

I agree, and how to best deal with it depends somewhat on the sizes and phases of the trash.

I don't think it is a matter of the iron being larger size than the target. Rather, if you could know the phase of the iron trash, you could set your DISC control to exactly "zero" the response to the typical piece of iron you are worried might hide your real target. If set it that way, the iron should not hide your target, because the iron adds exactly "zero" signal without going negative.

The success of that depends on how much the phase of iron depends on the size of the piece, and the range of sizes of trash. I was hoping it is a reasonably small range so you could pick a good point to set your DISC pot to "zero" most of the iron without the "-ve" problem causing much hiding of targets.

If the data in the graph Davor posted is accurate, large pieces of Iron will make the phase go more positive than small pieces. So we know that to "knock out" iron from making false signals, we'd need to move the DISC setting to make large iron signal weaker (close to zero response) than our target signal. However, smaller pieces of iron would try to make the signal go negative and potentially hide targets.

And I agree, not much we can do about it if our trash has many different phases. Two targets of different materials very close together will probably combine in phase to look like a different metal. Digging is required.

-SB

[golfnut]

"you could set your DISC control to exactly "zero" the response to the typical piece of iron you are worried might hide your real target. If set it that way, the iron should not hide your target, because the iron adds exactly "zero" signal without going negative."


If you move your Disc to Integrate out the Iron - you loose half of the Target signal..

[simonbaker]

"you could set your DISC control to exactly "zero" the response to the typical piece of iron you are worried might hide your real target. If set it that way, the iron should not hide your target, because the iron adds exactly "zero" signal without going negative."


If you move your Disc to Integrate out the Iron - you loose half of the Target signal..

That's true, but anytime you have DISC and GEB channels, you lose some target signal. Because DISC and GEB channels have different sync pulse phases, one or the other will not be maximum target signal, and because both must exceed the threshold for a beep, you will never get "maximum" sensitivity no matter what.

We usually accept this -- for example, if you are looking for silver and you discriminate out copper, you will lose most of your silver signal, it is a fact of life. Iron at least is farther away from most target metals and zeroing it will not hurt as much usually.

To summarize, if you "zero" out iron, you will maximize whatever metal is 90 deg more phase. For example, if you "zero" the DISC channel response to large nail at 40 deg (referring to chart Davor posted above), you will maximize response to 40 + 90 = 130 deg, which is close to zinc and large gold ring. Of course if the GEB channel is set to "zero" ferrite at 0 deg, then the GEB channel is maximum at 90 deg, which is small gold ring / small gold chain, and that will limit the total response to the large gold ring (although not much).

So I guess we need to know what we are looking for and what type of trash is the most problem and check the chart and try to pick the best DISC and GEB settings to do the job.

To me, if I'm looking for gold far away from typical "picnic" trash, I would be tempted to use ALL METAL mode where both channels are set similar to the GEB channel and should be quite optimal for small gold pieces, I would think. But if much iron is around, then probably no choice but to zero out the iron as best as possible and lose some of the gold signal.

We also have to remember, this discussion is about not losing target when iron and target are very close together spatially. Otherwise, any DISC setting that makes iron zero or negative is fine.

I'm definitely going to think about modifying the TGSL AM sync pulse to get a square wave if possible, and maybe without phase shift, since it might make a useful gold hunter.

-SB

[Qiaozhi]

That's true, but anytime you have DISC and GEB channels, you lose some target signal. Because DISC and GEB channels have different sync pulse phases, one or the other will not be maximum target signal, and because both must exceed the threshold for a beep, you will never get "maximum" sensitivity no matter what.

You seem to be getting yourself a little confused here.

Although the GEB and DISC sample pulses occur at different times, the target signal is integrated by the channel filters, and changes slowly when compared to the raw RX signal.

[Davor]

OK, slow down a little bit. To remain accurate we must establish some nomenclature here, otherwise this is going to be a mess.
Simonbaker noticed correctly that sampling a target signal with two phases produces two different signals at lower amplitude (if I understood it correctly), and that is mostly correct. When you multiply (sample) some signal by two sampling signals that are at 90° e.g. in perfect quadrature, you'll get products that also have orthogonal properties. The funniest property of all is that noise is reduced by the same rate as the signal and noises from different channels are mutually orthogonal, and that is a good thing. So in fact you do have conversion losses, but your noise also has "conversion losses" so your S/N is not ruined. Ability to discern small signals comes from S/N so that's good news.

Regarding integration ... it depends mostly on the build. TGSL/IGSL have severe filtration here, but IDX doesn't. Thing to remember is that target response happens simultaneously in all channels, but may reverse phase and amplitude in a process. So copper coin will have positive response simultaneously in both Disc and GEB channels. In case we have an incompetent design ... which are practically all of them unfortunately, you have level dependent durations of channel response, and these lead to utter confusion. It is very easy to solve: just remove capacitors in feedbacks of the last channel amplification stages and you are saved. Severe integration in subsequent stages just helps sorting out the dominant signal, and reducing the "chatter" that is a direct consequence of design problems.

So first and the most important - target feedback signal happens simultaneously in all channels, and variations in duration are merely a problem in design that is easily solved. Ideally these would happen simultaneously, but different levels of saturation take care that they don't. Excessive integration to avoid these problems :duh: is yet another design problem introduced to fix the initial one. As a consequence of these two design problems -- you have masking.