PI Monocoil - sampling under 4us almost impossible?

[B^C]

Gday Nicolae,

Can you post a screenshot of the simulation you have done with these components--Voltage & Amps at the Inductor vs the Initial trigger pulse at V2, i haven't got the Mosfet or Diode models you have used.

I will run some simulations from the models i have & can compare them.

[Woody.au]

Hi Woody,
I think lowering the dead time improves sensitivity to all small targets (this includes 0.5g...5g gold nuggets). Tinkerer explained to me about the response of small targets. A 1g gold nugget's response decreases a lot with every microsecond and there are two ways to increase the sensitivity of the detector: either shorten the dead time, either increase the gain of the circuit. There are physical limits to both of these.
Increasing the gain also increases susceptibility to electrical noise.
Overall, I would like to make a circuit that can detect a target with a TC=5us (that's a 0.5 square inch thin aluminium foil) to over 15cm and a target with TC=10us (1 square inch alu foil) to over 25cm.
Could you tell us what's Minelab sensitivity to these targets in the standard (unmodified) version, and after your mods? I know that in the domestic environment you're using the fig 8 coil, that has reduced sensitivity, but I can also use a similar setup and we can compare the results. You'll have to bear with me though, it will be some time until I get my electronic components to do some improvements to my circuit.
When you get out in the field with a Minelab SD2000 or higher, maybe you can make a short movie about its discrimination abilities. I haven't seen such a movie on YouTube.

Regards,
Nicolae

Minelab P.I detectors discrimination is not worth even thinking about, it is useless on small Ferrous objects at medium depth. Ask how much rusty metal is dug by prospectors that use M/L P.I detectors. He who makes a full depth P.I discriminator will be a wealthy man.

As for faster timings, yes that is a good thing, Minelab cannot sample too close as the input switching fets have a tendency to "ring" after turning on and off. They also have tried to speed up the input opamps by clamping the input to ground with a Dual Gate Mosfet to either speed up or clean up the signal. This in on models after the SD2200. They also switch the dampening resistor in and out of circuit at critical times.
I think that the earlies sample they take is 15us, that is a long time and obviously can be greatly improved upon.

As for Video's I have to make some more, I want to field test comparisons as to showing older detectors outperforming the newer ones.

[hobbes_lives]

I think that the earlies sample they take is 15us, that is a long time and obviously can be greatly improved upon.

That is longer than I would have expected from a Minelab. What sample time have you been able to achieve on a modified unit?

[Woody.au]

That is longer than I would have expected from a Minelab. What sample time have you been able to achieve on a modified unit?

I take then down to around 10us but its a very hard task to keep them stable, especially with a Mono Coil. Then again with fast timings and a small coil the detectors run ok. The "problem" with speeding up the timing is the speed of the front end switching keeping up with its 2 series fets.
Get too close to the flyback and it takes the first fet out. There is more to be gained in matching the input impedances, coil = 0.5 ohm fets RDS on = 20 ohm Opamp input is 10 ohms. Can anyone see a power transfer problem?

[nick_f]

Hi Nicolae:

My LTSpice (ver 4.02c) does not have 1N4937 diode. Where do we get that?

Regards,

-SB

Hi guys,

I use the version 4.02o and that one didn't have the diode either. I downloaded heaps of models from http://tech.groups.yahoo.com/group/LTspice/

Here is a crash course that will raise more questions than give answers.

In order to add diodes, transistors to the library, you can check these folders:
C:\Program Files\LTC\SwCADIII\lib\cmp - it contains models for components. The file standard.dio is a text file with diode models.

You can add these lines at the beginning of the file (from line 4). Some of them are already present, do not add them twice.

.model 1N4148 D(Is=2.52n Rs=.568 N=1.752 Cjo=4p M=.4 tt=20n Iave=200m Vpk=75 mfg=Motorola type=silicon)
.model 1N4937 D ( IS=830n RS=34.0m BV=600 IBV=5.00u CJO=26.5p M=0.333 N=2.46 TT=288n mfg=Fairchild type=silicon)
.model 1N34A D (Is=2e-7 Rs=7 N=1.3 Cjo=0.5e-12 M=0.27 Eg=0.67 BV=75 IBV=18e-3 Vj=0.1 type=germanium)
.model 1N4001 D (Is=14.11n Rs=33.89m N=1.984 Cjo=25.89p M=.44 Eg=1.11 BV=50 IBV=10u Vj=.3245 Ikf=94.81 Xti=3 Fc=.5 Tt=5.7u mfg=Motorola type=silicon)
.model 1N914 D(Is=2.52n Rs=.568 N=1.752 Cjo=4p M=.4 tt=20n Iave=200m Vpk=75 mfg=Motorola type=silicon)
.model BAS16V D (IS=412p RS=0.140 BV=75.0 IBV=1.00u CJO=1.72p M=0.333 N=1.70 TT=5.76n mfg=Fairchild type=Schottky)
.model BAS16TW D (IS=68.6n RS=0.377 BV=75.0 IBV=1.00u CJO=1.72p M=0.333 N=2.34 TT=5.76n mfg=Fairchild type=Schottky)
.model BAV99 D (IS=261p RS=0.140 BV=75.0 IBV=2.50u CJO=1.19p M=0.333 N=1.70 TT=5.76n mfg=Fairchild type=Silicon?)
.model MMSD4148 D(Is=2.52n Rs=.568 N=1.752 Cjo=.64p M=.4 tt=5n Iave=200m Vpk=100 mfg=Onsemi type=silicon)
.model 1N5817 D(Is=31.7u Rs=.051 N=1.373 Cjo=190p M=.3 Eg=.69 Xti=2 Iave=1 Vpk=20 mfg=Motorola type=Schottky)
.model 1N5818 D(Is=31.7u Rs=.051 N=1.373 Cjo=160p M=.38 Eg=.69 Xti=2 Iave=1 Vpk=30 mfg=Motorola type=Schottky)
.model 1N5819 D(Is=31.7u Rs=.051 N=1.373 Cjo=110p M=.35 Eg=.69 Xti=2 Iave=1 Vpk=40 mfg=Motorola type=Schottky)
.model BAT54 D(Is=.1u Rs=2.2 N=1 Cjo=12p M=.3 Eg=.69 Xti=2 Iave=300m Vpk=30 mfg=Vishay type=Schottky)

These are some models which I found on the internet (for example, for diodes, check the web site: http://www.diodes.com/spicemodels/category.php?category=2_1_fast_recovery_rectifiers

For example, have a look at this link: http://www.diodes.com/products/catal...p?item-id=2920

There is a model for diode BAS70. You can copy the model and paste it to the standard.dio file and then edit it manually to fit the same pattern as on the file.

For bipolar transistors, the models are found in standard.bjt, etc.
The file standard.jft contains model for JFET transistors.

There are more complex models, called subcircuits and they are found in the folder C:\Program Files\LTC\SwCADIII\lib\sub
To make life for beginners hard, the subcircuits can have any extension we want or don't want. The most common ones are .sub, .mod, .cir, .301
You can change any of them to whatever you want, but you have to provide information in the .include statement with the model file name.

I put in there the models for the MOSFETS I added. In order to use the model for a mosfet, you have to know what symbol to use. New symbols can be added to the folder: C:\Program Files\LTC\SwCADIII\lib\sym

The file nemos3t.mod contains models for many IRF transistors (I provided the file in the zip archive).

For those who are serious about Spice simulations, please check the yahoo forum http://tech.groups.yahoo.com/group/LTspice/ and read some tutorials like http://www.uta.edu/ee/hw/pspice/pspice01.htm

The symbols used in diagrams are placed in C:\Program Files\LTC\SwCADIII\lib\sym and their extension is .asy.

If you open in Spice my file IRF720_HH.asc (.asc means schematic), and right click on a MOS transistor, you will see at the top the symbol used (C:\Program Files\LTC\SwCADIII\Nicolae\nmos4.asy). Since I am very much a beginner as well, a few days ago I didn't know I should place all the symbol files in the "\lib\sym" folder, so I placed it in my personal folder.

Next, you check the prefix. We use X when we want to use our imported (or created) circuits, not part of the LTSpice libraries. Also, the value field has to contain the part we want, in this case: XIRF740

The model of that part is found in the file nemos3t.mod
We need to tell LTSpice where to look for the transistor XIRF740, and we do this by using the directive .include nemos3t.mod

If you have questions, please ask one thing at a time, so I can provide clear answers (in case I know). Usually the answer of any question can be found on the internet.

Please let me know if any of you found my messy tutorial helpful

Regards,
Nicolae

[nick_f]

Minelab P.I detectors discrimination is not worth even thinking about, it is useless on small Ferrous objects at medium depth. Ask how much rusty metal is dug by prospectors that use M/L P.I detectors. He who makes a full depth P.I discriminator will be a wealthy man.

Hi Woody,
I am very surprised to hear ML are not very good discriminators. I was expecting that for that price, to have a discrimination comparable with the VLF. As far as I understood, there are some detectors out there that have better depth than Minelab, is that true? I am talking about the commercial, off-the-shelf ones, not modified. Anybody can confirm this?
Woody, I am regularly checking your videos, keep up the good work!

Regards,
Nicolae

[simonbaker]

Hi guys,

Please let me know if any of you found my messy tutorial helpful

Regards,
Nicolae

Yes, very helpful - program config is always 95% of the difficulty.

Somehow I can't get the ".mod" include to work, even with absolute path. Puzzling over that...

Cheers,

-SB

[nick_f]

Hi Simon,

Please check if you placed the nemos3t.mod file in the folder \sub and and use my original .asc file for the beginning. You should be aware you can have text comments and directives placed on the schematics. The directive you have to use is:".include nemos3t.mod" (with a dot at the start). You can create a directive either by clicking on the .op button (top right), or Edit->Spice Directive or simply by pressing S key.
You can right click on a directive and change it to a comment or the other way around (comments are blue, directive are black).
If you still can't make it work, please send me an email with the error details, maybe I can help you.

Regards,
Nicolae

[B^C]

Thanks Nicolae,

Your tutorial is good--thanks.

Nothing compares to Moneylab detectors, in air testing everything else is better but in ground conditions moneylab are far better than anything else, especially in bad ground, if there's a better detector i would buy it today regardless of cost or name.

All the bench & air testing or talk in the world won't make a detector better, i test everything in the worst soil i have found, this tells you what's what & why trying to compare things is difficult.
Air testing has a way of giving false information & wasting a lot of time.

People may not agree but i have run a 4500 against all other detectors including the modded ones "to date" & have outdone them all by a good margin. I prospect for gold in very bad ground because this is where it is, so we need detectors that work in these conditions & not air or bench testing.

I have a mod that is extrememly useful & is powerful in bad ground, i have been setting up test equipment & should be able to show some results by next week hopefully. Finding out why this mod works so well is difficult but i have found various very interesting things so far.

It seems people are trying to build a better detector without know what's required, to me it's more valuable to start with the target response in bad ground & work backwards from there so you can build what's required.
Most people start with the detector & leave out the important part, the target in soil conditions.
The target response in air has no resemblence to a target response in soil or bad ground conditions.

Not much good trying to reinvent things, firstly when you don't test where things work & secondly when you don't know what's required from the target response side of things.


As for the discimination on the 4500 or previous moneylab detectors, Woody is correct, for larger deep or smaller near surface ferrous targets it's ok but as soon as the ferrous targets are a little deeper or smaller again, you have to dig them.
Even a boot tac near the surface will not discriminate, if you test the target response signal in soil conditions you will see why there is so much trouble trying to discriminate. Again discriminating in air testing won't do anything once the targets are in the ground.

I think people get misled, on some of the older detectors you could infact discriminate--a bit, but lets not forget most of the older detectors & most of the newer detectors besides moneylab are near surface detectors.
Most have trouble finding a penny at eight inches in good ground where a money lab will find this deep in bad ground.

From some recent in soil testing a deep or very small ferrous target response looks much like a shallow small gold nugget target reponse so trying to disciminate between the two is nigh impossible. The reasons are obvious when you look at the signals.

I wish people would spend more time looking at the Targets signal response than trying to build a better detector without knowing what's required, then we will get somewhere.

Woody,

As soon as i get a handle on this mod i will contact you, we may be able to do something worthwhile.

[nick_f]

Hi B^C,

That's one of the best insights I read about metal detectors and Minelab characteristics lately. Thank you very much for sharing your knowledge.
I have some comments about the air tests. I consider it is very true that they don't accurately reflect the capacity of the detector of working in bad ground conditions. However, I think they are still very useful up to a stage. For example, without air tests, how would I know if I improved the sensitivity of my circuit? And if a detector can't detect a coin at 5cm in the air, don't even bother going outside with it.
Also, with the discrimination, if a detector can't make the difference between iron and copper in the air, I doubt there is any chance to do that in the ground.
Therefore I see the air tests as the first stage of testing a detector. The second stage would be testing targets in the ground (or bad ground if available ) .
You're in Australia, aren't you? Maybe one day you'll take me with you for some prospecting. Never been searching for gold before and I am looking forward to the first adventure


Regards,
Nicolae

[porkluvr]

Yes, very helpful - program config is always 95% of the difficulty.

Somehow I can't get the ".mod" include to work, even with absolute path. Puzzling over that...
-SB

Forgive me if I tell you stuff you already knew...

You can edit third party symbols ("3rdP" symbols) by either (1)clicking on the file name, or (2)dragging it into an open LTspice window,
or (3)right clicking it from within an open project and then clicking "Open Symbol".

If you want to edit (or examine) a symbol that was supplied with LTspice, method (3) will not work.

CTL+A will open the "Edit Attributes" window for a symbol. If you examine the attributes of an LTspice symbol (LT1037.asy, for instance)
you will notice that they have filled in the blanks a little differently than I have in my examples.
So, obviously there is more than one way to create usable symbols.

I create a different symbol for each part. Although it may seem like unnecessary trouble, I make notes about the part
within the symbol, and can edit attributes so that LTspice knows where to find the model file, making the "include" statement unnecessary.

You can use one symbol as a template for similar symbols, so making new symbols is a snap.

I rename all model files to *.sub, unless the extension describes something, like ".501" indicates a model characterized for a
single 5V supply (as TI's method).

edit:

[nick_f]

Thanks porkluvr,

I also found some new stuff. Thanks for posting. I will also include the model in the symbol, I think it's more convenient this way.

Regards,
Nicolae

[simonbaker]

Thanks porkluvr,

I also found some new stuff. Thanks for posting. I will also include the model in the symbol, I think it's more convenient this way.

Regards,
Nicolae

Hi Nicolae:

I was able to run your circuit after doing this:

I right-click on the U1 XIRF720 part, and change the Prefix from "MN" to "X".

Does yours work without that?

Thanks everyone for all LTSpice advice, I always have problems with configuring it, so always welcome tips.

Regards,

-SB

[nick_f]

Hi Nicolae:

I was able to run your circuit after doing this:

I right-click on the U1 XIRF720 part, and change the Prefix from "MN" to "X".

Does yours work without that?
-SB

Hi Simon,
It's very good you were able to run the simulation.
I also had to change the prefix to X. I actually mentioned in the crash course we have to change it to X (I was in a hurry and I didn't specify what was the original prefix MN).

("Next, you check the prefix. We use X when we want to use our imported (or created) circuits, not part of the LTSpice libraries. Also, the value field has to contain the part we want, in this case: XIRF740")

Regards,
Nicolae

[simonbaker]

Hi Simon,
It's very good you were able to run the simulation.
I also had to change the prefix to X. I actually mentioned in the crash course we have to change it to X (I was in a hurry and I didn't specify what was the original prefix MN).

("Next, you check the prefix. We use X when we want to use our imported (or created) circuits, not part of the LTSpice libraries. Also, the value field has to contain the part we want, in this case: XIRF740")

Regards,
Nicolae

Here is the updated file with the X prefix as convenience.

Regards,

-SB