need osciloscop test who can help

[okantex]

Hi Everybody,

I am planing to have discrimination with PI s.

to do this I need some scop results.

I want to see the end (die) times in microseconds and movement of grafic at
these conditions ;
we will use two meterial types.gold and iron.
coil can be small coil.and must be constant during all tests.

1) change of grafic while changing distance of gold(ring or bracelet) to coil.
measurement distances are 10cm,20cm,30cm,40cm,50cm

2) same for iron meterial.but size must be close to gold sample.

3) this time distance is constant.you can chose suitable distance (10,20,30,40) according to tests.
we will change size of target. both for iron and gold.
ı know it is difficult to find gold .but use anything you have with different sizes.
and also double reflection area of iron at each readings.like surface area=A
first A ,seceond 4A ,than 8A,16A . we must start A from coin size.
with an other aspect increase diameter .for example
D=2cm
D=4cm
D=8cm
D=16cm
D=24cm
D=32cm

4) last one .for this test important result is end times.(microseconds) of different targets.
targets are all gold.but different values gold. I do not know how to say it in english. kirat ,degree, like 16k (for rings) ,18k ,22k (for bracelets)

[Max]

Hi Everybody,

I am planing to have discrimination with PI s.

to do this I need some scop results.

I want to see the end (die) times in microseconds and movement of grafic at
these conditions ;
we will use two meterial types.gold and iron.
coil can be small coil.and must be constant during all tests.

1) change of grafic while changing distance of gold(ring or bracelet) to coil.
measurement distances are 10cm,20cm,30cm,40cm,50cm

2) same for iron meterial.but size must be close to gold sample.

3) this time distance is constant.you can chose suitable distance (10,20,30,40) according to tests.
we will change size of target. both for iron and gold.
ı know it is difficult to find gold .but use anything you have with different sizes.
and also double reflection area of iron at each readings.like surface area=A
first A ,seceond 4A ,than 8A,16A . we must start A from coin size.
with an other aspect increase diameter .for example
D=2cm
D=4cm
D=8cm
D=16cm
D=24cm
D=32cm

4) last one .for this test important result is end times.(microseconds) of different targets.
targets are all gold.but different values gold. I do not know how to say it in english. kirat ,degree, like 16k (for rings) ,18k ,22k (for bracelets)

Hi okantex,
I think that is not a good idea...cause different objects made of different materials could show same decay end time...
I read about in some older post or article...can't remember now.
But the problem is that the decay is much more related to eddy-currents that flows in the target than the actual target composition.
I think that many have tested before this approach without good results on the discrimination problem.
If e.g. you think at Minelab they use 3 or more frequencies to (and different tx pulses) to make some kind of disc...that anyway is accurate in PI machines.
Disc by conductivity/eddy current density is too limited to be useful in the real field with real common targets.
Real disc system need to be much more complex cause requires a lot of computation and really fast hardware-mcu.

Anyway, do it if you want cause could be an interesting experience to know better PI designs.

Best regards,
Max

[okantex]

hi max
I thought if we use big coils.which means we wont get signals from small items.we could have chance to disc.
only gold and irons delay is same and it is 250us .but there is difference in grafic.and it can be analyzed by PICs.

to be sure of this .I need to see some test results.

[okantex]

why there is no interest to this topic.
do not you want disc at dept targets

[Carl-NC]

What you are asking for is difficult, if done correctly. You need to measure multiple points along the decay, let's say every 5us from 5us to 100us. That's 20 measurement points. This needs to be done, like you say, for different distances, different target sizes, and different metal compositions.

But the hard part is resolution. You can't just put an oscope on the coil and visually measure. You will need a significant preamp stage, preferably one that does not alter the decay. You need to feed that into a high-res ADC and capture the measurement points. To remove noise, you need to capture the same decay a bunch of times, and digitally average the data. To make it all meaningful, the coil and target should be set up in a jig, not just handheld.

This a lot of work, which is probably why no one has jumped on it. But it would be interesting to do.

- Carl

[okantex]

hi Carl
resolution can be 15 us.bu t someone says I can increase resolution why not.
not need different metal compositions.just gold and iron samples can be enough.because their delay is same.
rest can be seen after PIC programming.but before software I must differ gold and iron. this test is just for it.

[Max]

hi Carl
resolution can be 15 us.bu t someone says I can increase resolution why not.
not need different metal compositions.just gold and iron samples can be enough.because their delay is same.
rest can be seen after PIC programming.but before software I must differ gold and iron. this test is just for it.

Hi okantex,
would be a good idea to automate the sampling process:
- one could build a data-logger interfaced to a conventional PC using some interfacing chips e.g. max232 to get RS-232 levels...or simply by the parallel port (LTPx)...that works at 5v. Interface could be made by e.g. an old audio-card...isa-eisa-pci etc...giving sampling of received signal amplified by an ultra low noise op. amp (e.g. op-37).
Just make the program...then have some kind of sequence when you press e.g. RETURN...meaning starting from 50cm...then 40...and so on.
This will speed up the sampling process...and one can also manage multiple sampling (e.g. 10 or 100) at every step and then leave the program making some average. Then the program could give some output e.g. for excel or something to get visual indications of meaninful data...etc etc
this way one could investigate in a scientific way how many different metal targets interfere with decay signal...and then find an alghoritm or something to be implemented in some MCU to do that kind of comparision using similar hardware e.g. ADC, low-noise amp, sampling gates etc.

Seems very interesting but requires also much time to be done the right way...and problem is that there isn't warranties on results...so one could discover after all the work that there are too overlappings between iron and gold decay if some condition happens...etc

I think one could do...but it's a big time investment and not for an average hobbist , with many time constraints and also other difficulties...but for sure one could do if time-skills-hardware aren't problems.

Best regards,
Max

[okantex]

ofcourse this can be only done by programmer but it is too difficult than using ociloscop.cause firts he has to program it as to use osc. than evaluater results than program for hunting mode.
and programming for osc is difficult than programing for huntng mode.because needs too many parametters.
so need assistance of osc.users.
the test I discribed is not so hard to do

[Max]

ofcourse this can be only done by programmer but it is too difficult than using ociloscop.cause firts he has to program it as to use osc. than evaluater results than program for hunting mode.
and programming for osc is difficult than programing for huntng mode.because needs too many parametters.
so need assistance of osc.users.
the test I discribed is not so hard to do

Hi okantex,
it's not too difficault to do. For an expert programmer it could be made in few hours (or minutes ...depends)...using also e.g. visual basic (that is really simple language) instead of more serious programming e.g. C or assembly code. Just one need to access low level services like ports i/o that could be tricky in windows2000/nt/xp/vista without proper driver installed...but I want to say that using that driver(s) it become really easy to do...one driver useful is e.g. "porttalk" for windows , but there are many others...and also some programming IDEs have their own.
Yes...using e.g. a digital oscope and reading values by RS232 or similar is good...but I was thinking on made a specific device...with scrap hardware...e.g. old computer sond-card...and some few external components...a pic...some 4066...and the like. So also without having a digital scope one could measure and logs readings.
Yes, using an oscilloscope directly could be much easier...but to have good results and make good measurements one must repeat many times...manually...and better if using a digital scope cause many errors could be made by direct lecture of values...you know...need continuos calibration and visual errors are even present in old fashioned oscopes...the analog ones.

Anyway...it's time the big problem here...all this stuff requires some work and so also time do be done.

Best regards,
Max

[okantex]

Hi Max
I am trying to learn PIC programming.so you can say I am seriosly programmer.
I confess I cannot do osciloscop by program.
but it is a must that we have to do this test to achive some disc.
yes , you can not carry labtop in field searches.but we can do metal analyzzes with pc and than carry that knowladge to PICs.

so we have to make path for us.needs time and work.you are right.
and we have nights for us to do such a project.

do you have old components around there to make PC osc.

[Max]

Hi Max
I am trying to learn PIC programming.so you can say I am seriosly programmer.
I confess I cannot do osciloscop by program.
but it is a must that we have to do this test to achive some disc.
yes , you can not carry labtop in field searches.but we can do metal analyzzes with pc and than carry that knowladge to PICs.

so we have to make path for us.needs time and work.you are right.
and we have nights for us to do such a project.

do you have old components around there to make PC osc.

Hi okantex,
yes I have dozens of old pc hardware cards etc...and also digital scope...and many other stuff...my real problem is that, at the moment, I work a lot during day...and when a come back home I feel really tired...and just watch the tv for say 5minutes and then fall in the arms of Morpheus...(sleep)...well not just this way...but pretty close...I do some two or three other things before sleeping... watching the tv...anyway
Night hobbies are not for me...I need to sleep for at least 8 hours a day....cause if it's not so...I'll be tired the next morning and can't work...or work so bad (you can't imagine which kind of problems could be created there...).
I spend some time in my electronic hobby where things become easy on work and I have a bit more time to do something ...some circuit or another. Sometimes I make everything in one afternoon...e.g. a pic programmer...or other simple device...and sometimes takes weeks or months to get something work...cause I can't spend more than 30minutes a day or even less or no time at all!
My workload is variable...but always present and I can't do the kind of testing you need to elaborate your alghoritm for disc in PIs...but I can maybe contribute to the hardware design of data-logger if someone jump in this adventure...and maybe some maths too or some programming...but not the real testing - measures - pcbs... and the like. Takes much time that simply I haven't !

Do you belive me if I say that I've tested my bandidoII only 5 times at the moment on the real field ??? It's so depressing...having little time for your preferite hobby.

Sorry, my friend ...but time is one of my big limitation !

I hope you will find some people helping in this project cause seems really interesting to me.

Best regards,
Max

[okantex]

Hi Carl

why do you stay away?
as you say we must measure from opamp.you are right.
we can use first or second opamp at DP.why not to use LF357.
I know you have gold bar if I do not remember wrong.and sure like every woman your wife must have gold ring and bracelets.good samples.

and you can easyly find iron items at different size.I mean ring size to ball size
if we place small coil motionless and draw this targets by hand or hepl of plastic or wooden aparatus. we can do this experiment.
distance is not so much important.it can varry 1-3 cm at each step.like
11-19-32-41.
this will help us to see decay times and decay curve's change according to metal and distance.

we must start from somewhere

[Carl-NC]

Like Max, I just don't have the time. I still think the most useful results will come from a high-res digitizer with digital averaging, not an oscope. Even a digital oscope, most of which have only 8 bits of instantaneous resolution.

This is not a quick and easy test.

- Carl