CHANCE Pcb's Finally Here

[baum7154]

Chance PI Build Instructions

Silverdog, I received the Chance PI board and parts from you and the board looks great! I am wondering if anyone on this forum can point me to a site that has construction information and details. It appears the there are 3 wire jumpers to be soldered onto the board near the microprocessor. Can anyone confirm this?

Also does anyone have coil build information for this PI, i.e. coil inductance and resistance specifications? Joe Rogowski's tutorial on this site, "Making a Fast Pulse Induction Mono Coil" is a great help in coil design. At the moment I am plannning on building both 6" and 8" 'fast' spidercoils for the purpose of small nugget detection.

Toward that end I am considering changing the IRF9640 mosfets out to much lower (Coss) capacitance units in order to keep most of the speed inherent in the spidercoil design. I am looking at using the Fairchild FQD5P20 mosfet as it has only 75pf versus 370pf for the IRF9640. My concern is the device dissipation for the FQU5P20 is at best about 1/3 that of the IRF9640 so heat sinking will definitely be included. Even so I think the mosfet can handle pulse currents up to 14 amps dependent on duty cycle and repetition rate.

I appreciate any assistance you and the other members can provide in this effort.

Regards,

Dan

[silverdog]

Chance PI

Chance author andy_f has many details on his site http://fandy.vov.ru/Chance.htm

[baum7154]

Thanks for the info. Now I'll find a russian translation on the web and really make some progress. It does show 3 jumpers on one of the drawings on Andy's site.

[mario]

Silverdog, I received the Chance PI board and parts from you and the board looks great! I am wondering if anyone on this forum can point me to a site that has construction information and details. It appears the there are 3 wire jumpers to be soldered onto the board near the microprocessor. Can anyone confirm this?

Also does anyone have coil build information for this PI, i.e. coil inductance and resistance specifications? Joe Rogowski's tutorial on this site, "Making a Fast Pulse Induction Mono Coil" is a great help in coil design. At the moment I am plannning on building both 6" and 8" 'fast' spidercoils for the purpose of small nugget detection.

Toward that end I am considering changing the IRF9640 mosfets out to much lower (Coss) capacitance units in order to keep most of the speed inherent in the spidercoil design. I am looking at using the Fairchild FQD5P20 mosfet as it has only 75pf versus 370pf for the IRF9640. My concern is the device dissipation for the FQU5P20 is at best about 1/3 that of the IRF9640 so heat sinking will definitely be included. Even so I think the mosfet can handle pulse currents up to 14 amps dependent on duty cycle and repetition rate.

I appreciate any assistance you and the other members can provide in this effort.

Regards,

Dan

Coil specs, connected to X1:

L = 400 uH , R = 1.7 Ohm.


Has any one built a working unit yet?

The board has an ISP header for programing the chip. How cool is that?!?

I also find it interesting that the atmega's DAC is relegated to battery functions, while offloading the more important processing to Microchip's MCP3201 instead.

Still, how is this design based off Minelab Pi tech? I see 1 channel for the signal: OC1A? Another channel (OC1B) for sound, which would explain the use of that odd crystal.

I personally like rotary encoders instead of switches, but Andy is really pin constrained with the atmega8. Then again, the next step up would be a 40 pin atmega, which would be overkill. A drop-in replacement would be the atmega328p, easily quadrupling his memory and interrupts with very little code modifications.

He's able to fit his program within the 8k restrictions and is awesome. He probably writes in assembly.


Interesting and all very cool....

[silverdog]

Chance PI

If you install google toolbar it will translate each page as you view it

The three jumpers should be linked, as should a jumper around C4, this links 2 ground planes underneath

good luck

Silverdog

[Tepco]

Has any one built a working unit yet?

All your gold are belong to us, but whit this thing, not much we will get. I have built this thing something over year ago, (on different PCB) and i was not " enlightened" whit this detector at all. It works, but...

First of all, sampling time is fixed in software, 17uS, and way too slow for any criteria. So you don't need to use exotic fast MOSFETS, drive circuit for them is slow anyway and if you modify it, no change in performance will appear. Detector uses Minelab multipulse technology, but how well it is implemented, whit monocoil, im not so shure. Coil output is alternative combination of "normal" 120uS PI pulse and "burst" of 20+ asymmetrical short pulses, whit total coil charging time being the same (MOSFETS short the coil between pulses), at repetition rate of abouth 200Hz. Discrimination is based on taking 5 subsequent samples, two after first (normal) pulse, first one starting at 17uS, another 20uS later, and 3 samples after "burst", now delayed, so totally 5 samples are taken, 10uS apart, starting from 17uS. This is latter used for calculated software guess of "discrimination", whit questionable effect. Coil have to be fast enough for this 17uS, faster coils will work too, slower ones will mess up everything, use original coil for reference. Main disadvantage of this relatively slow sampling is very poor sensitivity to small conductive targets (jewelry, small nudges etc), but for coin-relics hunting may be good, this is coin shooter basically. So, if it sees any gold, will be on correct VDI position, if not, sorry. And detector is not especially deep, normal analog PI will kill it, also audio response is very interesting (but may be not something you can use to it). Really good thing is that this "multipulse" trick seems to work very well on very bad ground, thing can probably find its target in pure ferrite, well, if it can see it at all. So far, most advanced DIY PI detector, simple and cheap, most of stuff done in software, build and test it anyway. Just we may have to wait for some other "incarnation" of similar concept.

[mario]

All your gold are belong to us, but whit this thing, not much we will get. I have built this thing something over year ago, (on different PCB) and i was not " enlightened" whit this detector at all. It works, but...

First of all, sampling time is fixed in software, 17uS, and way too slow for any criteria. So you don't need to use exotic fast MOSFETS, drive circuit for them is slow anyway and if you modify it, no change in performance will appear. Detector uses Minelab multipulse technology, but how well it is implemented, whit monocoil, im not so shure. Coil output is alternative combination of "normal" 120uS PI pulse and "burst" of 20+ asymmetrical short pulses, whit total coil charging time being the same (MOSFETS short the coil between pulses), at repetition rate of abouth 200Hz. Discrimination is based on taking 5 subsequent samples, two after first (normal) pulse, first one starting at 17uS, another 20uS later, and 3 samples after "burst", now delayed, so totally 5 samples are taken, 10uS apart, starting from 17uS. This is latter used for calculated software guess of "discrimination", whit questionable effect. Coil have to be fast enough for this 17uS, faster coils will work too, slower ones will mess up everything, use original coil for reference. Main disadvantage of this relatively slow sampling is very poor sensitivity to small conductive targets (jewelry, small nudges etc), but for coin-relics hunting may be good, this is coin shooter basically. So, if it sees any gold, will be on correct VDI position, if not, sorry. And detector is not especially deep, normal analog PI will kill it, also audio response is very interesting (but may be not something you can use to it). Really good thing is that this "multipulse" trick seems to work very well on very bad ground, thing can probably find its target in pure ferrite, well, if it can see it at all. So far, most advanced DIY PI detector, simple and cheap, most of stuff done in software, build and test it anyway. Just we may have to wait for some other "incarnation" of similar concept.

I don't think Andy intended to design a gold unit. On the gold fields, using discrimination means passing up those really small or deep nuggets. Depending on the mass of gold, it can behave as a very low or high conductor, so discrimination is not a very useful feature. Mind you, I have not used the $4k+ Minelabs, so I don't know how reliably Minelab implements discrimination in those units.

Still, if true, what he has done is nothing short of amazing. With this unit, I see a chance at fighting those guys who litter our beaches with cheap pennys. One day, when vacationing in California, I remember seeing a guy hunting for jewelry. I found it odd that every once in a while he would drop a penny or two. It's not until afterward that I caught on.

Those pennys must have frustrated the PI guys hunting for deep jewelery.

It's the coin and relic hunters that considerably use discrimination. I think the nugget guys consider sensitivity much more important. I don't mind digging every target if it means occasionally finding a small nugget.

Also, even though this unit may not be very deep, you can always build a bigger coil. And If you want sensitivity, you may easily experiment with a faster crystal to increase sampling -albeit, you may lose LCD and/or discrimination features.

On a side note, have you taken a look at the timings of the gpx? How is the sampling handled? It's a matter of time before some one emulates the timings of a modern unit.

[Tepco]

Chance is completely useless as a beach machine, it can be actually good for any other type of search, except this one. In short, good for anything else except jewelry. Even Andy himself mentioned this. After all, give it a chance and try it.

[Altra]

I simulated the code, here are some screen shots of the timming. Sorry for the quality

I think because the sample timming is fixed. The coil must be well matched and be exactly the same as Andy's when he set the timmings in code. For decent results.

I hope somebody will post some scope pictures in the future with different targets

Regards Mark

[Tepco]

Very good simulation, excellent job, works exactly like this in real world, observed on scope. I had some pics of different w-forms, but it was while ago, i will post if i manage to find them.

[mollax]

in fact such
http://www.md4u.ru/forum/viewtopic.p...033&start=1350

[baum7154]

Chance PI is built and working?

I have completed the circuit board,switches, and an exact replica of Andy's coil using #24 AWG silver plated teflon wire on a basket weave HDPE plastic form (18 turns + 24 turns and 1.5 ohms due to the silver plated wire. It does detect and discriminate iron from gold nuggets in an air test on the bench.

I have some questions about operations and the buttons:

1. I assume that the BARRIER settings are essentially sensitivity thresholds...correct?

2. What are the purposes of the four MASK functions?

3. Is there a way to make the display backlight to stay on?

4. Is the ZERO button used after all other parameters are set?

5. How does Andy provide shielding for his coil? I am thinking of using polyetheylene anti static bag material for this purpose.

6. Has anyone else built this same coil and if so was the 1 watt 390 ohm resistor the best choice for it or was a different value better?

My unit draws between 100ma and 160 ma in operation and all the functions seem to work as expected.

I did test detection of a small vial of gold flakes with the largest being about the size of a match head and it seemed to detect it at about four inches in an air test. It also appeared to detect a long nose pliers at about three feet in an air test.

I appreciate your help.

Dan

[silverdog]

As far as I have worked out;

Barrier is sensitivity
Mask is discrimination - off,1,2,3,4 - 4 seems best for iron rejection
I tend to use zero as a reset if it becomes a bit noisey - which it does on my bench !
I haven't shielded my coil
390 works ok for given coil, not tested others yet
my lcd has no backlight, from what I have read light only comes on with each press, to save power I guess
install google toolbar with the translator and head here http://www.md4u.ru/forum/viewtopic.php?t=3033

78 pages of discussion on chance - andy_f the author is actively involved

[baum7154]

Silverdog,
Thanks for the response. Any idea on the 390 ohm resistor value. I haven't had a chance to put it on my scope to optimize it.

I did set the coil up on my slate pool table last night and the sensitivity has really gone down from that . I'm not sure why. It just about will not detect a gold sample when it is lying on that slate surface. In addition it will only detect a long nose pliers at about 8 - 10 inches on the slate. I'm going to take it outside my house today to get it away from other influences.

I'm also going to get it into a metal case ASAP so it will be better shielded.

Thanks again!

[Tepco]

I have completed the circuit board,switches, and an exact replica of Andy's coil using #24 AWG silver plated teflon wire on a basket weave HDPE plastic form (18 turns + 24 turns and 1.5 ohms due to the silver plated wire. It does detect and discriminate iron from gold nuggets in an air test on the bench.

I have some questions about operations and the buttons:

1. I assume that the BARRIER settings are essentially sensitivity thresholds...correct?

2. What are the purposes of the four MASK functions?

3. Is there a way to make the display backlight to stay on?

4. Is the ZERO button used after all other parameters are set?

5. How does Andy provide shielding for his coil? I am thinking of using polyetheylene anti static bag material for this purpose.

6. Has anyone else built this same coil and if so was the 1 watt 390 ohm resistor the best choice for it or was a different value better?

My unit draws between 100ma and 160 ma in operation and all the functions seem to work as expected.

I did test detection of a small vial of gold flakes with the largest being about the size of a match head and it seemed to detect it at about four inches in an air test. It also appeared to detect a long nose pliers at about three feet in an air test.

I appreciate your help.

Dan

Lovely, it is time to put together my Chance once again, i didn't even noticed new software versions (mine was built in 2009, when only oldest version was available, don't remember which one, and at the end i scrapped this project). So, first question is: did anyone tried to compare different software versions, and what is changed? (What is your version?) And few answers i can give:

1.Barrier is threshold control, keep it as low as you can, depending on interference etc.

2.In old software it was DISCRIMINATION 0-4 setting

3.Connect backlight directly (in one modified version of old software, it turns on when target is detected, battery saving or something, otherwise just at a press of the button)

4.ZERO detector on soil (and do this often)

5. Original coil is unshielded. Antistatic material is not nearly enough conductive for effective shielding. And if you shield this coil by other means (graphite etc.) this will slow it down and then it wont work at all. This is probably intended to be used whit unshielded cable, not coax.

6. 390R seems to be optimal value for original coil. (actually "must have" value).Coil is somewhat underdamped whit this value, don't attempt to speed it up, machine expects exactly that part of waveform when it takes sample, leave everything as close as possible to original specifications.

Current consumption is OK. Only fully assembled, battery powered and outdoors it will show full sensitivity, not bad considering coil is in essence 8''. What is very strange behavior of this detector is that it can see very small nudges, but some large (almost 30g) specific shaped jewelry can get completely undetected, in fact you can put it in a middle of searchcoil, ZERO and detector will work quite normally detecting everything else!!! Obviously some software crap whit "fast decay" targets or something, i hope fixed in newer releases. Fortunately, my PCB and coil is not completely cannibalized, i will give another chance to Chance and report results ASAP.