Chance Pi problem....¿¿¿????

[Tepco]

MOSFETS are probably good, some small leakage is normal, in case of failure current consumption will be abnormal, whit some smoke plumes occasionally. They usually fail catastrophically, but sometime leakage problem may appear, change it, just in case. Sound card based scopes are quite limited (freq. response etc), not exactly suited for PI stuff.

For chip programming, all what is needed is a piece of rocket science, like this:
After all, this is electronics, not fine art, it works!

[Hannu]

I just order new parts (Q1 ... Q3), but it will take about 1 week when they are here.
Anyway here is one more measurement. Coil is connected and D6 is de-soldered. PSU is car battery (12.7V / digital multimeter) . It's 13.1V on LCD screen - I think this is not so accurate value.
http://aijaa.com/v.php?i=007918671138.jpg

[Tepco]

This type of measurement can be highly inaccurate, due to intrinsic difference between different types of DMM, how it will integrate pulsed waveforms, RMS or average, freq. response etc. I repeated measurements using average hand-held 3.5 digit DMM under same conditions, replaced Q2 from NMOS to bipolar, and obtained following results, in very good agreement whit yours: (Just note my detector is not from kit, built long time ago on single sided PCB, uses European BC transistors instead 2SC types and may be some minor differences, but not so important). Consumption was 68mA,D6 removed, ignore display voltage monitor value.

Point-your value-my value:

T0- 255mV/253mV (good starting point to compare DMMs, very close)
Base Q2 42mV/43.5mV
C Q2 11.5V/11.46V
E Q1 and G Q3 11.3V/11.13V
D Q3 27.3V/29.9V
Coil: 82mV/48.8mV

Only latest value differs slightly. Chance that you have leaky MOSFET is 1 in 1000000, but if you are lucky enough...Try to interchange Q3 and Q9 on board and measure again.

[solmer2]

Thx

Hi !, thanks everybody, I've been talking with Silver Dog and he told me that another buyers have the same problem so I bet the error is not in the coil or battery but in the scheme.... He told me that I'll keep me informed about the progress of the Chance, Silver Dog is working to find a solution.....

Anyway thanks for the imputs... I'll post the solution asap.

[silverdog]

I am a bit confused and admit a bit out of my depth, the pcb and verson 0.84 worked ok.

As soon as I attempted the hardware changes for V1, I have been unable to get a working solution - the hardware changes are pretty simple so I am fairly sure they are correct.

The timings have changes a lot in V1, so I was wondering if my coil was not close enough to spec. to work with V1, but the continuous sound appears without a coil connected.

I will go back to 0.84 without hardware changes and work my way thru the changes to see what I find.

I'll update as soon as I have any news

Silverdog

[mschmahl]

Chance problems

Hello,

i have a Chance-Kit from Silverdog. The new one. It was with Version 1.1.1. This was working but the searchdeep was not so good. Than i have orderd a new chip from Silverdog with Version 1.2.1. After Adjustment Guard interval (down to 25, my coil doesn´t work with deeper Guard interval) it works very good. It reacts extremly to iron in the ground (5 cm long nail was 50 cm in the ground) My Coil have 1, 5 Ohm and 0,525 mH.

A problem was the endingwires of the coil. If i have it in the center of the coil, the md gives a permanent signal. so i have lay it to one side. This is better.
So i think, the coil reacts of the coilwire endings.

Sorry for my bad english.

[silverdog]

Hello,

i have a Chance-Kit from Silverdog. The new one. It was with Version 1.1.1. This was working but the searchdeep was not so good. Than i have orderd a new chip from Silverdog with Version 1.2.1. After Adjustment Guard interval (down to 25, my coil doesn´t work with deeper Guard interval) it works very good. It reacts extremly to iron in the ground (5 cm long nail was 50 cm in the ground) My Coil have 1, 5 Ohm and 0,525 mH.

A problem was the endingwires of the coil. If i have it in the center of the coil, the md gives a permanent signal. so i have lay it to one side. This is better.
So i think, the coil reacts of the coilwire endings.

Sorry for my bad english.

Hi mschmahl

its good news that yours is working, so pcb and components are ok - I wonder why a few of us have problems ?
Could you post pictures of your completed pcb as well ? it may help ?

Silverdog

[mschmahl]

Completed PCB

here are the pictures. A few changes i made:

1. I have many sockets for my part´s
2. I add a power-switch
3. L1 smokes once by experiments, but it should be ok

[Tepco]

Word of caution: Using "resistor like" chokes in L1 position is not a good idea at all. They are not suitable for required current, have too high ohmic resistance and will not provide adequate filtering. Instead use ferrite or ring core inductors, like ones intended for MC33063 or LM2575 regulator chips (cheap and available), or something cannibalized from old car phone charger, PSU or similar, no need to be exactly 220uH. Or rather short it, this will make less trouble.

[mschmahl]

thank´s i will try this.
Is it possible to make a better response of silver? the response of iron is extrem but we have a lot of iron trash in the ground.

[mschmahl]

What is the best?

What is the best for L1? I don´t know how many Ohm and how many Ampere.

Thanks a lot

[Tepco]

thank´s i will try this.
Is it possible to make a better response of silver? the response of iron is extrem but we have a lot of iron trash in the ground.

It is possible in this software version, at least to some degree. Old versions of Chance had fixed, long sampling time (17uS), not suitable for small conductive targets. In this version "guard interval", that is, minimum delay for at least some samples (detector uses 28 in this version) can be adjusted (10-99, 30 is default for original coil, not in uS), the best and most useful feature of new software. Using faster coil whit proper dumping R (R13), preferably flat spiral design, and reducing guard time to minimum will do the trick. If you go below minimum time coil can provide, detector will just "die out" and stop responding to metal completely. Fast enough coil and minimal guard time is what you need, no more gold (or silver) "blindness" and still good behavior in bad ground. Same way, detector now can be adapted to larger coils or frames. Original coil is too heavy, complicated to build, not so fast, it is only 8", probably the weakest point in this design. Playing whit different coils can reduce sensitivity, change ground behavior, shift VDI/ Iron mask position (don't count on it too much anyway), but at least now it is possible.

[Tepco]

What is the best for L1? I don´t know how many Ohm and how many Ampere.

Thanks a lot

As a general rule, select one whit minimum ohmic resistance and highest current rating. Usually, these "small signal" inductors are rated based on permissible dissipation due to ohmic losses. So, small cross section of ferrite will be in saturation (reducing inductance) long before rated current is reached. Inductors intended for SMPS use have different specifications.

[WM6]

Fast enough coil and minimal guard time is what you need, no more gold (or silver) "blindness" and still good behavior in bad ground.

.

That is what we are searched for, Tepco.
What else to do in circuit, to improve sesitivity to natural gold (small gold nuggets).
Thanks in advance for all suggestion.

[satdaveuk]

It just means the author of the project is refining his work, which is great. I would rather follow a FREE evolving project than a dead one because there's always room for improvement.

I agree with your comment about providing a programmer. Some can be found cheap, $12 at ebay. I wouldn't recommend the cheap usBASP clones though.

Ditto
This forum must stay the way it is, after all thats what draws us here like a magnet attracting metal, A good example of that was and still is the tgsl which avconics started off which has gone from head to foot then onsite working as designed In education and development, contributed by some very clever engineers who dont ask for anything in return apart from new Ideas from other like minded people looking from a different angle.
Ive contributed a small mount here in the past but far as designing and alot of the theory behind such a difficult subject to actually make and operate in the real world is a different ball game to the radio and tv trade I was brought up in.
It was like a jungle when I first joined here, but im learning fast because I find it interesting to build metal detectors and using them which Ive done for the last forty years.
So please keep things as they are, anyone wanting to contribute in new Ideas/designes/schematics etc or maybe see a glitch in someone elses project, bring it on, thats what Geotech is all about, if need to alter design even 20,000 times so be it, that shows to me and many other like minded people that Geotech wont take second best they want perfection, and for me thats the way it should be.
The brains behind this forum are priceless
If anyone wants to build a final project thats going to work to there needs then just wait and let the forums thrash out the mods glitches shortfalls problems etc, but just bare in mind you maybe waiting rather along time because in Geotech theres often someone at some point comes along with a improvement or maybe not such a improvement to your requirements, and it will go on, but as I said previous thats what this forum is all about
My thoughts on the matter hope it helps
Regards
Dave