chance pi build

[Waikiki_Sweep]

Issue with resized PCB happened to me more than once, print from .pdf, or wrong paper format, letter instead of A4 or... Always noticed when PCB is etched, tin plated and drilled.

Ohhh. Yes. Next time I will print it on paper than put Atmega on it and check if pins fit.

One more Chance PCB - sandwich of two pieces: http://c2.at.ua/3m/L/Metal/Metd-4-.jpg
Link with layout file: http://c2.at.ua/load/avr/impulsnyj_m...zhki/17-1-0-62

[baum7154]

As I can see from picture it is Chance layout from DesAlex. It is smaller, one side, easy to make PCB. I have etched one.
Only my problem was that lay file was printed a couple % smaller so 28 pin controller hard to fit in it.
Probably Sprint or my laser printer were wrong. Better check dimensions before etching.

You can download all files including "Chance layout from DesAlex.lay"
http://forum.cxem.net/index.php?app=...ttach_id=75463

Comments on Russian on forum:
http://forum.cxem.net/index.php?show...dpost&p=510118

[ATTACH]28375[/ATTACH]

[ATTACH]28374[/ATTACH]

[ATTACH]28369[/ATTACH]

[ATTACH]28370[/ATTACH]

[ATTACH]28371[/ATTACH]

[ATTACH]28372[/ATTACH]

[ATTACH]28373[/ATTACH]

Makes for some interesting reading. Will take the time to do all of it!

Dan

[Waikiki_Sweep]

Just build a Chance PI.



Software version: 1.2.1.

First impression:

Device looks interesting. I have tried it with pretty lame coil - just a AWG 24 vinyl insulated wire.
But even with that lame coil sensitivity is good. Also I can clearly hear difference by sound between USA 25c, 10c, 1c, 5c.
Gold and platinum rings all falls close to USA 5c range with steady middle tone. Big chunks of iron have a low tone.

Disappointing with small iron pieces: hair pins, paper clips, beer caps. They plays like music box changing tone widely so you have no idea what is it.

I have tried to replace that lame coil with some better shielded and connected by coax coils but they work even worse.

Sure I have to make a proper coil as described by author and I will try.
But something tells me if I shield that original coil or use coax it will work bad.

So now I need a time to learn what I can squeeze from Chance PI because my goal is make it waterproof and search in ocean. (If possible.)
Because I expect salt water will force me to shield coil anyway and to use coax or that conducting environment will act like a shield that changing properties with every wave and trigger falsing for sure.

I was dreaming for a target ID detector for salt water for a long time. I remember I have purchased Garrett AT PRO for that purpose but sold it because simple VLF is useless in salt water.

When building Chance I had two major problems. First was absence of ground layer on my PCB so I have added a lot of jumpers.


Also couple times I have replaced transistors damaging PCB a lot because of pin-out of substitutes. Finally I got original A733 and 2SC945 from eBay.

Second major problem was with proper fuses. Different programmers coding fuses differently so "checked" bit may be interpreted as "0" or as "1" depending on programmer.
From first try I get it wrong. So fuses were made inverted in comparison with PonyProg screen-shot from website. Sure Atmega8 were "bricked" immediately.
Fortunately I have found mistake and cured it. But to "unbrick" micro controller I have build an external oscillator with quartz resonator and few logic elements.
Than I have connected oscillator to XTAL1 (pin 9) of Atmega8. Once micro controller starts I have fixed fuses. Now it works well.

I know author do not like to write instructions so some settings in menu may need explanations.
So I have tried to write my understanding of menu for 1.2.1 version. May be I am wrong. Please correct me.

Barrier ---- ??? probably it is sensitivity.
Volume --- sound volume.
Zero -------- set a zero for proper ground balance and initial conditions for search. Click it from time to time to fresh results.

Settings
- Iron mask ---- some complicated set of exclusions. For which targets to turn sound off. Anyway you can see all targets on indication.
- Volume raise ---- I am not sure. May be speed of reaction on target like a "SAT" in Surf PI.
- U min of batt --- when start to signal that battery is low.
- Light ----------- to keep light ON / OFF.
- Guard interval --- Delay in microsecond before sampling. (Delay pot analogue from Surf PI) Have to be set depending on coil quality or ground properties.

If I have learn something interesting about Chance PI I will post it here.

[baum7154]

Congratulations on getting your CHANCE built and working! I look forward to your progress.

1. The VOLUME RAISE setting raises the volume when a target is detected. I set mine to a value of '16' and when a target is presented the volume goes up noticeably with signal strength increase.

2. I believe that BARRIER is meant to set a signal strength threshold that must be met before audio output is allowed. When set to '5' most of the signals below that are not heard, but some are. I think this is because the VDI signal strength displayed is averaged or lags the actual value a bit. The instantaneous signal may be greater than the VDI shows and it is sometimes enough to exceed the Barrier setting and so audio output is allowed at that moment. Generally it works pretty well.

3. Guard Interval is definitely the delay after the last long pulse to the coil when flyback begins. This delay setting is limited on the low end (setting of 10) to 8us and on the high end (setting of 99) to about 88us. The steps are continuous in 0.9us increments.

What is the resistance and inductance of your coil? If you can find RG62B/U coax it is only 13.5pf/foot.

Good luck on the coil!

Regards,

Dan

[Waikiki_Sweep]

That "lame" coil on photo that I have used to test Chance has a parameters:

8.5" diameter

25 turns of AWG 24 (0.5mm) solid vinyl insulated wire.

325 uH inductance.

1 Ohm resistance.

390 Ohm dumping resistor was used for that coil.

Looks like it is pretty close to required parameters so Chance works relatively good with is.

Sure it has more interwire capacitance than reqiured by authors design but vinyl insulation provides some spacing reducing that capacitance a little.

Probably with proper dumping resistor and correct set of Guard interval Chance can work with different coils too.

I will try authors design coil soon.

Thanks for advice, RG62B/U I must have.

Also I can see that Chance can work even if case with PCB is close to coil. So for example for "diving" version cable can be very short.

I remember that Surf PI was falsing and oscillating if PCB was close to coil.

Why not to get that chance to differentiate by sound tone between concrete steel rods and stainless steel Rolex-es underwater?

[baum7154]

Yes, you really need to dump that 390 ohm damping resistor as it is nowhere near the proper value for any of the coils I have tried. Try to get the Guard Interval as low as possible with your coil and damping resistor. You can always set the GI (delay) higher for use in saltwater.

Best of luck!

Dan

[Silver Dollar]

I recently bought a chance board and wonder if there's a layer missing? It has the back side and just 3 jumpers on the front side.
Is is a 3 layer board?




Also were subs for the transistors ever determined? The 2SC945 and A733 are rare birds around here...

I'm sure this was discussed on the Russian forums but they are tough to read through the translators.
My hat goes off to folks doing the same with this site!

I had a look at disassembling the Chance code and it's doable but it runs 60 pages! It might take a while...
I wonder if we can convince Andy to release the code after all this time. (Mainly to allow substitute parts).

[baum7154]

It looks like you have one of the Silverdog boards. Will check mine against your photos.

Dan

[baum7154]

Yes, your board is a first release SILVERDOG product. It will require the modifications necessary to run the latest Version 1.2.1 firmware. No big deal to do the mods and this is what I am currently using for all my testing reported here. The mods are detailed on Andy's website. This is not a three layer board.

It would be great to open up the code and make a few tweaks and I have a few suggestions along those lines if it ever becomes possible. 60 pages of code does not surprise me considering all the features and signal complexities of this detector. Seeing the code would be instructive, especially in the area of how discrimination is implemented, and there a lot of people here to assist in the task.

Regards,

Dan

[Silver Dollar]

Thanks for your help! The mods for the new software aren't a problem. I was refering to #35 in this thread (page 2) where
someone said

"I have started to solder parts on Chance PCB and have noticed on that PCB just missed whole ground layer on the top.
So I will have to make several wire jumpers to ground all components properly. Probably it was one of that epic failure
PCB's from the first Chance with mistakes."

The PDF from Silverdog shows 3 layers;







I just checked and sure enough I have one of the original boards with the ground plane (2.jpg) missing.
It would have been nice to have a corrected one but I can do something like this to fix it;



I can see where the screen is written to and places where he talks to the external A/D but
to find the disc code will take a bit of study.

[sinclairuser]

@silver dollar, c915 and c945's appear in crt televisions here in the uk, can't imagine it would be that different where you are, at least its an avenue to investigate, ive found in the past with radio projects that scrounged/secondhand parts must be used, as so called replacements just did not work well, or other parts needed to accomadate the alt part were even rarer than originals.
i guess we are all different but i would rather use the exact part, even if that means tested secondhand.
as for a733 ant help there.

[baum7154]

I did not consider the three top side jumpers as a layer but it looks like that is what is depicted in the jpg. Sorry for the confusion. I believe I got my 2SC945 and A733 transistors from SILVERDOG, have you asked him if he has any?

Dan

[baum7154]

I did not consider the three top side jumpers as a layer but it looks like that is what is depicted in the jpg. Sorry for the confusion. I believe I got my 2SC945 and A733 transistors from SILVERDOG, have you asked him if he has any?

Dan

After carefully looking at my Silverdog Release 1 and 2 boards, the jumper positions shown in your 3.jpg pic are actually wires to be installed by the builder and are not printed as a board trace. Does your board actually have the traces visible or if not can they be ohmed out? If they don't ohm out you'll have to put the wire jumpers in. As I recall this was recognized a while back as the reason some constructions did not initially work. I know it is pretty confusing when the documentation shows another layer.

My Version 2 board ohms out just fine on the ground planes. Considering both versions are through hole plated boards I'm not sure there is a ground plane problem in either the Version 1 or 2 boards that I have.

Regards,

Dan

[Waikiki_Sweep]

I recently bought a chance board and wonder if there's a layer missing? It has the back side and just 3 jumpers on the front side.
Is is a 3 layer board?

[ATTACH]28721[/ATTACH]
[ATTACH]28722[/ATTACH]

Also were subs for the transistors ever determined? The 2SC945 and A733 are rare birds around here...

I'm sure this was discussed on the Russian forums but they are tough to read through the translators.
My hat goes off to folks doing the same with this site!

I had a look at disassembling the Chance code and it's doable but it runs 60 pages! It might take a while...
I wonder if we can convince Andy to release the code after all this time. (Mainly to allow substitute parts).

Check eBay. You can see 90 results for 2SC945, 73 results for A733.

I got my 2SC945 from eBay US seller and it was fast.
A733 I have purchased on Digikey but you can see a lot offers on eBay too.

It is good to use original transistors because substitutes have many different pinouts and you will have a headache to insert them correctly. I have made a lot of damage to PCB replacing transistors back and forth several times.
Also do not buy KP505 transistor because for new versions it has been replaced with 2SC945.

Good luck in soldering absent ground layer. I did it for my Chance. Pay attention for missed ground layer near pot R24 or you will see nothing on display (as it was with me).
Also I think that big solid jumpers I put everywhere to fix ground are good to lower resistance of ground connections and make Chance to work more stable.


Properly assembled Chance works perfect.

[Silver Dollar]

I see the A733 on ebay but they also call it 2SA733 which is a different animal (much lower HFE).
I ordered most parts from Mouser but they didn't have the transistors. I bought a 2N7000 before
I checked out the mods so too late!

The 2N3904 looks pretty close to A733 so might work ok. I see some venders selling 2SC945's and
2SA733's bundled so must be for Chance builders... I'm not sure the provided devices will meet
original specs coming from the far east. I've noticed the asian vendors don't provide much electrical
data with anything...

Here's a web site with all transistor data on it, to help in finding a good match. http://alltransistors.com/spisok.php?n=95

I'll use your picture to add the ground in. Should work OK just look a little ugly. I'm getting excited to
have a good working PI machine when the weather finally turns good again!