chance pi build

[green]

Green, do you recall the TC of the 6mm square of AL can sidewall? Was it the 3us or even shorter?

Two charts from another thread. The first chart was generated in Excel using recorded data TC's. 5 and 10usec TC lines plus ground lines with -1 and-1.49 slopes were added for comparison with linear and linear log charts. The second chart was done with Excel using the recorded data for 5x5 and 6x6mm squares cut from the side and bottom. The 6x6mm side TC is 1.3usec.

[baum7154]

Thanks Green,

1.3 TC for the 6mm x 6mm Al sidewall gives us a 6.5us Tau and it is not clear to me how it is seen @ 1cm distance with a 28us Sample Delay using the original configuration Chance coil. That Sample Delay is more than four times too long. I have to be misunderstanding what is going on.

[baum7154]

Dantech,

Can you post a screen shot of an O'scope trace showing the A/D sample delay relative to the coil flyback in the same shot?

Thanks

Dan

[dantech]

Baum7154
I shall try to make screen shots.
With 6x6 the signal is very weak and looks like ground signal but it is visible on dedicated coil.
What does the abbreviation "TC" mean?


Returning into signal masking...I have to explain how it works.
In the v121 there was something called "barrier". Let's imagine logarithmic table with values 3, 6, 13, 26, 54, 110, 225, 459, 936, 1911, 3900.
The table there is in the v121 and any newer release.
The signal has the the range 1-3900 and 3900 means "overdriving".
Each value from that table is a "step" of signal strenght indicator.
The purpose of barrier was masking up to 3.5 bar of signal strenght only because the barrier range was 1-20 only and like you see the value 20 there is in above table somewhere between 3-4th signal bar
My solution masks every signal bar one by one according to the table.
So, if you set signal mask on 3 bars it will work exactly the same way like barrier set to value 13.
In that case every signal from the range 1-3 bar will be muted (or value 1-13 with barrier)

There is no common sense to mask signal strenght by "units".
Too much button clicking and the level of precision is unusable.
The only reasonable way is to mask by bars.
Much quicker and simpler to use - only 10 clicks and you see what you set at once

Of course you can set mask on each signal bar if you need so we have real signal masking feature.

[baum7154]

dantech,

TC is the time constant for a given target. The mass, shape, and material that a target is composed of will give it a specific TC. One time constant is the time it takes a target to decay or give up about 63.2% of its energy stored as eddy currents. It takes 5 TC to deplete 99% of target energy.

In the case of the 6mm X 6mm sidewall of an aluminum soda can of about 0.1mm nominal thickness, forum member Green has measured the TC as 1.3us. 5 TC is then 6.5us and after this point in time there is no more detectable target energy. This is why it is critical to sample at or below 10us when detecting small gold perhaps in the range of 1/3 to 1/2 gram size. A true 28us Sample Delay opens the sample window 21.5us too late to see the 6mm X 6mm AL target decay.

Dan

[baum7154]

Baum7154
I shall try to make screen shots.
With 6x6 the signal is very weak and looks like ground signal but it is visible on dedicated coil.
What does the abbreviation "TC" mean?

Returning into signal masking...I have to explain how it works.
In the v121 there was something called "barrier". Let's imagine logarithmic table with values 3, 6, 13, 26, 54, 110, 225, 459, 936, 1911, 3900.
The table there is in the v121 and any newer release.
The signal has the the range 1-3900 and 3900 means "overdriving".
Each value from that table is a "step" of signal strenght indicator.
The purpose of barrier was masking up to 3.5 bar of signal strenght only because the barrier range was 1-20 only and like you see the value 20 there is in above table somewhere between 3-4th signal bar
My solution masks every signal bar one by one according to the table.
So, if you set signal mask on 3 bars it will work exactly the same way like barrier set to value 13.
In that case every signal from the range 1-3 bar will be muted (or value 1-13 with barrier)

There is no common sense to mask signal strenght by "units".
Too much button clicking and the level of precision is unusable.
The only reasonable way is to mask by bars.
Much quicker and simpler to use - only 10 clicks and you see what you set at once

Of course you can set mask on each signal bar if you need so we have real signal masking feature.

-----------------------------------------------------------------

Thanks Dantech,

A few questions,

1. Is it possible to have no mask at all?

2. Is the mask simply eliminating unwanted interference audio chirps? Or is it a mask to eliminate Iron responses?

Dan

[dantech]

1. No, you cannot set signal masking on 0 because 1st - 2nd bar step has value 3. This is unusable to play with units (bar 1-3) has too small difference. Remember that this is logarithmic scale
Look into table I have written earlier. I guarantee that internal noise of your hardware has higher level than 1-2 bars
2. Signal masking is simple sound muting without interfering in the process
3. Metal masking is simple sound muting related to 2 virtual points on LCD called VDI1 and VDI2. You can set points it in the menu and switch different variant of masking.
From left hand side of lcd to VDI1, between VDI1 and VDI2 and from VDI2 to right hand side of lcd. There are 6 possibilities of metal masking.

You have see it - to much to write

I have made a few test with 6x6 and it behaves rather like background - sometimes visible but for short period.
Visibility is rather unusable.
10x10 is fully visible from 5cm
A few screen shots from my O'scope below

[dantech]

The shots confirm my coil delay. Like I said it is 28us. It is point when signal strenght has the highest level with target in the coil. 27us has half sensitivity and 26us has total overdriving (audible 100Hz). So I find that 28us is the natural delay of my coil
Like we see on shots (5us time base) decay starts at 27us but the edge is little bit rounded. I feel it as half sensitivity.
With D4 removed decay has started 0,5us later only. I did not test it on O'scope but I did "natural delay" test as above.
Overdriving was on 27us and full sensitivity was on 28us so I find that the difference is 0,5us (settings resolution is 1us like we know)
So like I said it is not important modification for dedicated coil.

[baum7154]

dantech,

I agree with your description of 28us being the natural delay of your coil with its present damping resistor value. I have used this same process when using an adjustable damping network to quickly find the critical damping resistance for a new coil. I think you should be able to go lower than 28us by optimizing your damping resistor. It is worth trying. You can use a simpler adjustable network by simply adding a 1k pot is series with your current 470 ohm damping resistor. However once you have gotten to the lowest natural coil delay with the adjustable resistor network replace it right away with a fixed resistor or resistors because the wiper contact on the pot tends to get burnt with the flyback voltage.

Ideally the coil should be upgraded to a 300 to a 320uh 3DSS coil. I can also see that the preamp goes into saturation and that can be improved with the 2 stage amp mod. On my Chance PI this amp mod shortened the flyback signal elongation fed to the A/D by 9us. If you look at the flyback signal at the input to your preamp you can see it will be shorter than at the amp output due to the saturation.

Thanks,

Dan

[dantech]

a few screen shots from my Chance

[baum7154]

What did you se for a gold target? Was it a ring?

[dantech]

yes, the Au ring
Marked points on screen shot confirm settings in the menu.

[dantech]

http://chomikuj.pl/dantech/Chance+PI

v180

[dantech]

I did a few tests more with damping resistor on Chance. I have noticed that when I increase R13 value thr discrimination feature
works worse. Interesting is the fact that there is nothing bad visible on oscilloscope (eg. coil ringing)
My dedicated coil allows to use R13=1k without ringing. The natural delay shortens 2us, no ringing but discrimination works much worse.
I do not notice any sensitivity "in plus" so I find that original value of R13 is perfect in my case

[baum7154]

Dantech,

One thing I noticed when I installed the 2 stage amp mod in my Chance PI is that the amp offset trimmer gave the ability to set the input level to the A/D and this has a beneficial effect on the sensitivity and the discrimination. The original design does not have the trimmer capability to set A/D input level.

Dan