THREE COINS IN THE P.I. FOUNTAIN

[Ferric Toes]

I loaded the GB code in and did a rough GB. With one 50c the output still goes positive, but when I add a second coin the output of the diff int goes negative past 0 volt common, so the target is being picked up more in the GB channel.

Somewhere I have some recent Oz coins. When I find them, and if I have a 50c, I will do a plot.

Eric.

[mickstv]

I loaded the GB code in and did a rough GB. With one 50c the output still goes positive, but when I add a second coin the output of the diff int goes negative past 0 volt common, so the target is being picked up more in the GB channel.

[mickstv]

Hi Mick,

I reconnected the diff-int and modded the Micro code to put in a late single sample, but the end result was the same as previously posted.

Later I'll reload the ground balance code and see what happens to the target response.

[Mechanic]

HI Mick, try it with a second very late sample and see if the results are different, though I spose they will be the same.

Cheers Mick

[mickstv]

Hi Mick,

Yes 0v for no target (TX coil common rail). I didn't have a second sample running, I also disconneced the other side of the diff-int and connected the pin 3 input to the TX coil common rail.


Cheers
Mick

[Mechanic]

HI Mick,?

For those results, if there was no target was the voltage measured as 0v?
And did you have a late 10us sample also, well out from where the target signal has decayed away?

Cheers Mick

[mickstv]

No I've only got an old analog CRO, ATM.

I've just monitored the same targets but now looking at the DC out of the main diff integrator with the Ground Balance circuit disconnected results below.

Sample is 10us in width and 10us from TX end.


No spacer...

1x 50c = 180mv
2x 50c = 176mv
3x 50c = 185mv


A4 spacer...

1x 50c = 180mv
2x 50c = 178mv
3x 50c = 192mv


5mm spacer...

1x 50c = 180mv
2x 50c = 252mv
3x 50c = 315mv

[Tec]

So is this what is happening ?

If this isn't happening, then I am lost.
Amplitude and slop(e).png

I notice your scope, does it have connection to PC ?.. I bought a DIGTECH 40MHz USB scope from Jaycar and am very happy with it.

[mickstv]

Here's some screen shots taken using 200us TX pulse and frontend gain of 100x, the targets are 50c AU coins. CRO set to 10us per div.

No Target.....





50c AU coin.




Two 50c AU coins no spacer.




Two 50c AU coins using a piece of A4 paper as a spacer.




Two 50c Au coins with a 5mm spacing.

[Tec]

Great Plots and data, thanks..

Are we still wondering what is happening here ?

Yes the three coins insulated should be the same as three coins not insulated ?
Yes, the three coins will decay slower than a single coin..

Insulated or not the are connected magnetically.

Move them a couple of mm apart and the decay curve time will decrease ?


Kick three fly wheels glued together, or kick one fly wheel that weighs as much as all three..

It will spin for the same amount of time ?

The big fly-wheel will spin for longer.
The small fly-wheel will spin for less time.

[Carl-NC]

At the frequency components we are dealing with in PI, a coin does have at least two, and possibly three eddy current modes.

I suspect #1 is from edge effects due to the small diameter of the coin relative to the coil. If you use a target much larger than the coil then this may vanish. #3 is due to the vertical eddies in a non-normal target, which then suffer the same edge effects as #1 but due to the thickness, not the diameter.

[Davor]

Granted

[Ferric Toes]

I could send you some Croatian coins if you wish. On a CW machine with two channel discrimination these sound as a machine gun, as their ID is strongly related to orientation. If they are laying flat they sound more like their metal plating, but if they are perpendicular they sound as iron. As moving your coil shifts the relative orientation, these have sharp transitions from Fe to Cu sound, hence the machine gun sound. People tend to lose several at once and you just can't misinterpret such groups: junk coins.

Look for a PM from me.

Eric.

[Tepco]

I don't think Croatian coins are plated, just alloy composition. Similar problem here, in neighboring country, even recent coins in circulation are made of at least 3 alloy types, maybe more, for same coin. Some attracted to magnet, some don't, different TC, on VLF machine VDI is everywhere. No Star Trek made VDI can go thru this. In just one run last summer, I collected 67 of them in just 30m in one resort, total value ~2-3$, and give up. Better to have smelter installed in MD, so far I could have my statue cast in natural size made of this alloy. Not only PI discrimination, but also VDI or multifrequency VLF wont help with this. On the beach, try to follow “trail” of first 2-3m from waterline, coins are usually grouped in trails, pay attention if just one response is different (from 15 usual VDI readings you have to remember), and, wow, you will get equally worthless foreign coin. Good old bottlecaps, I learned to ignore them with ordinary Surf, but this is a plague and no cure for it.



It is much better to use concentric, not DD coil for this, to avoid coil “phasing”, less susceptible to coin orientation.

[Davor]

Don't anyone ask me to do plated steel coins.

I could send you some Croatian coins if you wish. On a CW machine with two channel discrimination these sound as a machine gun, as their ID is strongly related to orientation. If they are laying flat they sound more like their metal plating, but if they are perpendicular they sound as iron. As moving your coil shifts the relative orientation, these have sharp transitions from Fe to Cu sound, hence the machine gun sound. People tend to lose several at once and you just can't misinterpret such groups: junk coins.