gold nugget simulation?

[6666]

Does anyone know how to lower fly back volts without lowering current or increasing decay time?

I think Davor invented a snubber to do that.

[green]

I think Davor invented a snubber to do that.

Thanks, I've seen his snubber before but wasn't smart enough to know what it was for.

[baum7154]

Thanks Dan, For reference how far away does the Chance detect a US nickel with what size coil? Approximate coil current and inductance? Maybe someone with a MPP or any PI could answer the same questions.

-----------------------------
Hi Green,

Using my 8" round fast 335uh spider coil on the Chance detector in an air test the US Nickle had repeatable first detection at 10 5/8" and repeatable full tone lockup at 8 3/4". I don't really know the coil current as the Chance has a complex multi pulse output with a 120us long pulse at the end of a 16 short pulse string of incrementally lengthening pulses with active coil shorting between short pulses. A larger coil would definitely increase the detection distance of a nickel. Sampling was set at the minimum 8us.

I just remembered that Tinkerer and I had a discussion of Chance coil current and his assumption was a 300uh coil on Chance would have about 3 amps of current.

Hope this helps.
Regards,

Dan

[green]

Thanks Dan

[Chet]

I charted some data in the (Bid foot style for a PI) thread replies #61 and 63 about correcting for different coil inductance.

Hi Green,

My memory just doesn't work so well, I thought so much of your coil inductance correction method when you posted it in #61 and #63 I promptly stored it in two different files. I just didn't remember doing it until now.

Thank you for pointing me in the right direction,
Chet

[green]

Been looking at snubbers. Wondering if adding C1 sized to limit voltage would work or what is the best way?

[detectormods]

C1 will be in series with the diode intrinsic capacitance (C across the coil) and delay settling the FB voltage, C is the enemy R is your friend.

[crane]

The zener is chosen for the desired clamp voltage. It behaves as a normal diode during the spike and behaves as a zener when the FET turns on. The cap can be a 1uF or similar non-polarised. Nothing gets hot. You could also use a separate FET and a normal diode in parallel with the zener with a resistor to slow down the discharge but the method does work as shown.

Don't touch the terminals when charged!

[green]

The zener is chosen for the desired clamp voltage. It behaves as a normal diode during the spike and behaves as a zener when the FET turns on. The cap can be a 1uF or similar non-polarised. Nothing gets hot. You could also use a separate FET and a normal diode in parallel with the zener with a resistor to slow down the discharge but the method does work as shown.

Don't touch the terminals when charged!

Thanks for the replies, looks good. I needed to order some parts and wanted to try eliminating the fet avalanching.

[green]

The zener is chosen for the desired clamp voltage. It behaves as a normal diode during the spike and behaves as a zener when the FET turns on. The cap can be a 1uF or similar non-polarised. Nothing gets hot. You could also use a separate FET and a normal diode in parallel with the zener with a resistor to slow down the discharge but the method does work as shown.

Don't touch the terminals when charged!

The circuit looks good. The 100n capacitor seemed to be enough for 1 amp peak current. Tried another circuit. Both add a little to the delay time. Is there a reason for the snubber other than protecting the Mosfet? Is a snubber needed at lower peak currents?

[green]

The circuit looks good. The 100n capacitor seemed to be enough for 1 amp peak current. Tried another circuit. Both add a little to the delay time. Is there a reason for the snubber other than protecting the Mosfet? Is a snubber needed at lower peak currents?

Didn't notice any difference monitoring integrator out with the different circuits yesterday. I have been playing with the coil current in a control loop and noticed the circuit with the zener diodes eliminated a glitch in the current loop, so I'll be using it.

[green]

Made a new controller. Tried the sequence from reply 102. Looks like it works ok. Did a ground decay in excel. The times I get in excel using a -1.25 slope match what I get on the bench with ground that charted about -1.25. Excel predicts a 23usec TC hole. What would be best hole TC to try for?

[green]

Charted a gold chain. Recorded as shown in picture. Lower TC than I expected.

[green]

Detector circuit not working to test distance measurement, so I tried the 10grain(TC=6.6usec), 18grain nugget(TC=5.4usec) and a 20mm piece of AWG16(TC=5usec) solid copper wire from charts reply#15 with target response tester. Signal strength not as high as I would like with 200mm figure eight coil but think results aren't bad. Need to finish my detector to see how distance measurement compares. Interested in someone trying detection distance with a 20mm piece of AWG16 solid copper wire.

Added links from reply#15(Gold Results)
http://www.geotech1.com/forums/attac...9&d=1543333008
http://www.geotech1.com/forums/attac...0&d=1543333065

[green]

Nugget sizing info:

We are often asked how many pieces per gram or ounce. It is very hard to predict how many pieces there are per gram or ounce as the # of nuggets by weight varies quite a bit per batch. But in general you can expect around 1-2 pieces of gold for 4 mesh, 2-4 pieces of 6 mesh per gram, 7-12 pieces of 8 mesh per gram and around 15-20 per gram for 10 mesh. You can expect many more pieces for smaller 12 (around 20-25), 14-16 (around 30-50 or more) pieces and hundreds for fine gold. Every batch is very different and each piece of gold is natural and of course therefore unique. Some may be flat and light or rounded and very dense (heavy).

How small a nugget can a good PI detect? What mesh size would make good test targets for smaller nuggets? 8 mesh, 10 mesh, both or other?

Any guess on typical TC for 8 or 10 mesh nuggets?