Optimizing PI Coils

[Mr.Jaick]

green you need to buy "scale for reloading ammo" which goes down to 0.001g and beyond

[green]

Would be interesting to know why #9 lead shot doesn't test closer to the same amplitude. But maybe doesn't matter, no one wants to find #9 shot. I was using it because someone suggested using lead shot as a test target. Did another test today (test B) to check test repeatability with the shot test A (blind test, covered results from test A so I couldn't cheat) and some wire targets (possible target replacement for #9 lead shot). Wire targets lot closer to same amplitude. #9 lead shot time constant between AWG24 and AWG26 time constant. Any thoughts on using copper wire for test targets?

lead shot number 5, 8, 9, 14, 15 and 17 tested the same for test A and B

[Carl-NC]

I'm surprised at the variation in lead shot. I haven't seen that but I also haven't tested a lot of pieces.
I like lead shot because it's commonly available everywhere and should be pretty consistent. It's also spherical and comes in a variety of standard sizes. #9 is roughly equivalent to a 0.5 grain nugget and represents about the smallest piece a Minelab will pick up. Cut wire will also work but because it's not spherical detection will depend on orientation.

[green]

Tried some #8 lead shot to compare with the #9 lead shot.

#8 lead shot should weigh about 69mg, #9 lead shot about 48mg. How accurate would lead shot weight measurement have to be to see if weight difference is similar to signal difference?

[Carl-NC]

I have a scale with 0.001g resolution so I weighed 15 pieces of #9 shot (in grains):

0.66gr
0.70
0.70
0.70
0.74
0.76
0.76
0.78
0.78
0.78
0.80
0.82
0.82
0.90
0.92

I then took the lowest and highest, put them on wands, and tested them on a GoldBug-2 (only detector on hand that goes this small). The 0.66gr shot breaks the threshold at 1" while the 0.92gr shot is maybe 1.25". When I get a chance I'll get the SDC-2300 out and try that.

[green]

Connected 133mm fig8 Rx coil (5.6inch mean diameter) to test for distance with #9 lead shot. To get repeatability, I used some foam spacers (1inch squares, 5mm thick). Added squares at Rx center until threshold wasn't exceeded almost every time. Shot was taped to a craft stick for a wand, shot placed on stacked squares and removed quickly. List distance threshold was exceeded almost every time. Tested #9 shot from reply #58, numbers 4, 7 and 14. #4(83mV, 45mm), #7(43mV, 25mm) and #14(19mV. touching).

Don't know how much the weight varies. Wondering if distance is affected more with a PI.

[green]

Connected 133mm fig8 Rx coil (5.6inch mean diameter) to test for distance with #9 lead shot. To get repeatability, I used some foam spacers (1inch squares, 5mm thick). Added squares at Rx center until threshold wasn't exceeded almost every time. Shot was taped to a craft stick for a wand, shot placed on stacked squares and removed quickly. List distance threshold was exceeded almost every time. Tested #9 shot from reply #58, numbers 4, 7 and 14. #4(83mV, 45mm), #7(43mV, 25mm) and #14(19mV. touching).

Don't know how much the weight varies. Wondering if distance is affected more with a PI.

Wondering why Carl got little difference in detection distance with .66gr and .92gr #9 lead shot. Some possible reasons. My shot varies weight a lot more than his, GoldBug-2 isn't affected as much as my PI circuit, I'm doing something wrong, probably others. Trying to learn something, any thoughts?

[EL NINO]

In my opinion ... shot N.9 on for the PI detector already has the so-called critical size / weight ratio .... which means that it is close to the sensitivity limit of this PI detector .... and VLF detector 71 khz Fisher Gold Bug 2 is this the shot N.9 is still relatively easy to detect ...


I owned a Fisher Gold-bug2 for a while, so I had time to test this detector on various large targets..also on micro targets ... up to 0.003 grams which already ../detection range Fisher GB2 71Khz is... 3.5cm in All metall ..vs 1.7 cm in Disc - for 0.003 gram target.. and 3.5x6.5" coil/... in "high Treshold mode" at 5x10"and 3.5x6.5" coils ..

High-sensitivity VLF detectors operating at frequencies 18.6-30-40-48-54-80 khz have a good "sensitivity / range ratio" ....even for very small gold targets ...on coils of optimal size....../7-9cm detection range in Airtest .. for my 0.02 gram 24k gold test target../

for PI detector, the target can be 0.02gram 24k gold, a more demanding target ... as an example, I will present PI Garrett ATX in tests for sensitivity and range in mode without Ground balance.


But I'm also curious about Carl's answer .., and his possible test on PI Minelab SDC 2300 ..

[Auto-Mation-Assist]

When I was building a lot of high end stereo equipment in the past I reduced cable capacitance from tape heads or turntable phono cartridges not only by reducing cable length but also using two coax cables. One cable for each side of the devices coil which reduces the capacitive loading and thus provided extended the high frequency response and resonant frequency. When such a method is used with metal detector coils it will likely provide a increase in speed if the termination resistor value is properly recalculated for this configuration. High quality cable with low capacitance per foot is required. The drawback is that it increases cable mass since two cables are required per coil which can be held together with heat shrink or other suitable wrapping.

An additional braided shield can be added around the two coax cables before adding the heat shrink covering too provide a ground for the shield in the coil while the internal two coax shield are connected together only at the far end at the RX circuitry and grounded there.

[bbsailor]

When I was building a lot of high end stereo equipment in the past I reduced cable capacitance from tape heads or turntable phono cartridges not only by reducing cable length but also using two coax cables. One cable for each side of the devices coil which reduces the capacitive loading and thus provided extended the high frequency response and resonant frequency. When such a method is used with metal detector coils it will likely provide a increase in speed if the termination resistor value is properly recalculated for this configuration. High quality cable with low capacitance per foot is required. The drawback is that it increases cable mass since two cables are required per coil which can be held together with heat shrink or other suitable wrapping.

An additional braided shield can be added around the two coax cables before adding the heat shrink covering too provide a ground for the shield in the coil while the internal two coax shield are connected together only at the far end at the RX circuitry and grounded there.

What you call the termination resistor is called the damping resistor to quickly damp the current oscillations that occur just after the TX pulse fully discharges. This damping resistor value allows the RX stage to be turned on more quickly with less delay. The amount of capacitance from all sources that the coil sees will ultimately define the damping resistor value. Coax capacitance is one of the high contributors of capacitance so reducing this is a good way to reduce coil seen capacitance and use a higher value of damping resistor.

Question- What high quality coax with low capacitance would you use? Please post a rough drawing showing how the two coax cables are used to connect the coil to the circuit board components.

One good way to test the benefits of using two coax cables compared to just one is to measure the coil self resonant frequency using both the single coax and the dual coax but at the circuit board attachment end of the final cable length. With a known coil inductance the self resonant frequency will allow you to calculate the capacitance as seen by the coil. A higher self resonant frequency will mean less capacitance, a higher damping resistor value and less delay.

Joseph J. Rogowski

[Auto-Mation-Assist]

I will work on a diagram for how I envision using the dual coax method. I will also do some research to see if there are cables already available with dual coax cables inside along with overall shield. Heat shrink tubing is not known for being tough.

Right now I'm not yet ready to wind my own coils since I wanted to develop my Analog PI detectors using coils that are commercially produced so that I have a standard to compare to. As such I have been using two Coiltek Elite 9 inch mono coils and also a Elite 14 inch coil. These have giving me an excellent performance base to work with for potential future DIY work and comparisons.

Picture of the control panels of my analog PI detectors and test samples used.

[Auto-Mation-Assist]

Attached is the drawing of my proposed dual coax use to reduce coil cable capacitance. I don't know if it it will provide enough benefits to justify the added complexity for this use until it is tried and tested.

[bbsailor]

Attached is the drawing of my proposed dual coax use to reduce coil cable capacitance. I don't know if it it will provide enough benefits to justify the added complexity for this use until it is tried and tested.


[ATTACH]57248[/ATTACH]

Does the detector connector shell ground connection, per your drawing, have any connection to any of the numbered pins on the connector?

Do you have access to an oscilloscope and a signal generator that can reach approximately 2 MHz? If so, you can see if the single coax or dual coax has any different resonant frequency.

Thanks for your quick answer to my question.

Joseph J. Rogowski

[Auto-Mation-Assist]

The shell ground connector does not connect to any pin at the connector itself.

Here is a picture of my internal wiring that connects the coil socket to the transmit and the receive circuit boards.





I have all test equipment required but not ready yet to work on constructing my own coils just yet. I have litz wire on hand and may try to do a flat coil using MACH4 CNC tangential mode on my small milling machine.

[Carl-NC]

Here is a picture of my internal wiring that connects the coil socket to the transmit and the receive circuit boards.

[ATTACH]57250[/ATTACH]

"Invalid Attachment specified."

Can you re-upload the pic?