Tweet
Greetings to all,
First post here. My name is Thomas. I am a n00b to both electronics and metal detection—my background is in organic chemistry. Anyways, I have a Clone PI-W, the SMD version with the row of LEDs (not the through hole / LCD). Doing some experiments at home winding smaller diameter coils out of 26 AWG FEP teflon (4.8cm former). I guess this would be called a button coil or super sniper coil. The instruction sheet that comes with the Clone PI gives a chart with coils of varying diameters, the smallest being 120cm, The recommended inductances are from 380 uH to 406 uH, with resistance varying from 1.6 to 2.3 ohm. I wound a coil of smaller inductance by accident when I ran out of wire, about 312 uH, 1.63 ohm, and this coil showed the best distance in detecting a 1 grain bar of gold through stacked test layers. I had various coils wound with magnet wire, standard silicone pick-up wire, etc., tried FEP teflon to reduce interwinding capacitance, and just by chance I wound this coil with the low inductance which outperforms all the other ones that I wound with 4.8 cm former trying to stick to the instruction sheet ranges for inductance and resistance. I measured capacitance of the coils with a NanoVNA, so I am factoring that value in, and based on all the data, the low inductance is what really seems to boost performance here.
So then I started gradually peeling off turns from another identical coil, coming down by ~25 uH at a time, and there seemed to be a steady improvement in performance all the way down to 200 uH (1.24 ohm), which was significantly better for distance than even the fortuitous 312 uH I started with, at which point I halted and started consulting AI about potential issues. I was advised to test the voltage spike across the drain and source of the MOSFET (IRF740 - 400 V rating), and was alarmed to measure a 470 V spike. The original 312 uH was testing around 372V, and a 400 uH coil, same wire and diameter, tested at 312 V. The thing was still working and still continues to work with that MOSFET but I’ve been nervous about leaving it on for too long with the best coil so far if it has that kind of spike. Maybe it didn’t avalanche but I may have been close.
So — any general suggestions for exploring down in this low inductance range?
AI suggested another MOSFET with a higher voltage rating: STP10NK60Z, 600V rating, I got a few, swapped it out, but it did not work at all. The start-up sequence was wonky without the usual beeps and there was no detector response to speak of (tried the 312 uH coil), with some very erratic beeping. I did not scope anything with it installed, so I don’t know why it doesn’t interface. The MOSFET passed all multimeter checks before and after. The most significant discrepancy from the datasheet between this and the original IRF740 MOSFET is with the output capacitance and reverse transfer capacitance: (IRF - 330 pF output, 120 pF reverse, STP - 156 pF output, 37 pF reverse)… AI seemed to think this could be an issue.
I installed the original IRF740 and the detector resumed completely normal behavior. This original MOSFET passed all multimeter checks. So the champion 200uH coil did not seem to kill anything outright.
The IRF840 has a 500V limit, so I will probably try that. This subbed for IRF740 in the PIKINI V4 according to one user. Any other general suggestions or advice for mods to manage higher voltage spikes?
Also, do people ever install a socket so they can swap out the switching MOSFET in PI circuits? It might add a little inductance? Is this ill-advised?
Thanks for any suggestions,
Thomas
First post here. My name is Thomas. I am a n00b to both electronics and metal detection—my background is in organic chemistry. Anyways, I have a Clone PI-W, the SMD version with the row of LEDs (not the through hole / LCD). Doing some experiments at home winding smaller diameter coils out of 26 AWG FEP teflon (4.8cm former). I guess this would be called a button coil or super sniper coil. The instruction sheet that comes with the Clone PI gives a chart with coils of varying diameters, the smallest being 120cm, The recommended inductances are from 380 uH to 406 uH, with resistance varying from 1.6 to 2.3 ohm. I wound a coil of smaller inductance by accident when I ran out of wire, about 312 uH, 1.63 ohm, and this coil showed the best distance in detecting a 1 grain bar of gold through stacked test layers. I had various coils wound with magnet wire, standard silicone pick-up wire, etc., tried FEP teflon to reduce interwinding capacitance, and just by chance I wound this coil with the low inductance which outperforms all the other ones that I wound with 4.8 cm former trying to stick to the instruction sheet ranges for inductance and resistance. I measured capacitance of the coils with a NanoVNA, so I am factoring that value in, and based on all the data, the low inductance is what really seems to boost performance here.
So then I started gradually peeling off turns from another identical coil, coming down by ~25 uH at a time, and there seemed to be a steady improvement in performance all the way down to 200 uH (1.24 ohm), which was significantly better for distance than even the fortuitous 312 uH I started with, at which point I halted and started consulting AI about potential issues. I was advised to test the voltage spike across the drain and source of the MOSFET (IRF740 - 400 V rating), and was alarmed to measure a 470 V spike. The original 312 uH was testing around 372V, and a 400 uH coil, same wire and diameter, tested at 312 V. The thing was still working and still continues to work with that MOSFET but I’ve been nervous about leaving it on for too long with the best coil so far if it has that kind of spike. Maybe it didn’t avalanche but I may have been close.
So — any general suggestions for exploring down in this low inductance range?
AI suggested another MOSFET with a higher voltage rating: STP10NK60Z, 600V rating, I got a few, swapped it out, but it did not work at all. The start-up sequence was wonky without the usual beeps and there was no detector response to speak of (tried the 312 uH coil), with some very erratic beeping. I did not scope anything with it installed, so I don’t know why it doesn’t interface. The MOSFET passed all multimeter checks before and after. The most significant discrepancy from the datasheet between this and the original IRF740 MOSFET is with the output capacitance and reverse transfer capacitance: (IRF - 330 pF output, 120 pF reverse, STP - 156 pF output, 37 pF reverse)… AI seemed to think this could be an issue.
I installed the original IRF740 and the detector resumed completely normal behavior. This original MOSFET passed all multimeter checks. So the champion 200uH coil did not seem to kill anything outright.
The IRF840 has a 500V limit, so I will probably try that. This subbed for IRF740 in the PIKINI V4 according to one user. Any other general suggestions or advice for mods to manage higher voltage spikes?
Also, do people ever install a socket so they can swap out the switching MOSFET in PI circuits? It might add a little inductance? Is this ill-advised?
Thanks for any suggestions,
Thomas