Updated Mirage PI (Ex Hammerhead)- Mono Coil, DD Coil or Super D Coil Options

[green]

It looks my my overall rig (including coil, coax lead, IRF840, etc.) produces a fly-back decay time of about <5uS (using my eye, not measured to 1mV above ground) once critically damped, so the impact of the clamping diodes and pre-amp is my next area of focus. I have noticed that when I increase the pre-amp gain, that the "clamped time" does increase a few uS, so the pre-amp is probably the priority area to work on. Green's suggestion of a two stage pre-amp is a good one. Can anyone point me to other posts that resolve this?

[ATTACH]47380[/ATTACH]

Looking at coil decay in spice helped me see better what is happening. Like Carl replied, it's hard to see when the coil volts has decayed to 1mV with a linear Y scale.

Picture with MUR460 diode and diode shorted with a fast coil(SRF 1MHz). Right click Y scale(zip file) to change linear or log.

[Elliot]

Looking at coil decay in spice helped me see better what is happening. Like Carl replied, it's hard to see when the coil volts has decayed to 1mV with a linear Y scale.

Picture with MUR460 diode and diode shorted with a fast coil(SRF 1MHz). Right click Y scale(zip file) to change linear or log.

Thanks Green. The wealth of knowledge on this forum is inspiring! I'm trying my best to keep up. As far as i can tell from your simulations, the main difference between your plots is the damping resistor value, which clearly/significantly effects the decay time to reach 1mV by about 2-3uS (green line?) I also note that the plot with the 965 R1 value has a table top peak fly-back voltage, which is not as good for creating back-EMF in the target I would have thought. It's like cracking a whip I reckon - the sharper/faster the flick, the more "sound" you get back. Also, you have a bypass link on D1 for the plot where R1=420R (which I assume is there to simulate the circuit without the effect of D1). Also, I don't understand the V3 power supply to the cathode end of D2. I haven't seen that configuration before and am not sure what to make of it. Have you chosen the use of the SPA11N60C3 for any particular reason over an IRF840? I tried to use my cro to assess the 1mV cut-off point, but like you said, a linear scale makes it impossible (why don't cheap cros' have a log scale switch????). I really appreciate the feedback! My other big challenge is to find a schematic for a really clean +5v/-5V power supply (all tips most welcome).

[Qiaozhi]

A reminder of this thread may be in order -> What is the purpose of the series diode?

[waltr]

A reminder of this thread may be in order -> What is the purpose of the series diode?

Good read. I must have missed that thread previously. so thanks for posting the link.

[green]

A reminder of this thread may be in order -> What is the purpose of the series diode?

Thanks for the reminder. Had forgotten it, easy when you get old. Read it again, not sure what the conclusion was. Picture reply #46 above suggests diode does shorten delay time.

[green]

I also note that the plot with the 965 R1 value has a table top peak fly-back voltage, which is not as good for creating back-EMF in the target

Change in coil current excites target, rate of change is maximum when voltage is flat topped. Target volts in the simulation is a little higher with the faster decay, not lower.

Also, you have a bypass link on D1 for the plot where R1=420R (which I assume is there to simulate the circuit without the effect of D1)

correct

Also, I don't understand the V3 power supply to the cathode end of D2

The spice simulation doesn't avalanche, it's added to simulate avalanche or a snubber.

Have you chosen the use of the SPA11N60C3 for any particular reason

it's in the list of components and should work.

[godigit1]

Hi Elliot,

Carl has a dual amp on the hh2 Schematic. I made a plug in adapter board that works but its could be redone and made smaller. Its on my to do list.
its a old thread on the hh thread Plug In Dual Front End.

On the power supply the original HH used a 7660 Im wondering why the MIrage design went to a 555?

I also have a seperate HH power supply board drawn up.
Im not sure if the 7660 desigh has the same problem.
Ill check it out when I can get a minute.

On the dual op amps A 4562 alone will nock a couple uS of your sample time versus the 5532.

Did you by chance get any scope shots of your coil before shielding was added. Im curious how much the shielding effected your sample time.
I would call your coil a heavy shielded coil.
Im thinking it would make a good salt water beach coil possibly.
Thanks

[Elliot]

Thanks for the reminder. Had forgotten it, easy when you get old. Read it again, not sure what the conclusion was. Picture reply #46 above suggests diode does shorten delay time.

That was an epic read! It would have been nice to compare the benefits of just using a very low Coss transistor in the first place (eg. IRF840B - discontinued). Is there a consensus for the best fast coil winding design and materials selection? I see amazing claims for coils with SRF's in the MHz. Are these results substantiated through a recognised measurement methodology - like in comment #15?

[Elliot]

On the power supply the original HH used a 7660 Im wondering why the Mirage design went to a 555?

My understanding with the 7660 (backed up by the datasheet) is that there are issues with a voltage drop under load (not good for the analogue sections, okay for the digital sections), voltage temperature dependencies, and upper battery supply voltage limits.

Thanks for the tip on the LM4562. I'll look into it.

Unfortunately I didn't get any cro shots of the coil before shielding. The "heavy" shield, fibreglass, resin and polyester gap filler did marginally increase the coil capacitance from 69.9pF to 84.7pF (14.8pF), but that is dwarfed by the coax (15pf per foot), coil driver Coss capacitance etc.... Even the ABS coil case added capacitance. I'm trying to track down a genuine low Coss IRF840B (discontinued) to see how it may help.

My next coil will probably use a combo of copper foil (very small length only, to provide a proper - fail-safe solder point), partially overlapped with Graphit 33 (if I can get it in Australia) / or the 3M metalised (Nickel on copper-plated polyester fabric tape - CN-3190 DS) for 65-70db attenuation.

[Elliot]

Mirage PI - New PCB Batch Now Available V1.2

Hi All, I now have the next batch of pcb's in stock (V1.2) for those interested. No schematic changes made. For improving the construction experience, I moved around a few components in the bottom right corner to make it a little easier to fit the pcb and speaker into the Takachi project box; deleted most of the top pads as I found that when experimenting with different component values and removing and replacing components, I'd loose some of the top pads (most of which weren't functionally required anyway but pcb makers like them as its easier for them to produce); slightly increased the size of the JST connector holes so that no drilling-out was required for the OEM ones (cheap copies had smaller diameter pins), and modified the vias a bit so that if any plated through-holes were drilled out (for any reason), it didn't affect connectivity between various relevant top and bottom tracks. Please direct any purchasing enquiries or requests for the free construction guide to [email protected] to keep these thread pages as relevant as possible. Your interest is my interest, so keep it going!

[godigit1]

My understanding with the 7660 (backed up by the datasheet) is that there are issues with a voltage drop under load (not good for the analogue sections, okay for the digital sections), voltage temperature dependencies, and upper battery supply voltage limits.

Thanks for the tip on the LM4562. I'll look into it.

Unfortunately I didn't get any cro shots of the coil before shielding. The "heavy" shield, fibreglass, resin and polyester gap filler did marginally increase the coil capacitance from 69.9pF to 84.7pF (14.8pF), but that is dwarfed by the coax (15pf per foot), coil driver Coss capacitance etc.... Even the ABS coil case added capacitance. I'm trying to track down a genuine low Coss IRF840B (discontinued) to see how it may help.

My next coil will probably use a combo of copper foil (very small length only, to provide a proper - fail-safe solder point), partially overlapped with Graphit 33 (if I can get it in Australia) / or the 3M metalised (Nickel on copper-plated polyester fabric tape - CN-3190 DS) for 65-70db attenuation.

Hi Elliot,
Mg Chemicals has nickle shield product called Super shield and also carbon sprays I grabbed the one called total ground its more for work surface so it has additives that make it abrasion resistant.
I was thinking it might hold up better at connection than the regular carbon paint.
I dont know if you can get them in Au though. I would think they would be all over the Au market as well.

I was running the max412 but telano turned me on to the 4562 so I have to give credit where due.
thinking you will gain more than a couple us with the double amp set up and then another couple by using a 412 or 4562 over the 5532.
These shots are just swaping amps on a MPP same coil folded 8 mono.

Stats on your coil are similar to the folded 8 mono I have for the TDI except your Capacitance is way better than the pro coil it has 2.1 ohms 280 uh 361pf 500Khz srf.

Im wondering if you have target tested any small gold with it. Curious if it gets gram size and at what depth?

Ive been looking at low loss forming similar to waht is done for radomes. Using a polyurathane foam can be almost invisible to the coil. There is no Diapole allignment in polyurathane or polystyrene. Pvc or abs on the other hand have instantainious switching of the diaploes as sonn as its hit with current.
My thinking is the foam as a form reduces capacitance pcv is good for the shell even though it increases capacitance it positive diapole allignment in my mind propagates signal admittance.
A example of this construction can be seen in the coiltek nugget shooter coils.
I was allready working this way when 6666 provided a link to their advertisement. It does not spell things out the way Im putting it together but to me that is exactly how those coils are made. And coiltek uses a proprietary shell material Polycarbonite. I think that also is diaploe reactive. cant find tests yet. And I think they put anti static aditives in it Not sure.
So wheather Im right or not Im still testing but thats my reverse take of a fast coil.

[green]

Looking at coil decay in spice helped me see better what is happening. Like Carl replied, it's hard to see when the coil volts has decayed to 1mV with a linear Y scale.

Picture with MUR460 diode and diode shorted with a fast coil(SRF 1MHz). Right click Y scale(zip file) to change linear or log.

Spice_forgot to add .8V to supply volts(with diode)so peak current would be the same as diode shorted. Small increase in time to .001V.

Recorded coil volts decay with diode and diode shorted. Coil connected to my TRT tester. Measured circuit resonance(x10 probe across coil)no Rd, with diode and with diode shorted. Calculated Rd=pi*L*circuit resonance. Charted linear X log Y.

Spice and recorded data with diode similar, recorded and spice resonance(1MHz). Recorded with diode shorted slower decay than spice with diode shorted, recorded resonance(328kHz)spice resonance(430kHz).

[Elliot]

I was running the max412 but telano turned me on to the 4562 so I have to give credit where due.
thinking you will gain more than a couple us with the double amp set up and then another couple by using a 412 or 4562 over the 5532.
These shots are just swapping amps on a MPP same coil folded 8 mono.
[ATTACH]47404[/ATTACH][ATTACH]47405[/ATTACH][ATTACH]47406[/ATTACH]
Stats on your coil are similar to the folded 8 mono I have for the TDI except your Capacitance is way better than the pro coil it has 2.1 ohms 280 uh 361pf 500Khz srf.

I'm wondering if you have target tested any small gold with it. Curious if it gets gram size and at what depth?

Those screen shots of the 4563 and MAX 4512 are fascinating! The rise (fall) time in yellow is certainly faster with the 4563, and it looks like a couple of uS is shaved off. I note that the clamped diode width is still around 14-15uS in both scenarios though. Mine is about 11uS for some reason.

[6666]

godigit1, what was the blue spikey pulse in your 3 pictures of the preamp in post #56 ?

[Elliot]

Stats on your coil are similar to the folded 8 mono I have for the TDI except your Capacitance is way better than the pro coil it has 2.1 ohms 280 uh 361pf 500Khz srf.

Im wondering if you have target tested any small gold with it. Curious if it gets gram size and at what depth?

Ive been looking at low loss forming similar to waht is done for radomes. Using a polyurathane foam can be almost invisible to the coil. There is no Diapole allignment in polyurathane or polystyrene. Pvc or abs on the other hand have instantainious switching of the diaploes as sonn as its hit with current.
My thinking is the foam as a form reduces capacitance pcv is good for the shell even though it increases capacitance it positive diapole allignment in my mind propagates signal admittance.
A example of this construction can be seen in the coiltek nugget shooter coils.
I was allready working this way when 6666 provided a link to their advertisement. It does not spell things out the way Im putting it together but to me that is exactly how those coils are made. And coiltek uses a proprietary shell material Polycarbonite. I think that also is diaploe reactive. cant find tests yet. And I think they put anti static aditives in it Not sure.
So wheather Im right or not Im still testing but thats my reverse take of a fast coil.

Godigit1 - you're onto something here! My Mirage test coil (with shield, but not potted in the ABS shell) capacitance is 84.7 pF. The Canare coax cable is 15pF per foot (I have 2 metres attached right now for my test bench purposes = 98.4 pF). My total coil (potted) and cable inductance is 304.5uH. The SRF= 555kHz. The total calculated CPar is 270pF from CPar = 1/[(2*pi*SRF)^2 * L)]. Therefore, I can only deduce that the combining of the coil to the coax cable and the potted search head adds (270 - 84.7 - 98.4) = 86.9pF.

The idea of using expanded polystyrene foam and poly-carbonate (i.e. Non ABS) for the search head filler/cover is worth looking into for sure.

No, I haven't checked the coil's capability on small gold (1 gram), as I only carry 10 ounce samples on me (lol). Seriously though, I have an old gold crown I use (better in my pocket than the dentists!) that weighs 1.57 grams (with residual dental glue - yuck). I am still fine tuning the detector, but can get it at 3" at present. I need to reduce the sample time window more to get a better result, but the coil itself is more than capable, I'm pretty sure.