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

[6666]

To make a 300uH elliptical coil from scratch for the Mirage, I used this approach - refer attached: [ATTACH]47298[/ATTACH]

Wow thanks for the great document, well done and great detail, very impressed.

[MartyJ1963]

@Elliot,,,,DITTO,,brilliantly written guide,,Looking forward to building the new Mirage. Regards,,,Marty

[green]

[ATTACH]47287[/ATTACH] I haven't made any recordings yet. I've pulled apart the test jig setup and am in the process of transferring everything into the shielded Takachi project box. I really love this box. It's incredibly well designed and manufactured, and is the perfect size! It even has two small metal mesh covered conductive foam rods (2.5mm dia.) to fit inside the interconnecting crevice between the lid and the main box to ensure EMI shielding.
As soon as I get a chance, I'll make some recordings. Does anyone have any requests for which objects they want me to test (please don't ask me to test any foreign currency unless you're prepared to send me a sample - I'm in Australia)?

Some targets for discussion. 1x1 inch targets cut from regular strength aluminum foil(kitchen foil). Thickness can vary, probably need to include measured thickness. Should be easy to detect. 1 layer maybe not? 5+mm square and a 10mm square cut from the side of an aluminum drink can. Hard to cut exact so include micrometer measurements(example, 5.1x5.2mm). Charted decay for the foil targets, included US nickel, dime and quarter for comparison.

[Elliot]

Some targets for discussion. 1x1 inch targets cut from regular strength aluminum foil(kitchen foil). Thickness can vary, probably need to include measured thickness. Should be easy to detect. 1 layer maybe not? 5+mm square and a 10mm square cut from the side of an aluminum drink can. Hard to cut exact so include micrometer measurements(example, 5.1x5.2mm). Charted decay for the foil targets, included US nickel, dime and quarter for comparison.

What was your hardware test set-up to chart the decay times? Also, it is assumed that you are driving your output transistor with a +/-5V pulse, but the pulse length I use is 65 uS, not 160uS, which will make a relative difference too.

[Elliot]

Power Supply Noise

Has anyone come up with a design for a cleaner power supply for the +5V and -5V supplies. Using an NE555 timer to generate voltages adds small spikes to the power supply, and this gets amplified along the path. The attached photos show the following:

- Average voltages (+5V and output of pre-amplifier) - they look nice and clean
- Voltages by Time - lower trace shows the jitter as well
- Voltages by snapshot - can see the noise in the 5V transfers directly to the output of the pre-amplifier

Has anyone tried using a Low Drop Out (LDO) chip to smooth things out?

[green]

What was your hardware test set-up to chart the decay times? Also, it is assumed that you are driving your output transistor with a +/-5V pulse, but the pulse length I use is 65 uS, not 160uS, which will make a relative difference too.

https://www.geotech1.com/forums/show...567#post253567

My detector isn't working at the moment, so charted the targets with my bench tester. Doesn't show maximum detection depth, can't compare with your detector. Shows amplitude of the three foil targets are the same at about 5us after Tx off. The foil and can targets are possible targets that anyone could have available? One problem, thickness can vary. They are short time constant targets so Tx on time shouldn't have much effect, peak current yes. Just wondering if there are some targets that anyone could have available to compare detection distance.

[green]

Schematic reply #21 doesn't show MUR460 diode between mosfet and coil. Tried both ways, ended with diode.

[SVEN1]

Has anyone come up with a design for a cleaner power supply for the +5V and -5V supplies. Using an NE555 timer to generate voltages adds small spikes to the power supply, and this gets amplified along the path. The attached photos show the following:

- Average voltages (+5V and output of pre-amplifier) - they look nice and clean
- Voltages by Time - lower trace shows the jitter as well
- Voltages by snapshot - can see the noise in the 5V transfers directly to the output of the pre-amplifier

Has anyone tried using a Low Drop Out (LDO) chip to smooth things out?

Yeah there is some slight noise with the original power supply. Some get kind of fanatical about it and complain. The noise was found not to have any bearing on the way the Mirage operated or degrading performance. So it was left as is.
Will say, a new power supply was designed which no longer uses the NE555 but, two other chips replace it, its proprietary circuit design, so I was asked not to elaborate on it. Takes care of the noise and ties in with other parts of the Pi circuit to make a much more stable detector. So a cleaner power supply can be made.

[Elliot]

Is there a cleaner schematic of the Target Response Tester somewhere?

[6666]

Is there a cleaner schematic of the Target Response Tester somewhere?

That looks like a Green creation

Been trying to post a simple 555 power mod to stop the hash ripple but forum is acting up

[Elliot]

That looks like a Green creation

Been trying to post a simple 555 power mod to stop the hash ripple but forum is acting up

Keep trying to post it, I'm super interested. I'm having forum operational issues as well. If you have no luck, email it to [email protected] if you please.

The only other issue I've noticed with the Mirage is that the more you adjust the Target Sample Delay out, the more jittery it gets. On Whites detectors, they only let you adjust it up to 25uS, which is probably enough anyway. On the Mirage, with a 10k pot, you can adjust it up to about 65uS. If anyone has some mods for improving the jitter, please advise.

[6666]

email sent its just two parts to add

[green]

Is there a cleaner schematic of the Target Response Tester somewhere?

Just the hand drawn schematic. Enlarged it so a little easier to read.

[Elliot]

Pushing the coil driver boundaries - to 600V, with lower output capacitance

I am continuing to try and push the boundaries of performance of the Mirage PI and am working on some potential improvements. I was experimenting some more with the in-field operating permutations that PCB V1.1 offers - 3 choices of TX frequency, 2 choices of core pulse width, 3 choices of pre-amp sensitivity, and 2 choices of sample pulse width - that's equivalent to 3 x 2 x 3 x 2 = 24 operating modes to choose from before you start tweaking the delay pot, threshold pot, gain pot etc. In all this tweaking and testing, I soon found the thermal limit of the IRF840 coil driver, probably because I was changing modes using the temporary wire links and driving it full bore between wire connections. The IRF840 went into thermal overload and shut down until it cooled a bit (it seems to be ruggedly built!). I then thought, why not find a higher spec coil driver with a higher Vds (breakdown) voltage rating (i.e. >500V). After some research, I stumbled across a Vishay IRFBC40LC that has a 600V Vds rating (whether the coil winding can cope with that voltage is another matter). To my surprise, I also noticed in the specs that the Coss (output capacitance) is a LOT lower than the IRF840 (at all driven voltages), which means it should help produce a relatively quicker decay time, and enable faster Rx sampling, and be able to find smaller gold (and other low conductance metals). My question to the learned team is, have you ever tried the IRFBC40LC, and if so, what were your results? PS, the box-to-coil lead I use is a Canare DA202, which has a dielectric strength = 500V AC/1min., so I may be pushing things perhaps!

[waltr]

That IRFBC40LC looks good. Just added it to my 'parts ordering' list to try.

The lower C would be a benifit even if not pushing the fly-back Voltage.