Mirage PI Project, anyone interested?

[SVEN1]

Earliest versions were named Cobalt which eventually morphed into the Mirage.

[SVEN1]

Hi All. I purchased what I was advised were the last two PC boards off Sven just recently. I have made one up, and am fine tuning it as I write. The board has a few things that I want to modify to better fit components (e.g. several trim pots had to sit above a resistor due to them being too close together for the trimpot I used); there was a superfluous wire link position (carry over from previous mods no doubt), I wanted the option to use detachable wire connectors for the main pots (not be restricted to just solder joints), and I wanted to add the following features:

- 3 way selectable Centre frequency switch (for potential pesky high EMI sites),
- Differential amplifier boost switch (+20% gain),
- Main Amp gain switch (bypasses the 1k that was added back in later versions),
- SAT hold switch (for when pinpointing) and,
- a good 'ol volume control!

The bulk of these are for maximising flexibility in the field, when choosing between operating performance and/or stability, without having to rip the machine apart every time you want to tweak something.

I have spent considerable time designing the modified PCB using Pcbnew (using a two layer FR4 PCB due to all those pesky wire links), and will organise a prototype board to ensure all is working fine before I order any more boards (and, NO, I haven't used a 1000uf electro like someone else sounds like they've done lol).

If anyone is interested in purchasing a board in due course with my mods, let me know, and I'll order a batch to suit (PS: I'm in Australia).

Frequency control needs to be a multi-turn for fine adjustments. Selectable switch wouldn't even begin to cover the actual frequency needed depending upon conditions encountered. A single turn pot will work if your careful with making fine adjustments.

Many experiments were made using all sorts of amps, we settled what we had on the current version. Two stage amp mod was done, outcome was poor, switched back to a single stage early on.
Main amp switch to switch the 1K out and in will not work, has been tried, you will encounter feedback etc. Its either jumper or resister. We changed some other resistors to increase gain but, found the circuit was becoming unstable and harder to keep in tune.

SAT , a pinpoint switch was originally designed in. The pinpoint switch just reduced SAT to min. for a non-motion type pinpoint. Because the Mirage was so easy to pinpoint at 75% SAT setting, the pinpoint switch was not needed. Just stop the coil on top of target, let it retune, give a small wiggle and the target has narrowed, do it again if needed. Target will be in the coils center. By the time you use the PP switch, another person will have pinpointed the target and be on the ground digging.

Volume control was not placed on the PCB as you just wire the pot in......in our case, we use volume controls in our headsets and we rarely used the speaker on the control box. Mostly for demonstrations.

We used a lot of jumpers, since it was easier to work on a single sided board when making circuit board changes-revisions for testing purposes. A 2 sided board was considered for a final version. Just never happened. A SMT parts version was also considered, just too hard to build for the average builder, so it never happened, plus it meant buying a whole pile of SMT parts. We already had a pile of thru-the-hole parts...

There are a number of Geotech Forum members that have built the Mirage, like any project some with much better success. Also, many have experimented with the circuit and may give you some more insight. I was fine with my current version to suit my needs and purposes. The beauty of the project is for those who are able to, go ahead and experiment, just keep in mind what I posted above. I have built over 50 Mirage circuit boards of different variations and will say some builds were much better than others depending upon the parts used and the tolerances of them and search coil build. Any complete Mirage detectors that I have built for others needed to have equal performance to my personal unit . If not the circuit board was scrapped.

Side note, when I use the 8.5" search coil, rarely need to use more than 25% Gain, goes plenty deep. 75% Gain is usually max used. The 100% Gain setting is there for those who feel the need to adjust that high.

Please remember the Mirage was designed around finding small gold and gold jewelry, at max. depth the circuit would allow. More of a niche detector. At the expense of finding higher conductor metals at greater depth. If your interested ins such, develop the Mirage more in line with Carl's Hammerhead as did George when he designed the CrossBow. I have always said the Mirage is at one end of the Rainbow and the CrossBow at the other, they can compliment each other.

Also, a new circuit board design a forked branch of the Mirage that has been worked on for a few years is now revamped over quite a few versions. Final outcome is superior to the Mirage. It has two audio versions, one sound more like a VLF and the other the standard tic-tic-tic. Samples down to 6us. Frequency can go back to a single turn pot, super stable rarely needs adjustment. Lots of parts and IC's added. As for performance, equals the Mirage for depth or slightly better and has better sensitivity to very tiny gold nuggets. I am not at liberty to produce these new boards or give out the circuit design, as it has propriety circuit design incorporated into it. But, can say there are several Geotech members who have gotten and built one of these new version boards and can attest to what I have written. My electronics friend who designed the original PI circuits from the Hammerhead, with the help of Carl M., Reg Sniff, Eric Foster. He should be working for a major metal detector manufacturer..... but, he is 84 years old now, retired electronics engineer and enjoying his private life, family and friends. Now he does keep his mind active by experimenting with PI designs.

The Mirage started out as just a modified Hammerhead built on a bread board. I was able to take the schematics and produce the first actual circuit boards back in 2013 after Ivonic took the schematics and layed out the first Sprint board design for us. The circuit was refined so many times since then. Now, will say one of the early designs worked excellent and easy for the DIY person to build. It was a perfect Pi for hunting trashy parks, schoolyards and picnic grounds with the 7" search coil.

[Elliot]

Hi All. I purchased what I was advised were the last two PC boards off Sven just recently. I have made one up, and am fine tuning it as I write. The board has a few things that I want to modify to better fit components (e.g. several trim pots had to sit above a resistor due to them being too close together for the trimpot I used); there was a superfluous wire link position (carry over from previous mods no doubt), I wanted the option to use detachable wire connectors for the main pots (not be restricted to just solder joints), and I wanted to add the following features:

- 3 way selectable Centre frequency switch (for potential pesky high EMI sites),
- Differential amplifier boost switch (+20% gain),
- Main Amp gain switch (bypasses the 1k that was added back in later versions),
- SAT hold switch (for when pinpointing) and,
- a good 'ol volume control!

The bulk of these are for maximising flexibility in the field, when choosing between operating performance and/or stability, without having to rip the machine apart every time you want to tweak something.

I have spent considerable time designing the modified PCB using Pcbnew (using a two layer FR4 PCB due to all those pesky wire links), and will organise a prototype board to ensure all is working fine before I order any more boards (and, NO, I haven't used a 1000uf electro like someone else sounds like they've done lol).

If anyone is interested in purchasing a board in due course with my mods, let me know, and I'll order a batch to suit (PS: I'm in Australia).

[SVEN1]

To all those reading this post and have inquired about obtaining a PCB.
I no longer have any left for sale.

Just a note there are some unofficial copies made, of the Mirage PCB floating out there, it does not completely resemble any PCB versions I have personally made up.
If you didn't get the PCB directly from me, probably a copy and if you build it and it doesn't work, take it up with the person who sold it to you.
One copy I saw had a large 1000uf electro capacitor near the upper right corner of the PCB.
No Mirage PCB has an electro cap of more than 100uf.

[SVEN1]

Really need only one PCB mounting bolt to ground to metal case. We only use three as we originally only had three mounting holes due to lack of room on earlier version boards. After shifting parts around during revisions, it allowed enough room to put the fourth mounting hole on the board.


In this case, used a Vintage Garrett, the grounding plate was a copper clad one sided PCB board as the Garrett case was plastic. I have also used a solid copper plate as a grounding plate in other plastic enclosures. If using a metal box, ground the PCB directly to it using metal stand-offs.

[Elliot]

The speaker volume is very low with an 8 ohm speaker - even if the 10 ohm series resistor is bypassed. What workaround do you recommend for this?
Should be plenty loud, might try swapping out the 3906?

I also want to add the auto-track freeze feature as per the Hammerhead design. Do I simply cut the track and put a switch between pin 5 of TL072 and the 220k Sat Adj resistor?
Do you mean a static pinpoint? Originally there was such a feature on the original circut board versions. It was removed, we felt it was needed. You can add it back in, it basically switches SAT to min. setting. You will have to rework the PCB you have, see pix, can then add a toggle switch or a normally on type push button switch wired to the solder pads marked PP.
[ATTACH]45845[/ATTACH]

Thanks Sven. So for the static pinpoint function, you're effectively just placing a switch where shown in the attached image.

I'm also curious about why the earth point near the SAT adj pot (and one other corner mounting point on the pcb - i.e. bottom left corner when looking from the top side) isn't connected to earth on the pcb. The machine works well without these points connected to earth. Are you assuming the the pcb is mounted in a metal box and the earth points are connected via metal stand-off posts?

[SVEN1]

The speaker volume is very low with an 8 ohm speaker - even if the 10 ohm series resistor is bypassed. What workaround do you recommend for this?
Should be plenty loud, might try swapping out the 3906?

I also want to add the auto-track freeze feature as per the Hammerhead design. Do I simply cut the track and put a switch between pin 5 of TL072 and the 220k Sat Adj resistor?
Do you mean a static pinpoint? Originally there was such a feature on the original circut board versions. It was removed, we felt it was needed. You can add it back in, it basically switches SAT to min. setting. You will have to rework the PCB you have, see pix, can then add a toggle switch or a normally on type push button switch wired to the solder pads marked PP.

[waltr]

I have researched everywhere about how to optimise the damping resistor. Some articles say that a lower resistance can overdamp the return signal, while others say reduce it as far as you reasonably can, I assume that a lower resistance will deliver a faster voltage drop, but how do I optimise? Do I simply keep reducing it until I start seeing some ringing on the cro, and then back off of a bit?

Look at Pre-amp output on scope. Without damping R you will see ringing. Decrease value of damping R until ringing is gone and a smooth exponential decay curve is seen.

I prefer a shielded (metal) enclosure but seem many PI detectors are built into plastic boxes and work fine.

If you have shielded the coil then this shield needs to be connected to the circuit "ground" (virtual earth). Several ways to do this- either on the cable shield or on the "Ground" coil connection.
On my modified HH I do not shield the coil or run a shielded cable. Works fine with CAT-5 cable but YMMV.

Which cable and shield connection scheme works best may need to be determined experimentally.

[Elliot]

Hi Sven,
I have completed building the circuit board, and all is working well.
A couple of questions:
- Does the pcb need to enclosed in a metal (i.e. shielded) box, or can I install it in a plastic box with no worries?
- I have found a very low capacitance cable to go between the search head and the control box (7pf per foot between adjacent wire conductors - excluding the shield). It also has a woven shield that I can use. Do I need to use the shield, or can I simply use the twin core wires (twisted)?
- If I need to use the shield, should it be connected within the search coil to the coil shield/virtual earth star point, or could/should it be connected to the virtual earth (+ve in this case) at the pcb?
- The speaker volume is very low with an 8 ohm speaker - even if the 10 ohm series resistor is bypassed. What workaround do you recommend for this?
- I have built what I think may be a very fast coil using a planar spiral design (2.2 layers), with foam padded filler used before adding copper foil shielding (measuring the self resonating frequency seems to be a vague and fraught process). I have researched everywhere about how to optimise the damping resistor. Some articles say that a lower resistance can overdamp the return signal, while others say reduce it as far as you reasonably can, I assume that a lower resistance will deliver a faster voltage drop, but how do I optimise? Do I simply keep reducing it until I start seeing some ringing on the cro, and then back off of a bit?
- I also want to add the auto-track freeze feature as per the Hammerhead design. Do I simply cut the track and put a switch between pin 5 of TL072 and the 220k Sat Adj resistor?

Thanks in advance!
Elliot

[Elliot]

Hi Sven,
I have completed building the circuit board, and all is working well.
A couple of questions:
- Does the pcb need to enclosed in a metal (i.e. shielded) box, or can I install it in a plastic box with no worries?
- I have found a very low capacitance cable to go between the search head and the control box (7pf per foot between adjacent wire conductors - excluding the shield). It also has a woven shield that I can use. Do I need to use the shield, or can I simply use the twin core wires (twisted)?
- If I need to use the shield, should it be connected within the search coil to the coil shield/virtual earth star point, or could/should it be connected to the virtual earth (+ve in this case) at the pcb?
- The speaker volume is very low with an 8 ohm speaker - even if the 10 ohm series resistor is bypassed. What workaround do you recommend for this?
- I have built what I think may be a very fast coil using a planar spiral design (2.2 layers), with foam padded filler used before adding copper foil shielding (measuring the self resonating frequency seems to be a vague and fraught process). I have researched everywhere about how to optimise the damping resistor. Some articles say that a lower resistance can overdamp the return signal, while others say reduce it as far as you reasonably can, I assume that a lower resistance will deliver a faster voltage drop, but how do I optimise? Do I simply keep reducing it until I start seeing some ringing on the cro, and then back off of a bit?
- I also want to add the auto-track freeze feature as per the Hammerhead design. Do I simply cut the track and put a switch between pin 5 of TL072 and the 220k Sat Adj resistor?

Thanks in advance!
Elliot

[homefire]

Roger on the HOT. Just showing signs of cooling down here near the Dotted Line. I have two of these PCB in my TO DO SOME DAY BOX ! Have most parts on hand . Couldn't pass up a good deal . LOL

[SVEN1]

I can sell you the current 2018 version PCB along with the schematics. I do not offer the lay files. Nor do I offer the schematics free, it would be unfair to anyone who has invested $$$ to buy a PCB etc.

Since I have had some bad luck sending items to Indonesia, there will be no guarantee of swift delivery or actual delivery if you order a PCB. And I will no longer send a free replacement.

The only way I will ship a PCB and Parts Kit if it is shipped insured and tracked. But, shipping is very expensive. That way if the parcel gets lost I can file an insurance claim to recover all costs.

I did respond to your message to me on ebay, to search Geotech for a PI to build and a way to contact me direct. If your still interested send me an email thru my website link, click on the mailbox at https://www.treasurelinx.com/zm.html

[hendra]

[QUOTE=Polymer;247978]

I have used a couple different cases for the mirage ( see pics below).. the best fit so far has been the one below in the ebay link or "mirage 1" build and it is relatively cheap compared to other cases. its a nice case for the price too. Just spend some time on your layout before drilling and you'll be fine.

mirage 1 - inside
[ATTACH]38047[/ATTACH]
mirage 1 - front panel
[ATTACH]38048[/ATTACH]
mirage 1 - rear panel
[ATTACH]38049[/ATTACH]

Mirage 2 - front panel
[ATTACH]38050[/ATTACH]
Mirage 2 - completed (difficult case, solid one piece, only removable ends)
[ATTACH]38051[/ATTACH]

Nice build! What parameter do you use the 10-turn pot for? (Rdamp, Delay etc.)

Thanks ... Polymer

hello sir, do you have schematic Mirage PI 2 and the layout pcb and component ? i want to try make Mirage PI
regard

[qute1]

Homefire : No idea but got a slip through the letter box on Friday saying held at customs with £11.75 to pay (you can pay it on-line) and then RM deliver it 4 days later.
I have been soldering all afternoon never seen so many jumpers on a kit before, about 30 little bits of wire to bend and fit on each board (single sided pcb's) but now have the ic sockets and jumpers done on both boards so can start on the psu bits and the rest tomorrow.
It SOOO hot here i am cream crackered after just 4 hours even harder as trying not to drip sweat on the pcbs.

[Polymer]

[QUOTE=geoscash1;224396]I have used a couple different cases for the mirage ( see pics below).. the best fit so far has been the one below in the ebay link or "mirage 1" build and it is relatively cheap compared to other cases. its a nice case for the price too. Just spend some time on your layout before drilling and you'll be fine.

mirage 1 - inside

mirage 1 - front panel

mirage 1 - rear panel


Mirage 2 - front panel

Mirage 2 - completed (difficult case, solid one piece, only removable ends)


Nice build! What parameter do you use the 10-turn pot for? (Rdamp, Delay etc.)

Thanks ... Polymer