PimentoUK

It appears that it's the same cable as before, then. The Manti cable seems 'tougher' in general. Instead of the usual 'plain string' that's run along with the wires, it has Kevlar thread. The screened TX cable has a very thin clear wrap of mylar film. This is tough, and flexible, but won't have the potential to crack like the thicker PVC(?) insulation used in the Eqx cable. The 5 thin wires are all 'tinsel-core' wire, multi-strand copper with fine polyester/nylon etc threads mixed in , for strength.

Out of curiosity, do you think the cable is the same as the original model? I've had my hands on a Manti coil for repair recently, and noticed the cable is thinner ( by about 1mm ) than the original Eqx. It has internal differences, including thinner more flexible insulation on the internal screened TX coil cable. I was wondering if this may be something that filtered down to the Eqx, as I think the Eqx cable has some known weaknesses.

"Formula on conductivity was off several sites and all had a value*log7 or log6" The values are in error, probably the factor of 10 got overlooked. Silver and tin are nowhere close, about 106% IACS and 15% IACS, respectively. And yes, working from memory, I got tin and lead wrong, tin is about twice the conductivity of lead, not half. The hazards of having resistivity AND conductivity as commonly used figures. Here's a pretty comprehensive list of metals/alloys and their electrical characteristics: https://www.effectrode.com/knowledge-base/conductivity-of-metals-sorted-by-resistivity/ If I can track down my 'fake dollars' thread on Tom D's OLD forum, I will add a link here. It took some sleuthing, but here it is: https://www.dankowskidetectors.com/discussions/read.php?2,27267,27267#msg-27267

The flaw in the data is the electrical conductivity values. These are PURE element data: you need to use realistic accurate alloy data, then you may have something to analyse. For example: 24ct gold is almost never encountered, and the electrical conductivity is VERY influenced by alloying. Even 'coin alloy' 0.900 fine gold has a miserable conductivity, almost the same as pure tin, and only half Edit: twice the conductivity of lead, for example. [ I used tin as my metal of choice to make 'fake' gold dollars, many years ago on Tom Dankowski's forum - these were physically very close to the real thing in diameter/thickness, but using pure ( or close ) tin solder. As a result, they ID'ed the same as the gold original, and ideally would give the same signal strength, and other behaviour as them, too ] And 'drinks can' aluminium is quite a fancy alloy ( hence the very high strength ), and this results in conductivity dropping significantly ( from memory 67% IACS down to 37% ) Even lead isn't guaranteed pure, all kinds of contaminants can be included, tin being an obvious one, but bismuth and antimony can be there. ( we often find printers 'type' blocks in UK farmland. They are made from a lead alloy called 'type metal' , and when the printing industry is finished with them, they get repurposed as 'lead', for casting musket balls, and anything else that lead would've be made into )

"I would be interested in a similar physical-principle-based description of how gold, silver, and plated chains behave" I think the very variable nature of how every individual link contributes and how they interact, would make this a tricky target to model. Some links will be 'flat on' , some will be 'on edge', some 'tilted'. Links close together will 'talk' to each other magnetically, though this isn't a strong effect. ( likely relevant to hoards of coins in a pot, though ) If there is little interaction, multiple links can behave like a single link with a stronger signal. So 100 links will give a response 100 times stronger than a single one ( if they were all flat on ). 100 times would equate to ( roughly ) a detectable depth increase of about 2.1 times. In practice, many links will be 'edge on' or 'tilted'. If you have any specific sample chains you could provide some details about, I may be able to have a crack at analysing them?

Update: After examining Clive's coil, the wire break was found to be inside the coil, and hence unrepairable. It seems to be a weak point in the coil, this is the second such failure I've seen. The thin wires do a 90 degree bend after they leave the bottom of the strain-relief gland, and pass over to the small circuit-board in the 'fat spoke'. This bend seems to be a high-stress area, possibly also affected by the potting epoxy holding the wires firmly until they enter the gland.

Quote: "A stat that could be equally meaningless would be ring size. ...Size and shape can change numbers. I don't know here if rings size/diameter will graph into something meaningful?" Seeing as this is the thread for nerdy techy analysis, I thought I'd contribute: Loops can be analysed mathematically - we did some work on the Geotech1 forum about this. Electrical resistance around the loop, and inductance of the complete loop, are what matters. Then L / R ( inductance divided by resistance ) will give the target time-constant, and hence what ID value it gives. Basic electrical theory states that a length of wire has a resistance that's proportional to the length of the wire, and inversely proportional to the cross-sectional area of the wire. So if you doubled the diameter of a finger ring, you would double its loop resistance, for example. Inductance is much harder mathematically ... but, basically, a straight piece of wire has an inductance, that is proportional to the length, and bending it into a loop shape makes surprisingly little difference to the inductance. The result is that making a ring twice the diameter roughly doubles its inductance; it may be 5 or 10% more than double. So, if the resistance of the 'big ring' doubles, and the inductance almost doubles, the L / R changes very little. In plain English : large rings read almost the same as small ones. It's the cross-sectional area that matters. Clearly, the electrical resistivity of the metal is very important. Here again, it's not obvious. Very pure gold conducts electricity very well. But alloy it, and it's conductivity plummets, and it doesn't make much difference if it's .900 / 0.917 fine or poor 9 / 10 carat . What can be significant is the presence of nickel in the alloy. Nickel is a conductivity-killer, and low carat rings often have it included, to influence the colour. So, if you're wanting to do some analysis, studying cross-section area versus TID would be modestly informative, when combined with carat rating. Also, rings with big bezels, stone clasps etc, don't analyse like a plain band, so it's hard to pin down their characteristics. Edit: I should add: This isn't just 'mathematics' , it's easily testable on real metal test-pieces. We made copper rings 20mm diameter in a wide range of diameters from 0.1 mm to 3mm , and larger diameter loops from the same wires. Theory and practice agree well, the only real discrepancy is with the fat 2mm+ wire, where skin effect starts to make the wire appear thinner than it actually is. You can do it yourself, too. Get a length of about 1mm diameter copper wire. Solder 65mm of it into a 20mm loop, and make a big loop, eg. 200 mm across (from 65cm of the wire). Little difference in TID is observable.

"Is there a better way for me to do this given my minimal soldering skills?" Of course there is - get a professional to do the job. Which is my primary interest in helping diagnose the problem etc, I was thinking there may be a few pounds I could make ... Regarding use of C meter: Obtain a length of wire the length of the coil cable; any size, doesn't need to be insulated, even. Wrap it gently in a spiral down the outside of the cable, maybe use a few bit of tape to hold it in place. See pics attached. Connect one end of this wire to one C-meter connector, and connect the other meter contact to the coil connector: pin 3, pin 4, pin 5, and no pin , in turn. Keep it all away from metal, eg. on a wooden table. Based on what I've previously done, you'll get about 10pF / 0.01 nF with no pin connected, and 60 pF / 0.06 nF with a full length of wire connected, eg. Pin 3. What reading you get for the open-circuit pin depends on the break location. If you see the same 60pF as the other two, the wire is intact down to the coil. If it's a low reading, the wire break is near/in the coil socket area.

You borrow the C meter, I'll sort out a description of the test method, probably with a photo to explain. What you can try: measure the open-circuit 3 - 4 connection, and wiggle the cable close to the top of the gland, looking for any indication of the 2 kOhm. Pushing the cable together, rather than just random wiggling, may give a result. My worry is the break may appear to be at the top of the strain-relief ( based on 'wiggle' ), but in practice is much lower, where it's unfixable.

So, here's the pinout of the coil connector: Pin 3 is the 0V ground. Pins 4,5 are the two complementary outputs. Pin 7 is the +3.3 V supply. Resistance between pin 3 and pin 4 is 2000 Ohms +/- 200 Ohms Resistance between pin 3 and pin 5 is the same. Resistance between pin 3 and pin 7 is 3600 Ohms. I suspect you'll find one of pin 4 or 5 is open circuit. Q: Does you multimeter have a capacitance measuring feature? Often there's a small 2-pin socket on the front panel?

"Maybe we can get some insight into the design of the coil and control unit" It's pretty well understood, I was just sparing you the details. People have been poking around these machines since they came out. "I have some electrical skills including a multimeter. If you could give me some suggestions on how to test this .." No need for wire-cutting, it's just resistance measurements on the connector pins. Give me a day to sort out the info, I'l post it up.

It may be repairable, it just depends on the fault location. Skipping the details: There are two signals coming from the coil, a complementary pair. One goes up when the other goes down. If one is missing, the detector still appears to function , but will give lower sensitivity, particularly to low conductors ( low ID number targets ), and will give lower than normal ID values. Missing signals are normally caused by a broken wire, they are quite thin and fragile. If the break is in/near the coil connector, it's possible to fit a new connector. If it's in the cable , typically just near the top of the strain-relief gland on the coil, it may be repairable, though it's fiddly, and clearly water-tightness is 'lost'. But ... it can be a really low break, inside the coil, in which case it's unrealistic to fix. Telescopic shafts seem to be bad for flexing the cable close to the coil, and causing these issues. The good news, maybe: I'm in the UK, and have had a go at a few of these failures now. With some successes, and a failure. Do you have 'tech skills' , and an electricians multimeter? It may help pin down the fault if you can test a few things first, then repairs can be better considered. If it turned out to be unrepairable, I would still be interested in the coil, and could offer you a modest amount for it.

I think Regis 'Rat' will be pleased to learn this device has just recently been re-manufactured, as a hobbyist 'build it yourself' partial kit. Two versions - 'vanilla' original , and LCD screen / microcontroller adjustable-features improved version. Currently being discussed on the Geotech1 forum.

I think you chances of success are high. Problems you will have include: The screen of the shielded cable is sort-of-Litz, it's enamelled copper wire ( for better high-frequency response ). As a result it's a bit of a pain to solder. When I did a similar fix, I tinned each wire strand individually, burning off the enamel with the iron. Then I grouped them together in bundles of 5 strands, and made multiple 5-strand joints. I thought this would be better than one big blob joint of all strands. The 5 thin wires are pretty fragile, and a straight solder-joint is likely to have a short life. I used a 2.5" / 6cm extension wire ( correct colour would be best ) for each of them, with the excess folded up so it made a small loop. This then allows some wiggle-room and strain-relief for the wires. Also: the cable outer is made from some silicone-rubber like material, and is quite flexible/stretchy. You can pull back the outer from the repair zone by 20mm, eg. to make room for slipping on heatshrink sleeving . This may minimise the amount you need to cut off when making the repair.

Quote: "If anyone hasn't done the 2.0 update by now, I'm sad to say you're really missing out on serious improvement to the D2 overall." I mentioned this in another post, but here seems a suitable place to repeat it: Version 2 has some completely new/revised modes, running three frequencies at once, rather than the two in previous firmware versions. The transmitted waveform is a three-level one, in the way the DeepHC transmission is done, but obviously more complex. ( it's being discussed on the Geotech1 forum currently ) So it's not unexpected for V2 to behave in a noticeably different way.