PimentoUK

I think a more marketable detector is the digital/analogue hybrid, a good example of which is the XP GoldMaxxPower. This uses a micro to do most of the processing and audio generation, but does it in an analogue style. The control-box is largely analogue, rotary pots for GB, disc, volume.

One seemingly smart solution would be for the Vanquish to have battery-charging electronics built into it. Perhaps something similar to the XP GoldMaxPower - which used a spare pin on the coil connector ( and the existing ground pin ) for charging. The charger power-pack had a trickle-charge current limited output, for a 14 hour charge. The additional electronics inside the detector were minimal .. from memory, a diode ( to prevent discharge from the connector ) a resettable fuse type device ( polyfuse ) and a resistor to limit the current a bit.
The problem with this arrangement, is it allows you to charge regular dry batteries inadvertently, which is not a good thing.

 

You are right about the tedium of removing four AA's to charge them every time. I have a Fisher F75 that I run on AA NiMH's. Thankfully, it's amazingly frugal, 30+ hours runtime is normal, so it's not neccessary to charge after every session, so I could live with the charging inconvenience.

But I did devise a scheme where I could charge the NiMH's in-situ, using the headphone socket. Trouble is, I also came up with a scheme where the detector couldn't be accidentally powered up ( the rotary switch is easily disturbed ) unless headphones were plugged into the socket. I never actually made either mod.

( a slip of thin card inserted over one of the +ve battery pips sorts out the unwanted turn-on issue ... simple )

Guys ... he's not wanting it to collapse small for transportation, he's wanting it to be short WHEN HE IS USING IT. If you insert the lower rod up to the top locating hole on the central section, it's about 1 metre total ( I haven't checked on my one ) and he wants 50 - 70 cm. For Scuba-dive use, maybe?
The lower rod can be directly inserted into the top shaft, that will get the length into the 90cm zone. He could make a new pip hole in the lower rod, that would let it be inserted a few cm further inside the upper shaft, the limit being where the control-box fits on. But he's going to need a shorter lower rod if he wants to get 70cm or less. Assuming he can buy a Minelab Equinox legally in his country, he can thus buy a new Minelab lower rod, which he can cut down how he likes to get the 70cm or shorter total.

Minelab distributor in Turkey:

https://www.minelab.com/usa/where-to-buy?search=dealer&country=206

This idea has been on my mind since I first got my Equinox. But recently I've investigated the calibration of the target-ID scale:

https://www.detectorprospector.com/forums/topic/18158-equinox-tid-value-target-frequency-calibration-table/

That has allowed me to make a more detailed post.

The current target ID scale has its mid-range '20'/'21' calibrated for a 6 kHz target. This seems reasonable for modes like Park1/Field1 , which place their emphasis on the 7.8 kHz operating frequency. The scaling of the rest of the numbers encompasses a very broad range, from targets way beyond US silver Dollars at the top end, to tiny nuggets at the bottom end. See the attached table detailing the current ID range.
I always felt the range was too compressed - targets reading above '30' ID are very rare for me, the machine seems more like it has a 30-point ID . But if you're the 'milled silver/copper coin hunter' , it's acceptable.

However, in the sparkier Park2/Field2 modes, the machine is clearly emphasising the 18.2 kHz operating frequency, great for European farmland hunting, looking for tiny ancient coins and other lower-conductor artefacts. But I think the detector needs the option of a recalibrated ID scale to match these low-conductor hunting modes. Centre-scale should be a target matching the 18.2 kHz, such as a US 5 cent 'nickel' coin ( a 16.6 kHz target ). The low ID range can be stretched a bit to give more resolution, and extended a bit to go 'below 1'. The high-conductor end of the scale can sacrifice a lot, and still ID many real common targets.

I propose a scale with a US 5 cent coin as ID = '20'. Scaling of individual ID steps is for a frequency ratio of 1 : 1.1067 ,so that 4 ID steps represents a 1 : 1.5 change in target frequency. To see how this fills the ID range, see the attached table.

I think the user should have the option of 'dual-scaling' the TID readout. If Park1/Field1/4kHz/5kHz/10kHz is selected, keep the existing ID scale. If Park2/Field2/15kHz/20kHz/40kHz/goldfield is selected, the nickel-centred 'Low-conductor' scale is used.
[I've neglected Beach-modes, as I'm not familiar enough with them to decide]

This choice of 'Standard ID' or 'Dual-scale ID' could be implemented using the (under-utilised) 'coil disconnected' mode. The machine would then remember how it was configured when subsequently operated normally. [ See this post for 'coil disconnected' mode thoughts:

https://www.detectorprospector.com/forums/topic/14250-what-features-and-performance-improvements-would-you-like-to-see-in-the-next-high-end-minelab-coin-detector/?do=findComment&comment=188022

 

An additional idea : The '88' TID readout can clearly indicate a value above 40, so it's possible to still ID large milled coins in the 'low-conductor' ID mode. The only real issue relates to the 'notch' , where '40' would notch '40 and above', and the feature allowing you to notch a particular target after you've swept over it would need a re-think.

"But how can we use this calibration table to more accurately ID targets?"

It's no use for that.

It has academic interest, and it removes the mystery of 'what the numbers mean'.

It allows you to in turn calibrate the ID scale of any other 'ID' detectors you have.

Many of you will be aware that the 'Target ID' number given by a detector is related to the 'Target frequency'. 'High frequency' targets read low-down on the ID scale, typically because they are small, thin, or made of a metal that's a poor electrical conductor. Conversely, 'low frequency' targets tend to be physically larger, and more likely to be composed of better conducting metals.
By careful measurement, it's possible to calibrate the ID scale of a given detector. I've done this for my Fisher F75 previously, it has a wide non-ferrous ID scale, from 16 -> 99, so it can give quite precise target freq. values.

I finally got round to properly working out the calibration of the Equinox 00 -> 40 scale. I have quite a varied selection of test targets that I've used for this calibration. Some of this as a result of work done on the Geotech1 forum, where both PI and VLF's were used for tests. The real physical targets include squares cut from aluminium drinks cans, coins, precisely-made copper rings. The best target is a 'synthetic' one: a coil of wire, with a selected resistor as a load. Knowing the inductance and resistance lets you calculate the time-constant / frequency.

Having taken a mountain of readings, it became apparent what Minelab have devised to create the scaling of the Equinox. It's actually very nicely done, and probably only possible as a result of using a high performance microprocessor. Each ID number step represents a fixed frequency multiple change from the previous number. This applies over the entire range from '02' up to '39' , covering a range from 80 kHz to 0.5 kHz. If you plot 'Target frequency' against 'ID number' on log/lin graph paper, you get a perfect straight-line graph.

Here is some nerdy maths for you:
The ratio from one step to the next is 1 : 1.1472 , so every 8 ID numbers represents a target freq change of 1:3 . An increasing ID number means a lower target frequency, (and a larger target time-constant).

Target time-constant and target frequency are related by:

2 * pi * Freq * time_constant = 1

which may be useful to PI users, nugget-hunters.

The designers obviously have to choose some particular ID value as their 'reference', and every other ID relates to this. It appears Minelabs engineers have chosen 'centre-scale' to roughly match the 7.8kHz operating frequency ( which dominates the Park1 / Field1 modes). A 7.8kHz target would actually read mid-way between '18' and '19', so technically they were 2 digits out, as it would ideally be 20/21.

( The numbers in the table are calculated. I chose ID ='20' as being a 25 microsecond target, and derived my figured from that.)

A few things to note:
This ID calibration is done for 'Park1' mode; 'Field1' should be the same. If you're using modes with a higher freq bias ( eg. Park2/Field2 ), the numbers can differ by a point, as you will have no doubt observed.

The frequencies shown are for the middle of that ID value range. So for example, TID = '16' actually encompasses targets from 10.3 kHz -> 11.8 kHz

Frequencies for ID values over '33' are estimated mathematically. I didn't have any useful test targets reading that high up the scale to verify the numbers.

To get a true measurement of a test target, you need to measure it at a test frequency close to that of the target. If you use a different measurement frequency, particularly one higher than the target, you can get errors due to 'skin effect' , where not all of the sample's metal is being measured. So, for example a US silver Dollar has a target freq of something like 0.85 kHz. But measure it with a typical 10 kHz detector, it looks like a 1.2 kHz target.

 

Buy an additional Minelab carbon lower rod. Cut down the length, make a new hole in it for the metal spring-pip, and fit it directly into the upper rod. 60cm total length should be easy to attain.

 

https://joanallen.co.uk/minelab-equinox-lower-shaft.html

As he's not replied with his Country location, I'll point out that the coins are Portuguese.

Settings issue? I got the impression he's asking:
"Why could I find plenty of targets before, but since the tides have made this ripple pattern, there are no targets to be found?"
I've never been beach detecting, but presumably if this ripple is caused by lots of deposited sand, then many targets have simply gone out of range?
There is no scale on the drawing, these ripples could be 60 centimetres high ( 24 inches for those Imperial users )

I think he is saying:
"The seabed is not flat, the tides/waves have made an undulating ripple pattern in the sand, like in this drawing.
What has caused this sand pattern, and is this why I cannot find any targets, now ? "

What complicates the business of Ground Balance on multi-freq machines is the obvious: they have two, three, four operating frequencies. Hence multiple ground balance figures.

So on the Eqx: Park1 and Park2 both use the same transmitted selection of freqs, but the 2 modes will GB differently. It seems clear that as well as giving more emphasis to certain freqs in normal operating mode, the GB value also has a bias towards that same frequency.

( In my opinion, Park1 will have its GB dominated by the 7.8kHz freq; Park2 will be dominated by the 18.2kHz freq. )

Strong iron/magnetite etc ground will likely give similar GB values regardless of which mode/freq profile is being used, a near-zero phase angle is still near-zero at a different frequency.

Basically, yes, you've interpreted the relationship correctly.

But it's only valid if you're lowering the coil to the ground, if you're lowering it from 20cm to 10cm, you're not doing it correctly.

It's a good question, because it then brings up the question "What is the audio signal indicating in ground balance mode??"

It's not simply indicating ground strength. If this were the case, lowering the coil to the ground would ALWAYS give an audio response. II guess it's pretty obvious that if the coil is 2cm above the ground, it's going to have strong ground pickup, and if it's 40cm above the ground, the pickup will be very much lower.

What's not so clear, is that the phase angle caused by the ground also changes as the coil is raised, and not in a particularly intuitive way. It actually changes towards 'Salt' as the coil is raised. So lowering the coil from, say, 15cm height down to a low level will cause an increase in ground strength, and a change in phase lag from, perhaps 25 degrees to 10 degrees.

If you're performing the 'pumping' procedure, ( groundgrab, fastgrab etc ) the micro brains can continuously read this signal, and work out that at the strongest point, that's the ground-balance value it has to calibrate to. But it can't use that value alone , it also needs to know what signal the coil gives when it's in free air, a long way from metal/ground/water. How it works that out, when you only raise the coil to 15/20cm high, I dunno.

I believe that ground with strong magnetite will still have electrically conductive minerals in it, it's just that the magnetite dominates the response to a detector, and whether it's also medium mineralised or hardly any minerals, or wet, or bone dry, has very little effect on the phase angle, it's going to be pretty close to Zero degrees, which would nominally be '00' on the Eqx, and '90' on the Fisher F75.

[[ mathematically, it's Vector addition:

There is a 'Zero Degrees' signal caused by the microscopic iron / rust / magnetite etc.

And a '90 Degrees' signal caused by minerals making up the ground, that conduct electricity, particularly when wet.

These two add together like vectors. The total strength would be calculated using the Pythagoras equation:

(Total strength)^2 = (Zero degrees)^2 + (90 degrees)^2

And the resulting Phase angle = arctan (90 degrees / Zero degrees )

The phase angle is the figure that is manipulated heavily to end up presented as 'Ground Balance', whether it goes 0 >99, or 90 to 0 , or anything else. ]]

Minelab chose the Equinox GB numbers to go the opposite direction to what Fisher typically used on their range, eg.  F75.

So 00 is strong magnetite -> 90's for saltwater.

I think the owners manual states this, if you download it and take a look.

( it's not that clear to be honest. And the fact that in 'beach' modes, the default GB value is '00' doesn't help matters ...)

To follow up:

Steve and I have conversed via PM about this.

To summarise:

The 'subscript' and 'superscript' options are no longer on the top menu. While there are keyboard shortcuts for italic ( 'ctrl+i' ), underline ( 'ctrl+u' ), bold ( 'ctrl+b' ) , there are no such shortcuts for subscript/superscript, so that workaround is not an option.

The 'Source' button is still present on the top menu. It's just my PC that doesn't show it, regardless of level of zoom in/out. Again, no keyboard shortcut exists for this function, because it's an editor plug-in, rather than a basic function.

 

 

100 US Dollars for an almost fully working pod ? It's got to be worth over 200 Dollars, surely?

Better use could be made of the situation where the machine is powered up with the coil disconnected.
It's good that the battery strength bar-graph functions in this 'mode' , I frequently check on state-of-charge, and appreciate not having to plug in the coil just to do this. So I think that the '88' display could usefully display more detailed battery info, such as the actual voltage: '40' == 4.0 Volts , or a 1->10 indicator.

In addition, it would be convenient to see what firmware version is in the machine, and the 88 display could usefully provide this info in this 'no coil' mode. Possibly in a 'code' form, the lack of decimal point makes displaying 1.3.2 etc troublesome.

"The Coil heartbeat indicator: there is concern that it will cause premature drain of the coil battery"

Speaking as an electronics engineer, the extra power drain caused by the LED, and associated electronics, is completely trivial. As there is no proper On/Off switch for the coil, there is inevitably going to be some drain, you just have to trust the designers to keep this to realistic levels. And there's no need for them to go to extreme measures, either, as the Li cell will self-discharge anyway.

----

I've not followed the D2 threads closely, but wasn't the pinpoint function considered inconvenient to use?

Anyone who has used trigger-operated pinpoint, such as the Teknetics T2, will know how intuitive and ergonomic it can be. So an over-complex system is undesirable.

My guess is that it wasn't commercially viable. Assuming it was 'Typical Deus' in operation, with integrated electronics in the coils, it would have retailed at a pretty high price, probably the same as, or more than, dedicated two-box machines ( C-Scope / Fisher / Whites ).

The concept of making a two-box add-on for a conventional machine is reasonable enough, Garrett did it. Normally this would mean just 2 coils and the hardware to hold it all, so it would be a realistic price. But it's a pointless concept on a Deus, as almost a complete new detector is needed.

(Since the closing of Whites, there is less choice now if you're after a two-box, but I still don't think this help XP sell their version.)

Quote: "So the signal strength increases but the dTID stays the same.  Good to know"

Maybe I should add some pictures to that post, to make it less 'wordy'.

If you feel experimental, try testing a 10mm ( 3/8" ) length of wire. Same result, I hope. The test samples we used were 19.0mm ( 3/4" ) long, for consistency. When the wire gets thin, it's not easy getting a strong enough signal to measure properly. So two alternative approaches were tried : multiple 19mm lengths, or a single longer length, both worked fine.

Here's an example of a real-world target giving the 'wire' problem:

this is the fixing pin off a Roman bow brooch, made of bronze, found by a guy on a UK forum. He saw it sticking out the soil, ran his French stick over it, to be met with silence. Only after some fiddling with the settings and pretty much scraping the coil over it did he get a poor beep, with a super-low ID value. The two problems here: the diameter ( about 1.7mm estimate ) and the metal : bronze, AND corroded. Now if it were nice pure copper, it could ID in the 5c 'nickel' range. But the unknown metal conductivity is going to be anywhere from 1/5th to 1/20th of that. This pushes up the target freq to 100 kHz, 200+ kHz, so even an 18kHz machine will struggle to see it, in motion-mode, anyway.

Ahh, better data, thanks.

Your measurements are in good agreement with the scientific tests we did ( on the Geotech1 forum ).

I calculate  ( with Eqx TID )

2.0mm : freq = 13.5 kHz, TID = 14

2.2mm : freq = 11.3 kHz, TID = 16

3.2mm : freq = 6.1 kHz, TID = 20

There's a bit of leeway allowed, I'm converting from F75 readings to Eqx ones, I never got round to 'calibrating' the Eqx ID's, in part because they covered a rather small useful range.

In case anyone's slightly interested, here is the maths:

We found the target time-constant was proportional to the cross-sectional area of the wire, and proportional to the electrical conductivity of the metal, and the result was this formula:

TC = 0.029 * D * D * %IACS

where TC = time-constant ( in microsecs ); D = wire diameter ( mm );  %IACS = metal conductivity.

Example: 2mm diameter copper, TC = 0.029 * 4 * 102 = 11.8 usec

To convert time-constant into target frequency use this:

target freq, f = 1/ ( 2* pi * TC ) with TC in seconds

using the above TC as an example, this gives 13.5 kHz

It's a good straight-line match for wires up to 2.5mm, then it starts to drift away a bit , due to skin effect; the full diameter of the wire is not all seen. PI machines don't see this so badly.

But regarding: "Not surprisingly the in-field dTID's vary with length and shape"

I think you may have missed the main point of my earlier post : TID does NOT vary with wire length, unless you have very short lengths. It can depend a bit on shape, irregular objects commonly give erratic ID's.

 

Quote: "I find quite a bit of copper wire -- single stranded.  The gauge is in the 12-16 range.  They typically hit in the USA zinc penny and aluminum screw cap range."

Assuming that's American Wire Gauge, AWG, that is 2.05mm to 1.29mm. Mathematics tells me that would be the target frequency range 13 kHz to 34 kHz, respectively. On the ID scale, 13 kHz is above US 5c coins ( 16.5 kHz) but way below 'zinc cent' ( 5 kHz ) , and 34 kHz is down in the foil zone . So I'm unsure why you're getting your wire come in at '5 kHz'.

Re: the Cu-Ag system, I found similar data on this page, for high-silver alloys up to the eutectic 28% Cu mix, including a graph that shows that heat-treatment plays a part in the behaviour:

https://www.electrical-contacts-wiki.com/index.php?title=Silver_Based_Materials

Another non-intuitive target is the long-and-thin type. Such items include wire ( copper or brass, typically ), plain finger-rings that are broken ( C-shape ), copper/bronze/aluminium nails.

Metal detectors induce circulating electrical currents in the target that tend to form loops as large as they can, without straying too far from 'circular'. So for a wire, these current loops are little more than the wire diameter in size, they don't go up-and-down the length of the wire at all. There are many of these loops along the wire, so they all contribute to giving a stronger response to the detector, but the 'target frequency' is still that of a very short snippet of wire, no more than twice the diameter.
So wires tend to have very a high 'target frequency' , that's independant of their length. This can make them hard to detect.
An extreme example : among my electronics junk, I have some test coils that are much like the windings inside a detector search-coil. 100 turns of 0.2mm enamelled copper wire, over 20 grams of highly-conductive metal in total. But place one 5mm from a detector coil, it is invisible. The target freq is about 2.5 MHz, so it's not going to match any commercial machine.

Copper and silver seem pretty compatible, maybe the alloy curve is rather dull, hence not worth studying?

The reference I used for gold-copper , and nickel-copper alloys is this weighty pdf:

https://srd.nist.gov/JPCRD/jpcrd221.pdf

is that the same as the one just posted .. ah yes, just the filesize is different, making me question it.