complete project

[Tinkerer]

Hi Tinkerer,
Sorry for the delay in replying, I had to nip away for a week.
I don't want to hijack this design I just realised that if you were tailoring it towards old coinage and artifacts as per satdaveuk's suggestion, then I could adapt it for my needs.

The type of country I hunt, is somewhat like Northwest European Roman sites, having lots of tussock grasses and peat that reduce depth markedly. I'll link to some photos when flickr stops having hiccups.

I expect I would need a coil size of around 60 cms able to detect a 10 gram nugget (20mm coin) at about 45 cms. The Minelab GPX4500 will easily detect such a target at 40 cms and deeper with a 45 cm coil. The problem with the Minelab is its weight and poor ergonomics. Hence my original thoughts of miniaturizing Tinkerer's PI (SMD, 4-6 layer PCB) as many of my sites are many miles from roads well away in the mountains, and I need to lug enough power for up to 40 hours detecting. A flexible roll-up solar panel is an option in summer, but not in spring or autumn.

Showerproof would be ideal as mountain weather is unpredictable, however the possibility of a waterproof fresh water discriminating machine could open up vast areas.

As previous posters have said on this thread, you have already developed a detector already that has such potential, it exists across many posts and threads in the form of developmental blocks. The problem is deciding which blocks to collect and assemble to get a working optimised machine. I think the Guys, myself included were very interested in your assembling such blocks in one thread as a complete design. However, the potential of a TEM solution could over shadow all your previous work, and so I don't want to discourage you from proceeding down this course instead.

I am interested in your 12V front-end, since I will need to go with a medium power option. I suspect that using the TEM system will actually produce results only achievable with much higher powered front-ends, do you agree?

Cheers
Kev.

Hi Kev,
Your situation presents an interesting challenge.
Let’s take a closer look and see if we can understand the problems.
I say WE, because I hope that others who have similar interests and problems might join in to help finding solutions. Many minds= many ideas= many solutions.
I am adding my comments between your lines. I will take some guesses, please correct me when I am wrong.
Hi Tinkerer,
Sorry for the delay in replying, I had to nip away for a week.
I don't want to hijack this design I just realised that if you were tailoring it towards old coinage and artifacts as per satdaveuk's suggestion, then I could adapt it for my needs.

The type of country I hunt, is somewhat like Northwest European Roman sites, having lots of tussock grasses and peat that reduce depth markedly. I'll link to some photos when flickr stops having hiccups.
Peat, why would that reduce the depth? It is not mineralized. Could it be that it is conductive? I think it is acidic. We might consider it an electrolyte. What is the underlying rock?
Is the ground response fairly constant? Or is it highly erratic?
Tussock grass, I guess you have to walk around the tussocks and sweep the detector around the base of the tussock? This makes the actual sweeping motion difficult.
Would it help if the motion could be slow, like 0.5m/s? Is it possible to maintain a steady motion of the coil?
Motion speed control.

I expect I would need a coil size of around 60 cms able to detect a 10 gram nugget (20mm coin) at about 45 cms. The Minelab GPX4500 will easily detect such a target at 40 cms and deeper with a 45 cm coil. The problem with the Minelab is its weight and poor ergonomics. Hence my original thoughts of miniaturizing Tinkerer's PI (SMD, 4-6 layer PCB) as many of my sites are many miles from roads well away in the mountains, and I need to lug enough power for up to 40 hours detecting. A flexible roll-up solar panel is an option in summer, but not in spring or autumn.
I don’t think that miniaturizing the PCB will change much to the weight problem. Also a 60cm coil will most likely be more heavy than a 40cm coil.
Personally I lean towards SMD in general, but I still have a lot of through hole parts to use up. Just a few years ago, I thought that the SMD will be the end of my hobby. These things are so small I can’t see them without glasses. But now I need a large magnifier for everything, so it does not make such a big difference anymore.
The ergonomics. Could you make some simple sketches of your ideas of ergonomic improvements?

Showerproof would be ideal as mountain weather is unpredictable, however the possibility of a waterproof fresh water discriminating machine could open up vast areas.

Would the freshwater machine be used for diving? How deep?
I used to build my own detectors and housings for recovering treasure from the bottom of the oceans, so I know what is involved.
Waterproof controls are expensive. This means we would try to have a minimum of controls. Make it extreme simple for a start.

As previous posters have said on this thread, you have already developed a detector already that has such potential, it exists across many posts and threads in the form of developmental blocks. The problem is deciding which blocks to collect and assemble to get a working optimised machine. I think the Guys, myself included were very interested in your assembling such blocks in one thread as a complete design. However, the potential of a TEM solution could over shadow all your previous work, and so I don't want to discourage you from proceeding down this course instead.
Some of my earlier ideas were not bad, but in general, there is constant evolution, better understanding and accumulated knowledge, so I really recommend going with the latest design.

I am interested in your 12V front-end, since I will need to go with a medium power option. I suspect that using the TEM system will actually produce results only achievable with much higher powered front-ends, do you agree?
The TEM method is very power efficient. Until now I have not tried to cut down on the power consumption, instead I have been trying the increase dept and the signal response amplitude and S/N, this ay eliminating a lot of problems that come with very high amplification.
I am working on a LP (low power) Lite solution. This might just be what you need.
Could you give some more information about your battery ideas? Let’s say we chose 12 batteries. Would you use LI-Ion? What is the maximum weight of batteries you would consider OK?
Tinkerer

[Kev]

Hi Tinkerer,
Here are some images of the country I detect in mainly. I also like river gorges, which is similar, large obstacles, deep & shallow overburden, damp and unpredictable mountain weather etc.,

Mountain camp in auriferous country


Thick grasses, tussocks and peat mean that a deep seeking detector is required (don't look at my wife using a Gold Bug VLF, actually she did OK with it in the rocky places, nice and light for the female enthusiast)


The quarry, a quarter oz found on this particular trip.




I meant to say that peat usually means that basement will be fairly deep so a deep seeking machine is required, especially when a thick mantle of grass overlies it.

I use the slow motion filters on my Minelab at present in this area, as you do need to poke the coil in and about the grasses, it achieves much better depth this way too.

The main thoughts on miniaturization, is that fact that all the gear has to be lugged in on ones back, so the less area and weight the control box etc., consumes the more space is available for food

The rivers and lake shores would give up many troy pounds within a couple of metres of the surface, that's all that would be required. I'd even consider using a pelican case and just opening it up and making adjustments before entering the water, or wading up a gorge channel.

Presently I'm using hip mount Li-ion batteries on my Minelab, it's called a Pocket Rocket system produced by Coiltek www.docsdetecting.com/lithiumionbp.html
I can purchase extra batteries locally for about $35 US they are listed as being 6.6 AH when new and each one will run my GPX4500 for about 6 hours, 5 after a couple of years.

I recently stumbled upon Dave Emery's patented Resonance PI. I was wondering if the TEM system uses a similar approach as far as Rx is concerned? I know Dave doesn't use Caps in the TX and the TEM coil isn't in a balanced configuration, at least that's what I remember, could be wrong. here's a link to the pdf. http://www.google.com/patents/US20090045813.pdf

Sorry for the delay again, busy with quake repairs to the house, I'm situated in Christchurch New Zealand. We've had a couple of good shakes over the last couple of years.

Best regards
Kev.

[6666]

G'day Kev
that sure is spectacular countryside you get to detect in
we have rellies in Christchurch, they have been ordered out of their home because of the quake damage, yet some houses in the street are undamaged.

[Davor]

I recently stumbled upon Dave Emery's patented Resonance PI. I was wondering if the TEM system uses a similar approach as far as Rx is concerned? I know Dave doesn't use Caps in the TX and the TEM coil isn't in a balanced configuration, at least that's what I remember, could be wrong. here's a link to the pdf. http://www.google.com/patents/US20090045813.pdf

One of my musings here on this forum is that PI Rx is not nearly efficient as it could be, and waiting for the perfect sampling moment is a main symptom of it. My thought is that some weighting function could squeeze much more of the target information than a mere exponential curve can. This resonant ... thing is in effect an energy conserving weighting function. However, I'm not convinced that (by looking to the patent) it is possible to increase depth by much with this particular solution - instead of exponential fall we have an exponentially fading sine.

[Kev]

G'day Kev
that sure is spectacular countryside you get to detect in
we have rellies in Christchurch, they have been ordered out of their home because of the quake damage, yet some houses in the street are undamaged.

Gid'day 666,
Yep, that's the way it went, all depends upon what lies beneath your house 10s of metres down. When the first Europeans arrived here way back, they filled in a swamp and built the city upon it, as it was the closest available land to the deep water port. Sadly if your house was over one of these subterranean streams or springs then water boiled up during a large quake or aftershock, and the ground turned to quick sand. Have you ever pattered the sand at the beach near the water line? You'll have a picture of liquefaction.

We get away to the hills as often as we can and leave all the stress of it behind us

I'm not convinced that (by looking to the patent) it is possible to increase depth by much with this particular solution - instead of exponential fall we have an exponentially fading sine.

You could be right, but the way I see it, a ringing coil and by extension the field, is going to have a much greater dV/dt than a critically dampened one. I imagine it wrenching the target hither an thither, wringing out the greatest possible response from it. Dave doesn't mention it in the patent, but at the time he was working on this there were a lot of experiments being done with constant current coil sources, he mentions in the patent high current pulses, perhaps he is using this typology?

What is really impressive is the full depth Fe discrimination.

Tinkerer is on the right track I think, maybe achieving the same results but with a public domain solution, I hope to have some time soon to experiment with his generously published findings.

Cheers
Kev.

[Davor]

I imagine it wrenching the target hither an thither, wringing out the greatest possible response from it.

Unless I'm terribly wrong, this detector does not do such thing at all. Instead it causes a Rx tank to ring in case of a target being underneath it. A ringing Rx coil is in a way a good thing because at start sin(x) ~ x (or a ramp) that makes for much lower initial voltages than traditional PI, while return pulse* energy is being conserved.

* In reality there is no such thing as return pulse at LF, just the energy being transformed by a Tesla - type transformer from Tx to target and back to Rx.

There is not much you can do about this transformation efficiency, except maybe building a better coil, but there just might be some approaches that may improve Rx sensitivity. I think applying a weighting function to Rx frontend may be that answer. I strongly believe that the ramp action of sin(x) ~ x could perform much better if only a first half sine is examined - even if it takes a lower resonance than Dave's solution.

[Kev]

Unless I'm terribly wrong, this detector does not do such thing at all.

Yep I think you're right, and I am getting the TEM and Dave's RPI Tx all mixed up.
Burning the candle at both ends at the moment, and the cider it good this year

Cheers
Kev.

[Kev]

Hi,
I can see clearly now........
What I was trying to put before Tinkerer with this patent, but didn't elucidate sufficiently, was a combination of TEM TX with a resonant receiver as Dave has done. Something not feasible with conventional VLF due to noise issues, but perhaps achievable with RPI. The tuned Rx coil may give you your weighting factor due to the transformer function.......Just rambling on about impossibilities probably, I must build some circuits and dabble.

Cheers
Kev.

[Tinkerer]

Hi,
I can see clearly now........
What I was trying to put before Tinkerer with this patent, but didn't elucidate sufficiently, was a combination of TEM TX with a resonant receiver as Dave has done. Something not feasible with conventional VLF due to noise issues, but perhaps achievable with RPI. The tuned Rx coil may give you your weighting factor due to the transformer function.......Just rambling on about impossibilities probably, I must build some circuits and dabble.

Cheers
Kev.

Hi everybody,

sorry I can not answer to all the posts right now, but definitely will as soon as I can find the time.
About Dave Emery's patent, I have been aware of that method quite some time before it was patented, but decided to go another way, which I believe is better.

Tinkerer

[Tinkerer]

Yep I think you're right, and I am getting the TEM and Dave's RPI Tx all mixed up.
Burning the candle at both ends at the moment, and the cider it good this year

Cheers
Kev.

Hi Kev,

thanks for the nice pictures. I wish I could be searching for nuggets there myself. I got a better understanding of your problems now and can see the challenge to find solutions.

One more question: do you have a lot of ferrous clutter in these areas? Would good iron discrimination be a help?

It seems to me that it is a real problem to swing a large coil in this environment.
This leaves us with 2 possibilities:
Design a detector for slow swing speed, something like 0.5m/s, or
Design a non motion detector.

Tinkerer

[Tinkerer]

Presently I'm using hip mount Li-ion batteries on my Minelab, it's called a Pocket Rocket system produced by Coiltek www.docsdetecting.com/lithiumionbp.html
I can purchase extra batteries locally for about $35 US they are listed as being 6.6 AH when new and each one will run my GPX4500 for about 6 hours, 5 after a couple of years.


About the battery:

The TEM TX method makes very good use of the power. Therefore I like to use a high power TX at a high pulse repetition rate, like 10A peak coil current at 5000 PPS.

This consumes about 600mA out of a 12V battery.

It has the advantage of a powerful TX pulse that generates strong eddy currents in deep targets. The high repetition rate allows for integrating or stacking many samples, thus giving a good S/N.

For this, it is good to have a relatively high TX voltage, 12V is OK, more is even better. You mentioned 12V earlier, so I have started designing around this voltage, but it seems that the battery pack you mention above, is only 7.3V.

Comments?

Tinkerer

[Midas]

Presently I'm using hip mount Li-ion batteries on my Minelab, it's called a Pocket Rocket system produced by Coiltek www.docsdetecting.com/lithiumionbp.html
I can purchase extra batteries locally for about $35 US they are listed as being 6.6 AH when new and each one will run my GPX4500 for about 6 hours, 5 after a couple of years.


About the battery:

The TEM TX method makes very good use of the power. Therefore I like to use a high power TX at a high pulse repetition rate, like 10A peak coil current at 5000 PPS.

This consumes about 600mA out of a 12V battery.

It has the advantage of a powerful TX pulse that generates strong eddy currents in deep targets. The high repetition rate allows for integrating or stacking many samples, thus giving a good S/N.

For this, it is good to have a relatively high TX voltage, 12V is OK, more is even better. You mentioned 12V earlier, so I have started designing around this voltage, but it seems that the battery pack you mention above, is only 7.3V.

Comments?

Tinkerer

Hi Tinkerer,

That 600ma figure you've quoted is that measured sitting on the bench or have you done some field testing? Because if its only a bench figure I don't think you can assume its going to stay constant in the real world. When you place the coil over the ground you will lose efficiency, and exactly how much needs to be determined. It could be very significant, for example lets say your 98% efficient on the bench and you drop to 90% efficiency in the field. Now you need 5 times more power to maintain the same coil current. Maybe those figures are overly pessimistic but still, I think its worth looking into.

Midas

[RKC]

Hi Tinkerer,
Here are some images of the country I detect in mainly. I also like river gorges, which is similar, large obstacles, deep & shallow overburden, damp and unpredictable mountain weather etc.,

Mountain camp in auriferous country


Thick grasses, tussocks and peat mean that a deep seeking detector is required (don't look at my wife using a Gold Bug VLF, actually she did OK with it in the rocky places, nice and light for the female enthusiast)


The quarry, a quarter oz found on this particular trip.




I meant to say that peat usually means that basement will be fairly deep so a deep seeking machine is required, especially when a thick mantle of grass overlies it.

I use the slow motion filters on my Minelab at present in this area, as you do need to poke the coil in and about the grasses, it achieves much better depth this way too.

The main thoughts on miniaturization, is that fact that all the gear has to be lugged in on ones back, so the less area and weight the control box etc., consumes the more space is available for food

The rivers and lake shores would give up many troy pounds within a couple of metres of the surface, that's all that would be required. I'd even consider using a pelican case and just opening it up and making adjustments before entering the water, or wading up a gorge channel.

Presently I'm using hip mount Li-ion batteries on my Minelab, it's called a Pocket Rocket system produced by Coiltek www.docsdetecting.com/lithiumionbp.html
I can purchase extra batteries locally for about $35 US they are listed as being 6.6 AH when new and each one will run my GPX4500 for about 6 hours, 5 after a couple of years.

I recently stumbled upon Dave Emery's patented Resonance PI. I was wondering if the TEM system uses a similar approach as far as Rx is concerned? I know Dave doesn't use Caps in the TX and the TEM coil isn't in a balanced configuration, at least that's what I remember, could be wrong. here's a link to the pdf. http://www.google.com/patents/US20090045813.pdf

Sorry for the delay again, busy with quake repairs to the house, I'm situated in Christchurch New Zealand. We've had a couple of good shakes over the last couple of years.

Best regards
Kev.

G'day Kev,

Good to see you have been doing well down in Otago ... and managing to get away from the city! Enjoyed seeing the pictures of the tussock country (I think I recognise that part of the country). I can't seem to manage to get down that way these days but might make a trip next summer.

Hope your house repairs are going ahead without too many problems. With Christchurch still getting shakes it must be a nightmare trying to repair a house.

Regards,
Rob (RKC)

[Tinkerer]

Hi Tinkerer,

That 600ma figure you've quoted is that measured sitting on the bench or have you done some field testing? Because if its only a bench figure I don't think you can assume its going to stay constant in the real world. When you place the coil over the ground you will lose efficiency, and exactly how much needs to be determined. It could be very significant, for example lets say your 98% efficient on the bench and you drop to 90% efficiency in the field. Now you need 5 times more power to maintain the same coil current. Maybe those figures are overly pessimistic but still, I think its worth looking into.

Midas

You are right, there is a big difference between the calculated or simulated efficiency and the real power consumption with the coil sitting on a conductive reactive ground.

In this case, the simulation shows about 350mA power consumption, while, with the coil sitting 10cm above a steel mesh, the consumption is near 600mA.

I feel 600mA is an acceptable figure for a deep seeking PI, but would love to hear other opinions too.

Tinkerer

[Mechanic]

Hi Tinkerer,

What effect does putting the coil near, say a sheet of alu foil have on efficiency?

600mA is fine for power draw, thats about what the ML's use

Cheers Mick