New commercial design

[ivconic]

How many people on this planet already are having smart phone and/or tablet?
So... those already bought "half" of the detector, right?

[Silver Dollar]

Here's how to run Android on a windows PC. http://windowssecrets.com/newsletter...-a-windows-pc/

The big problem is resources, if you do not have a target system you get into
support problems. IE one system has a 640x480 screen and another 768x1024
your software has to allow for this....

[ivconic]

Here's how to run Android on a windows PC. http://windowssecrets.com/newsletter...-a-windows-pc/

The big problem is resources, if you do not have a target system you get into
support problems. IE one system has a 640x480 screen and another 768x1024
your software has to allow for this....

Latest adt-bundle (Eclipse and Java SDK) overcomes all the past issues we met with earlier versions.
It has also very powerful and nice virtual device included.
I also completed it's APIs database starting from API 8 to latest API19.
I don't know exactly how much it took, but i guess over 10GB overall.
Don't really care much; my drive is 1TB now!
Before that i faced huge disappointment once i realized that my beloved Siemens (Intel D 3GHz) was not capable to run such virtual device in real time...
It was shock!
So i had to spend extra 800e on buying i7 laptop with top virtualization-capable features to be able to run virtual device in real time.
Just for rough comparison: it takes 22 minutes for Siemens to start virtual device and later it runs slowly with lot of problems, while it
takes ~10 seconds for i7 to start same virtual device and later runs in real time with no problems at all.
But these facts are completely irrelevant for end user because these are important only for code developer.
If you wish to become code developer; than be ready to spend decent sum of money on requirements, but if you wish to only enjoy in such technology than don't worry;
it will take you to invest just small fee overall.
Android & iOS machines are cheap.
Additional PC software will serve there only to further process already logged md data.
For example; GPS logged paths and points later can be nicely displayed on PC software in Google Earth, Maps etc... put in large SQL database and stored or shared... etc.
It will not be expensive for end user at all.
It is demanding only for code developers.

[ivconic]

There are also few alternatives on replacing virtual device from Eclipse with some prepared Android image.
Those are alternatives for someone who is not ready to invest additional money in equipment.
I tried such alternative but i don't like it because of certain limitations.
For serious intentions later will turn that those alternatives are not good enough.
Cheapest solutions are actually always the most expensive one!
Because of the tradeoffs in between...

[WM6]

Android & iOS machines are cheap.

Unfortunately still equipped with very easy breakable screen-glasses - even Gorilla Glass (scratch resistant, but not break resistant).
Mine was fall out of the pocket, when I bent down, from about 0.5m height and broken (got repair info: 76 euro - cheap?).
Not really suitable for hard terrain conditions.

[ivconic]

Unfortunately still equipped with very easy breakable screen-glasses - even Gorilla Glass (scratch resistant, but not break resistant).
Mine was fall out of the pocket, when I bent down, from about 0.5m height and broken (got repair info: 76 euro - cheap?).
Not really suitable for hard terrain conditions.

When i say "cheap" ; i mean it will be really cheap in such Metal Detector setup because it's overall price will be some 1/5 of prices we seen today on metal detectors market.
For example take Deus package price... or Minelab 3030 package price !
Android based metal detector will have much lower price and much better performances including endless list of various "jingle&bells".
I will rather replace Android part of my metal detector each month or two (anyway that's common reality, we already tend to follow new tech) than to stuck with limited Deus or 3030 abilities for good.
Search coil for such setup may remain the same in period of time, while Android & iOS part of such metal detector will always follow latest trends and fashion.
If you take things from this perspective; you must agree with me that this is indeed much cheaper solution, right?
Besides, after all: nothing is unbreakable in such technology. You can also easy break Deus as well as 3030... if you have such bad luck, right?
What's cheaper: to spend extra 620 euros on broken Deus's Remote ... or to spend extra 100 euros on broken Android? Ha!?

[Davor]

That would also depend upon the complexity of processing required by the android. If you supply it with raw signal and do all DSP in android, a choice of devices becomes problematic. OK, not for long, but serious number crunching also devours batteries, both in processing and wireless communication parts. If you do the opposite, a probe electronics have to do the crunching and battery devouring, while android gets the easy part. Perhaps it is the preferred path because you may wish to do many other things with android, and also you don't want to empty it's battery too soon.

It would also give you more room for integrating all the other sensors with the android application. There are lots of bells and whistles to play with.

One thing also came to my mind, there is also a NTRIP signal available in many countries that have a network of DGPS sites, and for a serious GPS path recording it would be a serious bonus because it improves precision to under 50cm of error in real time. There are drawbacks... in Croatia you have to pay "CROPOS" annual fee exceeding 100US$. Guess making your own DGPS receiver and transmitting your own correction data should be cheaper than that in 2 years or so. (oops, I see the next patent on the way)

The only NTRIP android client that I know of is https://play.google.com/store/apps/d...re.ntripclient
There is no such application for apple, as it relies on BT GPS, and apple does not support SSP.
Perhaps making a NTRIP application that uses internal GPS would make it useful for apple, and make a whole application more compact. But then again, apple does not have SSP to connect to a probe, so it seem a dry branch already.

[ivconic]

That would also depend upon the complexity of processing required by the android. If you supply it with raw signal and do all DSP in android, a choice of devices becomes problematic. OK, not for long, but serious number crunching also devours batteries, both in processing and wireless communication parts. If you do the opposite, a probe electronics have to do the crunching and battery devouring, while android gets the easy part. Perhaps it is the preferred path because you may wish to do many other things with android, and also you don't want to empty it's battery too soon.

It would also give you more room for integrating all the other sensors with the android application. There are lots of bells and whistles to play with.

One thing also came to my mind, there is also a NTRIP signal available in many countries that have a network of DGPS sites, and for a serious GPS path recording it would be a serious bonus because it improves precision to under 50cm of error in real time. There are drawbacks... in Croatia you have to pay "CROPOS" annual fee exceeding 100US$. Guess making your own DGPS receiver and transmitting your own correction data should be cheaper than that in 2 years or so. (oops, I see the next patent on the way)

The only NTRIP android client that I know of is https://play.google.com/store/apps/d...re.ntripclient
There is no such application for apple, as it relies on BT GPS, and apple does not support SSP.
Perhaps making a NTRIP application that uses internal GPS would make it useful for apple, and make a whole application more compact. But then again, apple does not have SSP to connect to a probe, so it seem a dry branch already.

Things are moving very fast in Android&iOS world, your points are correct but will be overcame soon.
Even "yesterday" i was delighted with my first Archos running on 2.3 and now i am watching it with huge "contempt"!
It cost me 150 euros than and now you can buy "beast" for such money.
Battery issue is serious issue and must be taken with lot of attention in final solutions.
There would be always certain tradeoffs.
I would rather pick solution with "stupid" coil and "smart" application, because things are moving fast and now at this moment we have multi cored platforms for not so large amounts.
I am playing right now with Infinity pad which is high end model and certainly way over than being capable for tasks which Deus usually is processing.
Multi cored, amoled gorilla, full hd, Nvidia Tegra 3, Sonic Master audio... 64Gb memory, 14 hours of mobile computing...
Let's analyze one typical day outdoor with metal detector.
I am spending all day long outdoor with Deus... but it runs effectively barely 4-5 hours, even those in rare occasions.
But ok, let's take that 5 hours is some standard.
Hopefully majority of modern Android & iOS devices will meet such demand, especially if BT is not using standard protocols.
On other side, search coil may run on 1860 battery with 4200mAh capacity which is hopefully quite enough.
Will see. Experiments are in progress.
Speaking of GPS accuracy... well that's beyond our limits, we have to deal with what we have available so far.
That's why i involved functionality as "Camera snapshot" to conjoin it with "GPS snapshot" data by free will of end user.
There are places and situations where you don't need exact accurate GPS location, additional camera snapshot will do just fine.
Though; i wouldn't complain on more accurate GPS data, of course!

[sled]

The battery drain could be reduced by using headphones and a good multitone signal. Maybe one could make a small external controller accessoire that wraps around your arm with a few buttons for basic adjustments so your phone is secure in your pocket all the time and you can adjust the Disc/Sens/GB on the go.

Or go one step further and do a VDI on a smart watch

The battery drain is hard to predict, however there are small external battery packs that can be connected to recharge the phone.

The big problem with doing everything in a standalone unit is that the overhead in hardware/software explodes and time to market will be much longer. The CTX3030 uses an ARM9 processor running linux on it and the software is actually written using Qt, so it's a desktop application running on an embedded device - they could have saved all this effort by running it on a smartphone, spend more time on things that matter instead of reinventing the wheel

[sled]

@ivconic: some time ago I did a little bit of reverse engineering on the CTX3030 firmware, I have posted the internal sqlite database structure here: http://www.geotech1.com/forums/showt...398#post187398

Maybe it can serve as "inspiration"

[ivconic]

The battery drain could be reduced by using headphones and a good multitone signal. Maybe one could make a small external controller accessoire that wraps around your arm with a few buttons for basic adjustments so your phone is secure in your pocket all the time and you can adjust the Disc/Sens/GB on the go.

Or go one step further and do a VDI on a smart watch

The battery drain is hard to predict, however there are small external battery packs that can be connected to recharge the phone.

The big problem with doing everything in a standalone unit is that the overhead in hardware/software explodes and time to market will be much longer. The CTX3030 uses an ARM9 processor running linux on it and the software is actually written using Qt, so it's a desktop application running on an embedded device - they could have saved all this effort by running it on a smartphone, spend more time on things that matter instead of reinventing the wheel

Exactly!
One of the reasons is also: money!
They can't earn filthy huge pile of money using smart phone GUI instead.

[ivconic]

@ivconic: some time ago I did a little bit of reverse engineering on the CTX3030 firmware, I have posted the internal sqlite database structure here: http://www.geotech1.com/forums/showt...398#post187398

Maybe it can serve as "inspiration"

Yes that's script dealing with some database elements.
It is easy to define elements for such database, guys from Minelab didn't done anything spectacular.
...
Initially i was thinking also to use some cheap embedded system in coil.
But power drain would be than even more emphasized issue.
( just look 3030 battery size !!!!)
No need for that.

[sled]

how do you plan to shape the filters and implement the algorithm? If one could pair the coil with a laptop one could use GNU Octave/Matlab to work on real measurements and quickly test it with GNU Radio on a live data stream

[ivconic]

how do you plan to shape the filters and implement the algorithm? If one could pair the coil with a laptop one could use GNU Octave/Matlab to work on real measurements and quickly test it with GNU Radio on a live data stream

That's the "prohibited" area man! No trespassing!
...
Yes, if it will pair with Android & iOS; it will be able to pair with Windows too, of course.
Your imagination is only limit in this, see?

[WM6]

Wireless is nice idea, but till we need shaft to move coil it is not a big improvement. Why wireless only? Why not shaft-less too?

Why not without those bulky shaft?

Why not sitting down and play with our smart-phone and waiting for signal from our self-moving robotic coil?

Who need shaft?

GPS in coil make sense only if we have robotic coil. Best submersible one.

So here project upgrade proposal to Robotic Coil: