Let's make a closely MXT like detector!

[KRinAZ]

I was surprised when I found a schematic in my archive! I don't remember where I got it from.

Wow, and I am beyond surprised - I didn't think we'd be so lucky as to have the schematic! Nice job, well done, thanks for posting to you and it's source. It's a wealth of information and ideas for this project. Now a zip of the C would make the complete set haha...

Earlier this evening I ordered Carl's ITMDv3, and George's Arduino Nano VLF Project books from amazon and look forward to absorbing them and applying them to this project. They should arrive Fri 4/17. I plan this week to open up my MXT and see what rev board I have and which micro. I'm also working on acquiring a MXT Pro to dissect, so I'll be doing research and posting insights.

If I find that the MXT Pro has a newer PIC16Fxxx with flash program storage I think I might know how to pull the code...would be a great source of examples and ideas...

[ivconic]

MXT & GMT, approx 60-70% of the layouts are quite the same.

[ivconic]

Or like this:

[ivconic]

I was surprised when I found a schematic in my archive! I don't remember where I got it from.

[KRinAZ]

The MXT Pro has identical performance to the MXT, it just adds tone ID, ground grab, and a backlight. The GMT was cloned because it had huge sales in Africa, the MXT was never cloned. I'll dig around and see if I can find what micro the Pro used.

If it helps any, in that pic that ivconic posted of a MXT PCB (probably a later revision PCB) I'm guessing it to have a 28 pin PDIP MicroChip PIC16F886 - but not sure as it's number stamp is not clear when zooming in...pretty sure ivconic found the pic on the web, and so probably doesn't have clearer pics...

[KRinAZ]

The MXT Pro has identical performance to the MXT, it just adds tone ID, ground grab, and a backlight. The GMT was cloned because it had huge sales in Africa, the MXT was never cloned. I'll dig around and see if I can find what micro the Pro used.

Very good, many thx for the Pro info, hope you can find the micro number.

It's interesting that Dave Johnson coded in assembly, didn't know that. I often felt like the odd one using assembly but I guess micro coders probably did so to keep their code as tight and small as possible. It can be hard to fit code and all the features desired into 8k or so of micro code space...

I have another application for this project - a sort of hand held version - hand held coil with ergo hand grip (maybe with display) on the 4x6 DD Shooter, control box on the belt, useful for scanning hard rock gold mine walls & ceiling. I have 2 tunnels and numerous cuts that I want to scan...a full size detector has proven quite awkward for this.

[Carl-NC]

The MXT Pro has identical performance to the MXT, it just adds tone ID, ground grab, and a backlight. The GMT was cloned because it had huge sales in Africa, the MXT was never cloned. I'll dig around and see if I can find what micro the Pro used.

[KRinAZ]

The original MXT board used an OTP chip that was read-protected. It was updated when I worked there to use a later version chip (for the MXT Pro) that could be reprogrammed, but it was still read-protected. It is possible to defeat the read protection and extract the code but it's a complicated process. An easy solution is to send the board to China and one of the hacker houses will send you back the extracted binary, but it costs around $500. Might be able to find someone who will do it cheaper. The binary can be decompiled to assembly language and then maybe AI can clean it up and even convert to C.

The Chinese cloned the GMT which uses almost the exact same circuitry and mostly the same code as the MXT. It's possible the GMT clones are not protected. In any case, they are a lot cheaper to buy and cannabalize. But if you want to extract the code with them most features, it would be the MXT Pro or the MX Sport.

The original MXT was developed by Dave Johnson and all the code was written in assembly language. There were a lot of fudge factors in the code, often with the comment "this seems to work." I had a co-op student convert all the code to C and port it to an STM processor -- that was released as the MX5, and later used in the MX Sport and GM24K models.

Hey thanks Carl for all this info! In terms of actual field performance how does the MXT Pro compare to the MXT - Prospecting mode, hot soil? I actually used a MX Sport briefly and it was interesting, certainly different from the MXT, I didn't have it long enough to really get in the groove with it though...

Wow I didn't know about all the detector cloning going on, guess maybe due to having blissfully lived so far without an account on aliexpress...but I suppose I'm not surprised...I think I would prefer to ignore the clones and focus on how the real MXT operates...what a purist, eh!

There's a local fellow looking to sell a nice condition MXT Pro... ...so pls let me know your thoughts on the MXT Pro...and is there a way to spot a clone?

Wonder what the controller chip part number is in the Pro? Would like to go over it's datasheet.

[Carl-NC]

I owned the GMT for more than 12 years before I even knew what GMT stood/stands for (Ground Measure Technology).

GoldMaster Tracking

[Carl-NC]

The original MXT board used an OTP chip that was read-protected. It was updated when I worked there to use a later version chip (for the MXT Pro) that could be reprogrammed, but it was still read-protected. It is possible to defeat the read protection and extract the code but it's a complicated process. An easy solution is to send the board to China and one of the hacker houses will send you back the extracted binary, but it costs around $500. Might be able to find someone who will do it cheaper. The binary can be decompiled to assembly language and then maybe AI can clean it up and even convert to C.

The Chinese cloned the GMT which uses almost the exact same circuitry and mostly the same code as the MXT. It's possible the GMT clones are not protected. In any case, they are a lot cheaper to buy and cannabalize. But if you want to extract the code with them most features, it would be the MXT Pro or the MX Sport.

The original MXT was developed by Dave Johnson and all the code was written in assembly language. There were a lot of fudge factors in the code, often with the comment "this seems to work." I had a co-op student convert all the code to C and port it to an STM processor -- that was released as the MX5, and later used in the MX Sport and GM24K models.

[KRinAZ]

Thx again for this pic, I've started a couple searches to try to acquire one of these boards and/or a broken MXT for research.
I'm also searching for info on how to download the code from the Microchip PIC16C76 that the MXT uses. So far I see this is the One-Time-Programmable version of the chip, that the programmer and IDE can still be found, and so far I haven't seen any code protections covered in the datasheet, but also don't yet see a method to download the code. Per the datasheet the IDE is used to compile and upload code to the PIC, not clear if the IDE has read capability from the PIC, I'm still reading the datasheet though...

And of course this project could use a more modern controller, in which case the existing code wouldn't be of much use, unless the existing code could be reverse compiled - unlikely but maybe...back in the day my first and fav language was Assembly, then C as a shorthand for Assembly...but that was a long time ago...

[KRinAZ]

"...The reason the MXT is as good as it is, is because of that long process of field-testing and revisions..."- from Daves letter.
That's the most significant claim. I like it.

I agree!

[ivconic]

"...The reason the MXT is as good as it is, is because of that long process of field-testing and revisions..."- from Daves letter.
That's the most significant claim. I like it.

[KRinAZ]

There are some features in the Goldmaster GMT that were left out of the MXT in gold prospecting mode that need to be included in a new updated MXT. (Ground Measure, ie ground mineralization strength).

I owned the GMT for more than 12 years before I even knew what GMT stood/stands for (Ground Measure Technology). "What the heck is Ground Measure Technology good for anyways ?".

Same thing with the DFX... there are some features in the DFX that would benifit a new MXT All Pro re-incarnation. Plus we could add some new utra-high-tech features not in any of these three detectors like advanced target seperation technology simular to what we see in the Manticor's 2-D indication mapping system and even beyond what we find in the Maticore.

The 13.9 kHz on the MXT is good for larger peices of gold deep say below 4" but will miss smaller peices of gold above 3" that the GMT will hit.

The GMT with its high 48 kHz frequency will struggle in higher mineralization and simply will eventually not work in hotter soils no matter what dials we turn or settings we select. The newer Goldmaster 24K and GMX were suppose to fix all of that, but like the GMT the 24K puts up a deceint performance in the air, but in highly mineralized soils... just cannot punch deep enough into the mineralized goldfield soils to hit the larger deeper targets we all want.

Simply put the GMT needs more coil energy and a lower frequency (say 24K). The newer Goldmaster 24K needs a lower frequency and the 14" double D coil to punch deeper into higher mineralized soil.

The MXT needs a multi channel system that can run a 6.9 kHz frequency for deeper gold and silver targets, 13.89 kHz for general use and a 24 kHz channel for gold nuggett prospecting.

This may well be outside the limits of the current stock coils including the aftermarket coils created later for the DFX and MXT.

But since there is a limited supply these origional coils... a new rebuild would have to include a new modern DD multi frequency compatable coil system and different sized coils.

...and obviously a new MXT would have a volume control and be waterproof like the MX sport and would have an updated filtering system for mineralized goldfield soils and a new powerhouse ARM processor for better target seperation.

To do it right would require the support of Garrett who ( in my opinion never going to develop a new Whites MXT that would just absolutly cannibalize there entire VLF line including the 24K.

On the DFX, there are two analog ground mineralization filters and 4 additional ground filters in the software that can be turned on or off by the user and this ability was not in the MXT or the GMT.

The MXT and GMT have both analog filtering circuitry and filtering in the software plus a ground balance Auto-Trac feature which even today is still well respected across the industry.

When hunting for gold nuggets in Washington, Oregon and Colorado and anywhere else where the iron mineralization is considerably high... We better have a pulse induction detector on stand by that has both an ability to hit larger deeper peices of gold and also be able to hit smaller peices of gold shallow.

For me... I have a new in the box TDI SL with the two stock coils and the newer Nuggettfinder MK-2 Sadie coil ready to go when the ground gets to hot for my ORIGIONAL LOS BANOS GOLD BUG and the newer Minelab Manticore running in multi-frequency gold mode... which combined are two of my more prized gold finders.

My guess is the newer Minelab SD 2300 will easily upstage the older TDI SL if anybody wants to spend the money... especially or defenatly when working around and in water.

Before I go all in on a total remake and update of the MXT/GMT... I would like to first explore options that will alow the MXT to hit smaller peices of gold shallow with the current stock Whites Eclips double D shooter coil without going all out hog wild on a bunch of big complex modifications for the MXT.

Because at the end of the day...

We still wouldn't get the target seperation of the Manticore, but running the MXT in relic mode for gold does cut out a decient amount of iron trash which is still better than nothing but still not even close to what we all want in a modern gold detector.

Ideally what we all want are vlfs that can run in hotter iron goldfield soils and still hit small gold shallow and also hit deeper gold all on the same pass and also seperate small gold from all of the small junk targets that makes detector prospecting unbearably frustrating and unproductive.

Because highly mineralized soil clobbers the depth performance of vlf detectors (like the GMT) they may never be able to hit deeper gold targets in highly mineralized soil without some currently unknown technology.like lowering the frequency back down to 6.59 kHz.

Right now and for the forseeable future hunting smaller peices of gold shallow and larger deeper peices of gold on the same pass with a vlf is gona be relegated to milder soils like many of the soil types we see in places like Arizona and in Southern California.

The newer Nokta Legend II is suppose to handle hotter soils better than the origional Nokta Legend plus better stone age VDI target seperation to boot.

The Manticore is a huge step forward in the direction we all wish to go for vlfs and to a lessor degree so was the Equinox 800 and 900... but these detectors still fall short of what is possible with todays newer advanced microchips, miniturized electronics and advanced software technology.including neural network computing and fuzzy logic.


Special thanks... not only to David Johnson and everybody at Fisher and Whites, but also all of the briliant engineers worldwide who have collectivly brought us all to where we are today.

"Now let's go find some gold!"

Lots of good thoughts...I think it would be wise to stick with Carl's advice in post 2 - start simple and build from there. Dave Johnson was a brilliant engineer and I think it will be a monumental task to try to recreate what he designed, and even more so to try to build upon it.
Just getting ground balance to work as well as the GMT/MXT will be quite the accomplishment.

I personally would be ecstatic to just build a functional MXT as I think it performs in the field as good as most any VLF, and as my above story illustrates, as good as a MineLab PI up to the MXT's VLF depth limitations.

FWIW the corner of Arizona I live in (generally the north side of the Bradshaw Mountains) has acres and acres of very hot soil and river bed. As in - I cannot simply set the coil to the ground with the default gain setting - if I do I just get the screaming Coil Overload, Turn Down Gain message and tone. No metal bits of any type, just the hot soil. There are outcrops with so much iron that the magnet on my pick sticks to the outcrop! Yet I can detect through and find sub gram little nuggets as well as tiny lead shot, tiny rust fragments, etc... Actually every gold hunting locale in Arizona that I have been to has very hot ground. So the MXT in hot field use performs very well - at least in the hands of someone familiar with it.
This is the area that Tesoro developed the Lobo SuperTraq to handle, specifically the Lynx Creek river bed that is so hot and known for tiny nuggets (as well as the rare lucky larger ones).

The local prospecting shop has a big tub of Lynx Creek river bed material and PVC tubes running horizontal through it so targets can be inserted down the tube, various detectors tested, and various targets detected (attempted anyway). We have spent time comparing detectors in the tub - the detectors that stand out far from the rest are:
Tesoro Lobo Supertraq (very close third best target signal, or maybe a tie with the MXT)
Fisher Gold Bug and II (sketchy target signal probably due to being difficult to ground balance over the tub)
White's GMT (the winner with the clearest target signal)
White's MXT (a close second place, second clearest target signal)
The other detectors we tried gave no diggable target indication and go bonkers with noise and ghosts.

On coils, I have all 3 of the White's DD coils for the MXT - The little 4x6 DD "Shooter", the standard 6x9 DD, and the 8x14 DD "Deep Scan", and so can test with those. I don't have any of the concentrics - the ground is too hot here to use them so I never bothered buying them, though in the past I borrowed them to try out.

[PetraDiamond5]

There are some features in the Goldmaster GMT that were left out of the MXT in gold prospecting mode that need to be included in a new updated MXT. (Ground Measure, ie ground mineralization strength).

I owned the GMT for more than 12 years before I even knew what GMT stood/stands for (Ground Measure Technology). "What the heck is Ground Measure Technology good for anyways ?".

Same thing with the DFX... there are some features in the DFX that would benifit a new MXT All Pro re-incarnation. Plus we could add some new utra-high-tech features not in any of these three detectors like advanced target seperation technology simular to what we see in the Manticor's 2-D indication mapping system and even beyond what we find in the Maticore.

The 13.9 kHz on the MXT is good for larger peices of gold deep say below 4" but will miss smaller peices of gold above 3" that the GMT will hit.

The GMT with its high 48 kHz frequency will struggle in higher mineralization and simply will eventually not work in hotter soils no matter what dials we turn or settings we select. The newer Goldmaster 24K and GMX were suppose to fix all of that, but like the GMT the 24K puts up a deceint performance in the air, but in highly mineralized soils... just cannot punch deep enough into the mineralized goldfield soils to hit the larger deeper targets we all want.

Simply put the GMT needs more coil energy and a lower frequency (say 24K). The newer Goldmaster 24K needs a lower frequency and the 14" double D coil to punch deeper into higher mineralized soil.

The MXT needs a multi channel system that can run a 6.9 kHz frequency for deeper gold and silver targets, 13.89 kHz for general use and a 24 kHz channel for gold nuggett prospecting.

This may well be outside the limits of the current stock coils including the aftermarket coils created later for the DFX and MXT.

But since there is a limited supply these origional coils... a new rebuild would have to include a new modern DD multi frequency compatable coil system and different sized coils.

...and obviously a new MXT would have a volume control and be waterproof like the MX sport and would have an updated filtering system for mineralized goldfield soils and a new powerhouse ARM processor for better target seperation.

To do it right would require the support of Garrett who ( in my opinion never going to develop a new Whites MXT that would just absolutly cannibalize there entire VLF line including the 24K.

On the DFX, there are two analog ground mineralization filters and 4 additional ground filters in the software that can be turned on or off by the user and this ability was not in the MXT or the GMT.

The MXT and GMT have both analog filtering circuitry and filtering in the software plus a ground balance Auto-Trac feature which even today is still well respected across the industry.

When hunting for gold nuggets in Washington, Oregon and Colorado and anywhere else where the iron mineralization is considerably high... We better have a pulse induction detector on stand by that has both an ability to hit larger deeper peices of gold and also be able to hit smaller peices of gold shallow.

For me... I have a new in the box TDI SL with the two stock coils and the newer Nuggettfinder MK-2 Sadie coil ready to go when the ground gets to hot for my ORIGIONAL LOS BANOS GOLD BUG and the newer Minelab Manticore running in multi-frequency gold mode... which combined are two of my more prized gold finders.

My guess is the newer Minelab SD 2300 will easily upstage the older TDI SL if anybody wants to spend the money... especially or defenatly when working around and in water.

Before I go all in on a total remake and update of the MXT/GMT... I would like to first explore options that will alow the MXT to hit smaller peices of gold shallow with the current stock Whites Eclips double D shooter coil without going all out hog wild on a bunch of big complex modifications for the MXT.

Because at the end of the day...

We still wouldn't get the target seperation of the Manticore, but running the MXT in relic mode for gold does cut out a decient amount of iron trash which is still better than nothing but still not even close to what we all want in a modern gold detector.

Ideally what we all want are vlfs that can run in hotter iron goldfield soils and still hit small gold shallow and also hit deeper gold all on the same pass and also seperate small gold from all of the small junk targets that makes detector prospecting unbearably frustrating and unproductive.

Because highly mineralized soil clobbers the depth performance of vlf detectors (like the GMT) they may never be able to hit deeper gold targets in highly mineralized soil without some currently unknown technology.like lowering the frequency back down to 6.59 kHz.

Right now and for the forseeable future hunting smaller peices of gold shallow and larger deeper peices of gold on the same pass with a vlf is gona be relegated to milder soils like many of the soil types we see in places like Arizona and in Southern California.

The newer Nokta Legend II is suppose to handle hotter soils better than the origional Nokta Legend plus better stone age VDI target seperation to boot.

The Manticore is a huge step forward in the direction we all wish to go for vlfs and to a lessor degree so was the Equinox 800 and 900... but these detectors still fall short of what is possible with todays newer advanced microchips, miniturized electronics and advanced software technology.including neural network computing and fuzzy logic.


Special thanks... not only to David Johnson and everybody at Fisher and Whites, but also all of the briliant engineers worldwide who have collectivly brought us all to where we are today.

"Now let's go find some gold!"