My stm32 powered pi detector

[Tinkerer]

I use LI 18650 cells in everything I do, then couple with a charge/balance controller.
There is boost circuit for the tx already but I would not go lower than a 4 cell +/- 16v, but how many in parallel would be up to you.
I have been doing some LTspice sims trying to cross over Thinkerer's IB research with mixed results, the problem is the bucking coil and bipolar operation.
On the ADC side, I think I will go with the AK4558EN as there is plenty of documentation out there for troubleshooting.

Bi-polar sounds like a good idea. Theoretically it could take care of quite a few problems. In real circuits, however, it makes things much more complex. Not 2 times more complex, more like 2x2 times more complex.

For IB coils, David Emery gives very good instructions. http://www.geotech1.com/cgi-bin/page...anar/index.dat

[waltr]

For IB coils, David Emery gives very good instructions. http://www.geotech1.com/cgi-bin/page...anar/index.dat

Also, see my thread on the HH2:
http://www.geotech1.com/forums/showt...ake-on-the-HH2
I have details on building and using IB DD & Concentric coils with a PI detector.

[green]

Bi-polar sounds like a good idea. Theoretically it could take care of quite a few problems. In real circuits, however, it makes things much more complex. Not 2 times more complex, more like 2x2 times more complex.

For IB coils, David Emery gives very good instructions. http://www.geotech1.com/cgi-bin/page...anar/index.dat

Carl has posted Bi-polar is easier, think he used it in his pin pointer. Looks more complicated to me, maybe some discussion on pro's and con's of Bi-polar.

[waltr]

Carl has posted Bi-polar is easier, think he used it in his pin pointer. Looks more complicated to me, maybe some discussion on pro's and con's of Bi-polar.

It does make the TX harder: either a center tapped coil and half bridge or a full H-bridge for a single-end (TX) coil.
Also, a full bipolar RX front end, differential, adds parts and needs a center-tapped RX coil.

Bipolar makes the processing easier. No EFE pulses needed.

These are good options but needed on a separate PCB from the STM32. This would allow easier experimenting with TX/RX circuits.

[chafik]

I think for PI metal detector using 2 microcontroller is a good point 1 micro for timing sampling and controlling battery level the 2 micro is for GUI setting discrimination communication processing 3d image GPS logging for the 1 micro I prefer Analogue Device arm micro M3 with 24 bit delta segma adc 128k flash, 2 microcontroller could be raspberry pi or TI beaglebone or any stm32f7 for low cost using the new dual core stm32 with external 24bit adc

[gtdavid]

I really like that mcu nice pick eval boards are overpriced. But the chip it's self is only $7 that's a real option I like the adc with its pga's how is there support and software?
I have ditched the LCD in favour of a ui over Wi-Fi setup cheap easy and allows programming with out opening up your detector everytime. I thought about a GPS but every phone these days has them so one could just use a gmaps app to flag finds.

[Mdtoday]

From experience, I can say that using a phone with blue-tooth or WiFi link will be hit and miss when it comes to interference and noise.
Even if you make all the necessary shielding on PCB around front end using best practices, you still have a coil to worry about and even when you shield that coil there is still the possibility of picking up interference.
Moodz had a design called the Magpi and I believe he gave up on the idea because of this problem.
Sure you can get rid of a lot noise but you can dumb down the detector response in the process. Everything is a trade off in this PI game.
For this reason, I would suggest you try out the idea using available evaluation boards first to see for yourself, the issues you face, then commit each area to PCB as you prove them.
This is the route I had taken early on and with using pre-compliance EMC EMI testing chamber, was able to nut out a lot of issues before committing the design to PCB.
BTW, there are simple little tricks like the use of ferrite shielding/sheeting on top of processors that help in this manner along with good layout practice.

Of course you may have already done some or all of the above, in which case, I'll just shut up



On the GPS, on my design I used the early Jupiter SMD module from Navman along with some derivatives, all autonomous and they were good for 2m accuracy at best, differential GPS is required for sub-metre accuracy but do we need it?
With the same design , I used 3 processors , 2 AD DSPs + 1 PIC32 but this was on a multi coil array with 2 x bipolar TX and 8 receive circuits, it was built for a specific purpose.
I don't think you need anymore than one capable processor on your design.

Personally, when going to my remote prospecting areas, I prefer a rugged GPS unit + the usual safety comms, in areas of good reception, phones are OK but there is a trade off, just my 2 bobs worth.

On the LCD, maybe you could include layout to cater for a simple character unit and have a port connection for a graphics module, at the end of the day, one really only needs it while experimenting in field.
Once that is completed, its into the real use of the detector which doesn't require such a display. A simple small character single line display is more than enough.

Anyway, this is not meant to be a negative post, more as observations from going down this path myself

cheers

MDtoday

[6666]

Just asking a question, what is the expectation of this project will it be something that can be reproduced ?

[Carl-NC]

Carl has posted Bi-polar is easier, think he used it in his pin pointer. Looks more complicated to me, maybe some discussion on pro's and con's of Bi-polar.

For sure, bipolar TX makes Earth field cancellation easier, you get it whether you want it or not. Bipolar can be done either with an H-bridge or with a CT winding. Each has pros/cons, but both require more clocking than a simple unipolar TX. Bipolar invariably adds more C to the TX, making aggressive early sampling more difficult. And if you want to use a preamp input isolation switch, bipolar makes that way more difficult.

[moodz]

Moodz had a design called the Magpi and I believe he gave up on the idea because of this problem.

MDtoday

I didnt give up ... just gone to sleep ... too busy working for the man LOL ....

I use serial over bluetooth ... to talk to GUI on bluetooth phone .. the other version was just a bluetooth terminal app with text commands / status to change things in the detector.

I detect with a mate as the digger .... ( we swap turns ) ... the digger adjusts the detector using his Bluetooth phone whilst the searcher concentrates on swinging. That way the digger and the phone is well away from the search coil.

There is noise ( random ) and there is the PI signal ( not random ) ... sorting them out is not so hard :-)

This patent does the sorting ( and earth field cancellation ). http://pericles.ipaustralia.gov.au/o...nNo=2017100579

Ive got a demo cct somwhere that uses no micros to demonstrate this ...

moodz

[gtdavid]

Yes I want others to make it..
Just still cleaning up schematics then will post the ideas and suggestions have been great and have led to me changing the design hence the delay

[gtdavid]

My design been based on a h bridge from the beginning but the problem I have been having is with use of a bucking coil . Basically having to switch it from one side to the other.
The Wi-Fi "should " be OK I have my BIG Wi-Fi router sitting right above my lab bench and have never really had problems from it.
On the ui side I will be going with Wi-Fi as the esp32 module can broadcast Wi-Fi ap so any phone with Wi-Fi can use it with out the need of an app and then I will be utilizing the Bluetooth for Audio.

[gtdavid]

My design been based on a h bridge from the beginning but the problem I have been having is with use of a bucking coil . Basically having to switch it from one side to the other.
The Wi-Fi "should " be OK I have my BIG Wi-Fi router sitting right above my lab bench and have never really had problems from it.
On the ui side I will be going with Wi-Fi as the esp32 module can broadcast Wi-Fi ap so any phone with Wi-Fi can use it with out the need of an app and then I will be utilizing the Bluetooth for Audio.

[Mdtoday]

I didnt give up ... just gone to sleep ... too busy working for the man LOL ....

I use serial over bluetooth ... to talk to GUI on bluetooth phone .. the other version was just a bluetooth terminal app with text commands / status to change things in the detector.

I detect with a mate as the digger .... ( we swap turns ) ... the digger adjusts the detector using his Bluetooth phone whilst the searcher concentrates on swinging. That way the digger and the phone is well away from the search coil.

There is noise ( random ) and there is the PI signal ( not random ) ... sorting them out is not so hard :-)

This patent does the sorting ( and earth field cancellation ). http://pericles.ipaustralia.gov.au/o...nNo=2017100579

Ive got a demo cct somwhere that uses no micros to demonstrate this ...

moodz

Hi Moodz,

Good to hear you still up and about then 😁

I thought I read in a post a while back you'd dropped the idea of using phone because of noise but anyways, seems like you worked it out by using a 2 man operation hope you are finding some yellow stuff 👍

GTdavid, was wondering, have you built up and tested a front end yet?

cheers

MDtoday

[Tinkerer]

My design been based on a h bridge from the beginning but the problem I have been having is with use of a bucking coil . Basically having to switch it from one side to the other.
The Wi-Fi "should " be OK I have my BIG Wi-Fi router sitting right above my lab bench and have never really had problems from it.
On the ui side I will be going with Wi-Fi as the esp32 module can broadcast Wi-Fi ap so any phone with Wi-Fi can use it with out the need of an app and then I will be utilizing the Bluetooth for Audio.

You could try a DD induction balanced coil configuration instead of the concentric with the bucking coil. The bucking coil problem is one of the complications of the bi-polar system. There are a few more.