New Bipolar Boost TX and Front End

[Mdtoday]

I have settled on the KEMET EC2-3NU. It is stocked by LCSC whereas the OMRON 3V version was not (and the OMRON cantact voltage rating was only 125V). The following circuit is the drive and connection configuration for the DPDT relay. It should provide adequate protection of the FPGA from transients in case of a relay malfunction. I have used this coil connection configuration in the Vallon with a DPDT manual 3 position switch.

Looks good JL, nice neat little circuit, thanks for sharing.
I'm currently working on a version of the sub-enclosure and will try post progress later today.

[KingJL]

Probe/Search coil switch relay

... I would stay with OMRON or similar quality brand then as I have had recurring nightmare experiences with Chinese relay manufactures, they are not all bad but mechanical reliability on most leaves a lot to be desired...

I have settled on the KEMET EC2-3NU. It is stocked by LCSC whereas the OMRON 3V version was not (and the OMRON cantact voltage rating was only 125V). The following circuit is the drive and connection configuration for the DPDT relay. It should provide adequate protection of the FPGA from transients in case of a relay malfunction. I have used this coil connection configuration in the Vallon with a DPDT manual 3 position switch.

[Mdtoday]

Also an update: I just changed the OLEDrgb support ip so that the SPI clock is upped to 6.25 MHz. I felt that this was justified by the fact that a majority of processing time will be spent in updating the display since there is a new received data set every 4 msec. This puts the SPI clock at the upper limit for the OLEDrgb module per the specification sheets. It has all tested out fine.

That is great news

[Mdtoday]

I think you may have miss-understood my "rectangular enclosure" vision. The "rectangular enclosure" than I was envisioning, was a sub-enclosure (metal or metalized plastic) much like the Vallons packaging of it's 2 main electronics boards. This was simply to provide more shielding to the sensitive areas of the electronics and to facilitate more abstraction for the support modules (power supplies (in my case 3 small boost regulators for 5V, 20V, and 3,3V) audio support circuitry, and coil/probe switch w/support circuitry) and how they are positioned. Sorry for the confusion. I like the designs that I have seen for the overall enclosure.

Not a problem JL, I should have realised, I have a Vallon too, it just didn't click in my brain.

There will be plenty of room in the housing still, I can just enlarge by the required amount if needed.
I spray the inside of my 3d prints with nickle screening which does the trick or a ready made metal enclosure as you say.

Here is a view of boards mounted horizontal to give an idea of available space and where it will require adjustment to accommodate the shielded enclosure.

[KingJL]

I'll finish off this attached design then revert back to the rectangular one and finish that off when the OLED module arrives in 2 or 3 days.

I think you may have miss-understood my "rectangular enclosure" vision. The "rectangular enclosure" than I was envisioning, was a sub-enclosure (metal or metalized plastic) much like the Vallons packaging of it's 2 main electronics boards. This was simply to provide more shielding to the sensitive areas of the electronics and to facilitate more abstraction for the support modules (power supplies (in my case 3 small boost regulators for 5V, 20V, and 3,3V) audio support circuitry, and coil/probe switch w/support circuitry) and how they are positioned. Sorry for the confusion. I like the designs that I have seen for the overall enclosure.

Also an update: I just changed the OLEDrgb support ip so that the SPI clock is upped to 6.25 MHz. I felt that this was justified by the fact that a majority of processing time will be spent in updating the display since there is a new received data set every 4 msec. This puts the SPI clock at the upper limit for the OLEDrgb module per the specification sheets. It has all tested out fine.

[Mdtoday]

We need to wait until we actually mount the 10 pin connectors and marry the two boards and then measure what clearance we need. I suspect it will be about 2"x1 1/2"x4".

I have been working on both enclosure designs using manufacturers 3D models and comparing with the real parts and they match quite well, this is the way I normally start enclosure designs. Also I set up parametric tables for the actual enclosure measurements and can change most things without having to redo the whole design, saves a lot of work.
I'll finish off this attached design then revert back to the rectangular one and finish that off when the OLED module arrives in 2 or 3 days.

The following pictures show the general idea with all of the mounting pillars and rear hatch.


I have 3D printed a half sized version to confirm it actually prints without problems and its Ok, so next stage is to correctly position the PCB mounting pillars and add some ribs, then do a full sized test print.

In the end I will have 2 versions for you to take a look at and decide on changes etc, I'll then either send STL files if you have a 3D printer or print a couple off and send if you would like them.

I will post progress on the rectangular version soon

Cheers

Mdtoday

[KingJL]

I had envisioned a metal or metalized plastic 2"x2"x4" (inside dimensions) rectangular circuit enclosure ...

We need to wait until we actually mount the 10 pin connectors and marry the two boards and then measure what clearance we need. I suspect it will be about 2"x1 1/2"x4".

[Mdtoday]

There are quite a few miniature DPDT relays with 3V or 5V coils (KEMET(EC2 series), OMRON(TX2 series), SAMSUNG, and quite a few Chinese manufactures). The KEMET and OMRON I know have contacts rated at 250VDC/225VAC. We would not want the ones with latching coils. After some thought and research, I believe the solid state switches would add in too much capacitance. It would be easy to add the AXI4 interface to the userinput module to control the DPDT relay circuitry. Having the switch controlled by the FPGA means that we can control the whole process (disable transmit, switch to mono operation if required, switch to probe, enable transmit). That way there are no spurious transients that can occur that are prone to occurring by just switching an active TX.

That is true, I would stay with OMRON or similar quality brand then as I have had recurring nightmare experiences with Chinese relay manufactures, they are not all bad but mechanical reliability on most leaves a lot to be desired.

I like the idea of controlling the switch-over process.

[Mdtoday]

For those not familiar with the Digilent OLEDrgb (96x63) display module, it displays 12 characters per row and 8 rows. My thoughts on the display for normal operation and the menu system are as follows:

These thoughts are extremely preliminary!

Looking good, menu layout and operational functions should do the job nicely and can be trimmed or extended as the project progresses.
It won't take long to figure out the most useful positions etc.

cheers

Mdtoday

[KingJL]

The switched probe is an excellent addition, a solid state switch would be my choice.

There are quite a few miniature DPDT relays with 3V or 5V coils (KEMET(EC2 series), OMRON(TX2 series), SAMSUNG, and quite a few Chinese manufactures). The KEMET and OMRON I know have contacts rated at 250VDC/225VAC. We would not want the ones with latching coils. After some thought and research, I believe the solid state switches would add in too much capacitance. It would be easy to add the AXI4 interface to the userinput module to control the DPDT relay circuitry. Having the switch controlled by the FPGA means that we can control the whole process (disable transmit, switch to mono operation if required, switch to probe, enable transmit). That way there are no spurious transients that can occur that are prone to occurring by just switching an active TX.

[Mdtoday]

I do not think cables pose any problems. .....
We may have a challenge with the Pmod OLEDrgb display.......
De-soldering components with multiple through holes without damaging the PCB can be problematic......

What are your thoughts on power source?
One other thought on the enclosure... Let's consider the option to use a switched probe. This requires space for a 3 position dpdt switch and another coil connector. Or the coil switch might be able to be accomplished with a solid state relay(s) and a small switch on the front of the enclosure with the relay control via the electronics. The later scenario would allow the probe operation even when using an IB main coil. The software could turn the IB receive path off during probe operation. I need to research for suitable solid state relay options.

I agree, SPI over a short screened cable at <5MHz will be Ok, I think the limiting factor is more likely the quality of the OLED module.

The draft design I posted earlier was designed to carry the PCBs in the vertical plane and accommodates a rear accessed bay for 2 x 18650 batteries. So all electronics mounted in the head.
Not shown on the drawing is the addition of the coil connector which is mounted rear facing at the base and on the side of the shaft clamp, a tear drop bump (I will post the updated models).
The internal coil cable runs up through the handle and is soldered directly to the PCB, it is unusual but effective as the coil connector is right next to the shaft.
There is another part to this enclosure design and that is the stand which will either be integral or a snap on, it is still to be decided.

There is a product called Chip Quick along with BGA flux for removing smd parts and connectors from boards with no damage, its not cheap but does last a long time.

On the power source, my preference is to keep things light as possible for gold prospecting therefore I generally use 2 high capacity 18650 batteries in my designs with 8 hrs run time.
The power pack is a great idea though and perhaps options for both should be included, I will work on this.

The switched probe is an excellent addition, a solid state switch would be my choice.

cheers

Mdtoday

[KingJL]

Menu/Display System

For those not familiar with the Digilent OLEDrgb (96x63) display module, it displays 12 characters per row and 8 rows. My thoughts on the display for normal operation and the menu system are as follows:

In normal operationthe display will be:

Bipolar PI

S1


S2


Sens


Where the special symbols will be illuminated depending on signal strength for S1 and S1. The Sens symbols will indicate the Sensitivity setting and can be adjusted with the encoder knob. Default sensitivity will be all illuminated.
To go from normal operation display to the menu system, press the encoder shaft button. When the main Menu is exited the display will return to the normal operation mode. The main menu will exit when "Exit" or "Probe" ("Search") is selected.
The "Select" or menu item is selected by pressing the encoder shaft button when the selection is high lighted.

The Main menus will be 8 items arranged in 2 columns:

Main Menu

GB PRT
Probe S1 (Probe/Search)
Maint S2
Filt Exit

In theMain menu in normal operation, the 2^nd entry text will be "Probe"; In probe mode operation the 2^nd entry text will be "Search".
The PRT, S1, and S2 sub-menus will be the display of the numerical value in usec and a "Select" item. The Filt selection sub-menu will be a numerical filter selection (that relates to speed of the response) for the mineralization compensation filter and a select item.i
The Maintenance menu will exit back to the Main menu when either "Exit" or "IB COIL" ("MONO COIL") is selected.
In all sub-menus where the numerical value is displayed with the "Select" entry, the numerical values can be changed by turning the encoder knob CW (+) or CCW (-).
The Maintenance sub-menu will be 4 items arranged in 1 column:

Maint

TX Off
Damp Tune
IB (Mono) Coil

Exit
In theMaint menu, the 2^nd entry text will be "TX OFF" if TX is currently on or "TX ON" if the TX is currently off. The 3^rd entry text will "IB Coil?" if current mode is mono coil or "Mono Coil" if current mode is IB coil.

These thoughts are extremely preliminary!

[KingJL]

... Or the coil switch might be able to be accomplished with a solid state relay(s) and a small switch on the front of the enclosure with the relay control via the electronics. The later scenario would allow the probe operation even when using an IB main coil. The software could turn the IB receive path off during probe operation. I need to research for suitable solid state relay options.

Maybe something like this. It would require a small PCB and a transistor/MOSFET to isolate the drive from the FPGA ( I would hate for the HV pulse to leak through).

[KingJL]

... yes sounds like cables are the way to go.
I have been working on 2 types of enclosure. The one attached [ATTACH]47028[/ATTACH] is a draft version which can be 3D printed in 2 sections split vertically in the centre and held together with screws.

I do not think cables pose any problems. The most critical would be the OLEDrgb as that is an SPI connection with <= 5MHz clock (I haven't actually measured it, but the parameters indicate that it should be 3.125 MHz). I have driven 5 MHz across 1 meter cable with no problem. All other connections are either in the audio or sub-audio range. Nice looking enclosure! We may have a challenge with the Pmod OLEDrgb display. When yours arrives, you'll see what I mean. It is the position of the male 12 pin connector. We may have to consider de-soldering the connector and flipping it to the other side to get clearance. There is also the possibility to de-solder and swap out with a vertical male 12 pin. De-soldering components with multiple through holes without damaging the PCB can be problematic.

What are your thoughts on power source? On my re-packaged Vallon, I discarded the use of batteries and used a USB power pack mounted in a waterproof enclosure on the tail of the shaft with the power cord going through the shaft. This provided some counter weight and provided a convenient means for ground stabilization. Just a thought. These modern day USB power packs can power a detector for days (30 operational hours or more) if needed.

One other thought on the enclosure... Let's consider the option to use a switched probe. This requires space for a 3 position dpdt switch and another coil connector. Or the coil switch might be able to be accomplished with a solid state relay(s) and a small switch on the front of the enclosure with the relay control via the electronics. The later scenario would allow the probe operation even when using an IB main coil. The software could turn the IB receive path off during probe operation. I need to research for suitable solid state relay options.

[Mdtoday]

From some previous projects where Pmod connectors were used, I have a small assortment of Digilent Pmod cables (8" I think). I have a couple of 12 pin cables, 1 splitter cable which splits the the 12 pin to 2 6pin connectors and 1 (or 2) 6 pin Pmod cables. The Pmod OLEDrgb uses a 12 pin connector (whole Pmod) while the rotary encoder uses 1/2 of a Pmod (6 pins). My thinking was to use a cable connected 12 pin Pmod for the OLED display and the splitter cable with one section going to the rotary encoder and the other section connecting the PP and GB switch wiring and audio PCM out..I also have used the through hole 90 degree 12 pin sockets as 90 degree socket adapters for Pmod boards/cables. I think a cable/wiring connector system gives the greatest latitude with enclosure placement/design.

I had envisioned a metal or metalized plastic 2"x2"x4" (inside dimensions) rectangular circuit enclosure with openings for the power/antenna wiring, the four side/end mounted Pmods, and the one bottom(or top depending on your perspective) Pmod which could be mounted in the main housing. Also there is the small hole opening needed for the reference voltage adjustment.

I do not think there is much more to consider as for inputs or outputs as there are only 4 (after adding the audio PCM output) more output ports that will support LVCMOS33. That little tidbit I learned from this past weeks marathon in getting the user input module integrated. I tried to temporarily add 6 test ports to the design so I could hook a scope up to observe some of the internal signals... well I could add only 5. But, in the end, 5 was enough.

Thanks for the info JL, yes sounds like cables are the way to go.
I have been working on 2 types of enclosure. The one attached KingJL_Bipolar_TX_Enclosure_V1.pdf is a draft version which can be 3D printed in 2 sections split vertically in the centre and held together with screws.
I'll do some more work on the other one now that I have your thoughts.

Cheers

Mdtoday