Universal PI Micro

[bugwhiskers]

The attached .pdf shows a preliminary detailed overview of a stand alone circuit board that will greatly aid those conducting experiments in PI metal detection. It can drive PI's like the Hammerhead, Gary's PI, Goldscan4 etc.

There are 3 outputs to drive analogue switches or JFET's, most PI's only use 2. The delay between each output and the sample width are adjustable from 0.5uS to 512uS and are displayed simultaneously on a 16 character * 2 line screen. The screen will have another page to show a digital readout and bar graph of the final stage signal strength.

Two pots and a pushbutton serve to navigate around the adjustments and displays. Turning the "mode" pot will cause stepping between various screen pages and variables. The "adjust" pot causes the variable to be changed with the "accept" button storing the new value. Adjustments will be live, so the user has instant feedback. There will be provision for storing (retained with power off) 16 sets of all variables into EEPROM.

Feedback and suggestions for additions are very welcome and the sooner the better as I am in the process of laying out the board.

To complicate things, I have finally bitten the bullet and started to learn to use "Eagle" light PCB program after using a DOS version of PROTEL since 1990.

regards
bugwhiskers

[bugwhiskers]

Layout of adjustment screen

The attached .pdf (which needs to be rotated right, clockwise) shows the intended layout for the adjustment screen. All the delays and widths will be adjustable from .5 to 512 microseconds. The CT (cycle time) will be shown in Hertz and the maximum frequency is dependant on the total of all the other widths and delays and internal processor tasks (driving screen, reading AtoD inputs etc)

A suggestion from bbsailor was to have the "G" force sensor instigate changes in sampling dependant on sweep speed. No signal from the "G" sensor after a set period could also turn off the TX pulse to conserve battery power.

Additional 4094's on the SPI line could be added to provide extra digital outputs or drive a DAC to produce perhaps changes in offsets or gains.


regards
bugwhiskers

[bugwhiskers]

UPIM audio

One of the major problems with audio in a PI circuit is at the threshold of detection the change in output from the amplifying stages is very small. If the audio is a rising pitch with rising signal strength the change with small or distant targets will hardly be noticable. The use of an anolgue meter does help but clearly a better method is required.

In the preliminary overview posted earlier you will see a box labelled "Audio".
The audio will be generated by the micro's internal PWM. The way this works is fairly simple. The voltage from the final amplifier stage is read by the AtoD and converted into a number from 0 to 1023 (10 bit resolution). That number is fed to the PWM and the output pin produces a signal the frequency of which is dependant on the number it is fed.

Clearly, if the signal input changes from 0 to 1 as it will at the threshold of detection then even Mozart would have trouble hearing the change.

A way around this problem is at the threshold of detection AtoD read = 0 a steady tone (maybe 500Hz) is heard pulsed on and off at .25 Hz. If the AtoD goes to 1 then the rate of the pulsing of the steady tone is doubled, now the change is very discernable. After about 8 doublings of the pulsing of the tone the PWM would go to normal rising pitch to indicate target strength.

regards
bugwhiskers

[johnno]

great post!

The software required to program the atmega168, is this to be public domain?,or do we try ourselves..(Ihave no programming experience.only BASIC).
Thanks
johnno

[bugwhiskers]

Hi johnno,

Thanks for your interest in the project.
The boards will be shipped with the micro pre-programmed.
Source and HEX code (everything you need) will be included.
The board will have a 6 pin programming header for those who want to try their hand at making changes. However, the whole thing is being designed so that all the changes you need to make can be done with the front panel knobs and button.

The timing section of most PI's is very similar, this project supplies that and more, like a MOSFET driver, a +5(over battery +) charge pump, LCD screen, "G" sensor and audio.

regards
bugwhiskers

[johnno]

Re Atmega168

Hello Again ..what price are we looking at?
Thanks
johnno

[bugwhiskers]

Hi johnno,

A bit early to be able to quote a price. Some people may want to supply their own LCD screen, and others may not want the "G" sensor. Sometime soon I hope to have a price for the board with just the CPU soldered in (it's a SMD).

regards
bugwhiskers

[johnno]

Thankx BW I will wait till things settle out..
Thanks johnno

[johnno]

how's it goin' BW

Just a quick note as to how things are progressing?
Also that "G" spot sensor thing...you'll make a million on that alone!..
is that on Protel?
johnno

[bugwhiskers]

Hi johnno,

The schematic and board are being done with "Eagle" light.
As for the "G" sensor (G for gravity, not Groin) the schematic has been changed to incorporate a 3 axis type. The X,Y axis will be used to detect sweep speed/or lack thereof and the "Z" (up down) axis may be used to instigate ground balance.


Support for LED backlit LCD displays has been added also. The micro chip, which is only about 5mm*5mm with 0.3mm leads only 0.8 mil apart will be mounted on a small circuit board with standard 0.1" pin headers. The rest of the circuit board will be standard 0.1" thru hole components making it much easier to construct, modify and repair.

Another option under consideration is the use of a small joystick as the main input interface to the micro. The joystick would provide 5 inputs, up, down, left and right and pressed. This would make for a very uncluttered control panel.

regards
bugwhiskers

[bugwhiskers]

Joystick input

The attached .pdf shows how a 5 switch joystick type input device would be configured.
Each resistor has a different value(to be determined). When each switch is closed a unique voltage is presented to the AtoD. The software is set up to look for a "range" of a voltage, ie 3.7 to 4.1 then performs a predetermined action for that range. Using this arrangement diagonal movement of joystick and indeed any direction plus centre press will give 19 different voltages.


N=NORTH, E=EAST, S=SOUTH, W=WEST C=CENTRE PRESS

C
N,S,E,W
NE,SE,SW,NW
C+ N,S,E,W
C+ NE,SE,SW,NW

regards

bugwhiskers

[bugwhiskers]

Micro SMD to thruhole adaptor board.

Attached .pdf shows a 1" * 1" board to adapt the SMD micro to thru hole. The bypass caps for the micro are on it too. Contemplating putting the external XTAL on it also to cut down on radiated noise.

regards

bugwhiskers

[bugwhiskers]

SMD to thru hole

The attached pic shows the first prototype board ready for drilling and soldering.
The board now includes provision for an external XTAL to enable running the micro at it's full speed 20 MHz.

regards
bugwhiskers

[bugwhiskers]

SDM adaptor board

The attached pic shows the bottom side of the SMD adaptor board. Bypass caps, headers and XTAL components in place.

Work on the main board schematic is about half done. Very slow because I have to make the schematic symbols and packages as well because a lot of the parts aren't in the libraries.

regards

bugwhiskers

[bugwhiskers]

Revised UPIM overview

The attached .pdf shows changes to the schematic. Support for a backlit LCD has been added. Support for a 3 axis "G" sensor and either a Joystick or 5 push buttons has been added also.

There is one spare I/O pin, any suggestions for it's function.


regards

bugwhiskers