Universal PI Micro

[bugwhiskers]

64 SAMPLE BOARDS

The 2 boards that make up the 64 sampler are up and running. The sample caps have been changed from 100nF to 1 nF with an increase in sensitivity not only for signal but also for noise. The major culprit, the voltage doubling charge pump is being modified so it will still produce sufficient voltage when turned off during sampling.
Unfortunately the micro I am using has only enough RAM to store/process 48 of the 64 samples. The preferred micro, ATMEGA644 will be available around the end of April so we will have to wait until then to see the full picture.
The intention at this stage is to get as many samples as possible of the knee of the decay curve with the remaining samples spaced increasingly further apart to the end of even the longest decay. The later samples I suspect will really only indicate the size of the target.
The preferred micro also runs faster (20 MHz) which will allow sampling at 100nS intervals.

regards
bugwhiskers

[bugwhiskers]

Hello BW, Considering all the work you have done on your project, are you planning on building this to a workable PI.? I have read earlier posts as this was to "interface for PI's such as Gary's and Hammerhead".
Have I missed something?
johnno

Hi johnno,

I am indeed intent on getting this PI up and running. Initially it was going to be a Universal micro for controlling PI metal detectors then I had the idea of multiple switched capacitors and the potential they had to discriminate. The UPIM is still a motherboard as suggested by bbsailor with all the functions that are common in a PI, ie MOSFET drive, analogue switch drive for integrators, negative voltage generation for OP amps, Screen drive capability, control button read capability and 2 bus sockets for people to put boards of their own designs for experimentation or a re-hash of an existing design like Gary's PI or the Hammerhead.
The UPIM is fulfilling that function for me in that I now have an easy way to test out the switched multiple cap decay sampler.

Early results show that discriminating ironstone is not a problem. The setup I have at the moment that only gets 24 samples (64 or more samples would be better) also shows a distinct difference between a small gold nugget and a steel screw of about the same size.

Over the coming weeks I will be working on finishing off the necessary software to drive the screen and read the buttons and other housekeeping tasks and publishing it here.
It is my fervent hope that any that decide to go down the UPIM path share their discoveries and experiences gained for the benefit of all.

regards
bugwhiskers

[johnno]

Hello BW, Considering all the work you have done on your project, are you planning on building this to a workable PI.? I have read earlier posts as this was to "interface for PI's such as Gary's and Hammerhead".
Have I missed something?
Keep up the brilliant work!
johnno

[gef12]

Takers

Hi BW
Ye am sure you will get a few plus takers on this one .. including myself..
cheers 4 now
gef

[johnno]

A lot depends on the number of people interested in it as to whether PCB's are to be made professionally and sold.
As for cost,(all figures are approximates)

Micro $7
LCD screen $15-$150
PCB's $30
Other components $30

regards
bugwhiskers

I would definately be interested in buying one!
johnno

[bugwhiskers]

hello how soon is this to be in kit form?
any idea of cost yet?

A lot depends on the number of people interested in it as to whether PCB's are to be made professionally and sold.
As for cost,(all figures are approximates)

Micro $7
LCD screen $15-$150
PCB's $30
Other components $30

regards
bugwhiskers

hello how soon is this to be in kit form?
any idea of cost yet?

[bugwhiskers]

Hi gef12,

Thanks for your interest in the project.

The first preamp, the NE5534 has a gain of 1000 and provides most of the amplification. The AD8055's claim to fame is it is very fast and can drive low impedance devices like a cap that needs fast charging/discharging. The combined capacitance of the 64 caps is 64 * 0.1 uF or 6.4uF worst case which is quite large and must be done in 6.4uS at the fastest sample rate. Essentially the 2nd preamp is just a buffer so as to not load the first stage. Now that the offset adjustment is going to the 2nd preamp I will be able to have some gain on it also. Because the AtoD converter must be fed a signal that ranges between GND and +5 volts the offset circuitry must ensure the signal doesn't go too far below GND.

Because this is a project that is still on the fly there will no doubt be some changes along the way.

Hoping this explains your question.

regards
bugwhiskers

[gef12]

coil preamp

Hi BW
Looking at your coil preamp schematic I noticed 2 op amps .. seem to be running in parallel the AD8055 and NE5534. Is this so , is some sought of dual sampling going on here and then maybe averaging it out at a later stage. ??
Just curious. The finish product is looking ok.
gef

[bugwhiskers]

More pics

The attached pics are the reasoning behind the shape and size of the UPIM circuit boards. The plastic (polycarbonate methinx) are readily available, relatively cheap and light weight, they are also very tough and with the right sealing, rainproof too. The handle and fittings are also readily available and cheap.
The detector shown is an early design. The batteries can just be seen mounted on the underside of the box. The weight of the whole detector inc batteries and coil was just over 1kg, no steroid treatment required to allow a user to swing it all day.

regards
bugwhiskers

[bugwhiskers]

Coil-Preamp pic

The attached pic is of the revised coil-preamp board.

regards

bugwhiskers

[bugwhiskers]

Revised Coil - Preamp board

Attached are a revised schematic and pcb for the coil - preamp board.
The most important change is removal of the components for the second coil thus making room for a "poor man's" DAC. This DAC is driven serially by the micro and makes threshold adjust possible via software and the pushbuttons. It will be possible also for automatic adjustment via the AtoD reading a very late sample and the micro adjusting the threshold to ground potential. With the component values shown, swinging the DAC from 0 to 5 volts produces a +0.5 to -0.5 voltage change at the output of the last preamp.

It is very early days but with the 24 bit sampler board hooked up it was possible to see on the LCD screen graph a very distinct difference between fast targets (al foil, small coin, small nugget) and 2 quite different samples of ironstone (hot rock). It would seem that ironstone has a very distinct signature.


regards

bugwhiskers

[bugwhiskers]

Good news

I have worked out a way to get 64 samples and still use only DIP chips (not SMD). If you look at the previous photo you will see room above the CPU. I intend to make a board that "piggy backs" onto the sampler board and sits over the CPU. Six 74HC4051's and the associated caps just fit ! The front board will easily house the remaining two 74HC4051's, 74HC4053 & 74HC138. There will be room also for a dual OP amp to buffer the signals coming in and going out.

regards
bugwhiskers

[bugwhiskers]

Hi Ray-NM,

Thanks again for your interest in the project.
All is coming along well despite shortage of time due to paying work commitments. The routine that gets the current sample of each of the caps and compares it to the average for that sample is in place and working as expected. This routine makes the switched cap PI a motion detector, the faster you sweep over the target the greater the difference between successive samples.

The small PCB for 3 push-buttons is finished and today I will be writing the routine to read the buttons. Each button press generates a unique voltage that is read by one of the micro's AtoD inputs. Originally there was going to be 5 buttons but 3 take up less space and cost and is just as functional. The outer buttons = left or right(navigation around screen) or up or down(increment/decrement values) and the middle button = accept.

regards
bugwhiskers

[Ray-NM]

Bugwhishers
If everything works as you think then that will be one awsome Detector. been following this thread for awhile. Get it built and let us know how well it works.
RayNM