PI CHIP

[moodz]

moodz

i bought the hc-06, i'm going to get a hc-05, also not very impressed with thr nano usb adapters either....

is the hc-06 the master / slave version ? what was wrong with the usb adapters ( fell apart ?? )

cheers.

[moodz]

..small update on the detector project ... we are having alot of success with detecting small targets.

here is the frontend performance on a 1 gram test gold nugget ( its about 8 x 6 x 1.5 mm )

The cursor is set to where the first sample in the detector could begin at 5.4 microseconds ( we can sample down to 4 microseconds if I change out the amp )

The sample is placed one inch above the coil centre. Its a nuggetfinder 0.4 ohms etc etc.

First Pic shows the output of the amp with some nasty overshoot ( I am using an ne5534 ) Our gain is set to 100 ( this is the first amp in the signal chain ).

The second Pic zooms in to 100 mv per division on the CRO but the same timebase.

75 Millivolt deflection ....not bad if I say so myself .... Cannot even see the deflection at the first amp on our reference commercial detector ( though it does easily detect the nugget at this range ).

the signal to noise for these small targets is orders of magnitude higher than the same measure point on a commercial detector.

[hdphilip]

The HC-06, seems to be a slave only, not alot of programming options. even though i could access the AT command set, i wasn't able to change the baud rate from 9600, so i changed the baud rate in your program to 9600, reloaded it into the uno, and it worked.

the usb bluetooths i got, well, it's maybe best to get a "good one" instead of 2 for 3 dollars, free shipping from china..... They were deaf to discovry my bluetooth toys



on your detector project, are you going away from the 2 to 3 microsecond sampling then?


philip

[moodz]

That's interesting that they are limited like that ... Though for control 9600 is plenty.
I am fairly happy to get around 4 to 6 microseconds on these commercial coils ... Tho the Nuggetfinder coils are very good anyway and I am using relatively high tax current and long tx length at 200 ... I could do a lot better if I knock down the tx to say 60 microseconds. The production prototype detectors do better but I dont / can't post anything to do with them. I will publish a ground pic .

[moodz]

That's interesting that they are limited like that ... Though for control 9600 is plenty.
I am fairly happy to get around 4 to 6 microseconds on these commercial coils ... Tho the Nuggetfinder coils are very good anyway and I am using relatively high tax current and long tx length at 200 ... I could do a lot better if I knock down the tx to say 60 microseconds. The production prototype detectors do better but I dont / can't post anything to do with them. I will publish a ground pic .

heres the ground pic .... from some of the worst ground in OZ. ( see magnet pic ... you dont have to move the magnet through the dirt .... its all magnetic )

The pix show the behaviour over really bad ground ... huge response from the ground ... the 1 gram nugget still provides a clear signal though
the peak coil current is 3 amps.

[moodz]

.. tc infos for a couple of targets over bad OZ ground. Its interesting to compare the response of two smallish targets over bad OZ ground.

[hdphilip]

moodz

Humm, looks like your going to have the ground balnce circuit/ software self-adjusting just right, have you thouht about usuing a accelerometer to help the ground balance?

i thought about it for my "Qed inspired sheild board" i'm working on, but it could go on my 3rd layer controls board.

those sampling circuit "glitches" look pretty nasty,



Philip

[moodz]

moodz

Humm, looks like your going to have the ground balnce circuit/ software self-adjusting just right, have you thouht about usuing a accelerometer to help the ground balance?

i thought about it for my "Qed inspired sheild board" i'm working on, but it could go on my 3rd layer controls board.

those sampling circuit "glitches" look pretty nasty,



Philip

Actually ground balance is almost a trivial exercise ... my company is testing the Howard Rockey GB patent method.

http://ip.com/pdf/patent/AU2010101019B4.pdf

The accelerometer idea can be used for another purpose though .... however I wont make it public just yet.


Yes the sampling glitches are on a test rig and not present usually .. I keep letting the smoke out of surface mount version with missprobes so I stopped doing it.

[hdphilip]

a few test points outa do the trick...

one of the things i"m doing on my board is having my 16 bit adc circuit, with a switch so i can look at the intergrator voltage, or the snubber voltage, it'll work to fine tune, and it may help in troubleshooting the software writing.
I'm usuing a LT1372 for the V-, that chip works works well. it'll be controlled by the pic.

Philip

[fmtemp]

Schematic front end

Hi everybody,

I am very impressed by the results you obtain.
I am interresting in testing all your experiments and i am wondering if it would be possible for you to post the front end schematic, and if possible the pcb, you are using for obtazining such results ?

Thank you very much for your job that seem to be really performant.

FM.

[hdphilip]

got a new bluetooth adapter (Asus BT211), bluetooth between the uno and my computer works Great!



Also replaced my dual monitors, now their matching(great for programming and PCB layout), and a new video card to drive them

philip

[moodz]

...you won the lotto or something

[moodz]

We are still working out how to do that.

[hdphilip]

...you won the lotto or something

can't win the lotto, all the money i save on tickets, I buy toys with

[hdphilip]

moodz,

on your MCP3903, what freq is the xtal? and are you usuing a differential op amp driver configured as a single ended to differential conversion for the adc frontend?

I'm planning on trying it on my qed project.

but in the meanwhile, I'm putting together a few circuits on a pcb to send to the pcb fab house as a "test" board, and i'm including a circuit for the mcp3901

philip