Tweet
hello to one and all.
for the last few weeks I have been working on a new creation,
a pi detector with a few differences.
it's nothing magical,
it's just a pi detector.
it doesn't discriminate,
I've just spent some time doing some things just that little bit differently.
yes,
I could have just made a surf pi, or a barracuda,
and I did breadboard many ideas and variations from these detectors.
was never happy with the "scratchy" audio a lot of diy detectors give,
so that was one point to address.
motion detection is also fine,
and good in slightly "noisier" environments but it does have a slight depth
penalty compared to a "pre-set" level non-motion detector.
but then regular adjustment is required.
swings and roundabouts and all that.
so decided to design and build a detector with motion and non-motion modes,
poss done things a bit quirky,
and a little out of the box,
but,
it's in a box, and it works.
phew,
bonus there then.
little apprehensive posting a self designed project as it will be subject to
many different responses , positive and negative.
anyways,
circuit diagrams below,
i'm not a whizz laying out diagrams, so there all on separate sheets.......
brief basic explanation:
1st sheet.
used 2x555's for the pulse to keep pulse rate and on time separate.
used TS555's for lower power usage.
freq is about 1500hz.
with 70us on time, good compromise.
used a 1N4937G diode between the coil and mosfet to isolate the capacitance
of the IRF840.
the timing is pretty standard (aka surf pi )
2nd sheet.
2 stage pre-amp, been mentioned allot over the last couple of years,
non-inv setup appears to be faster than the inverting setup,
oh,
and a diode across the first stage to speed it up a little more.
used TL072's I like e'm and they seem to fast enough for the job.
rest of the op-amps are TL062's,
low power usage,
and,
unlike the TL072's can go to 0.65v from the rails.
earth field cancelling is something I saw on a post way back from sacho ( Bulgaria )
elegant solution I thought.
didn't go for the sampling intergrator method,
actually quite noisy I found ..........
the detection part of it is what's just that little bit different.
basically,
the input voltage level is held on the balanced 1uf caps,
buffered to the final op-amp used as just an on / off comparator,
whilst the + input has about 10mv removed to keep the comparator off.
the hysteresis is corrected / controlled by the diodes to be as near constant
over the input voltage range.
think that cleared up that section........
I think.....
oh nearly forgot, when the final comparator is on (bleep), the reference caps get isolated
via the 2n7000 and the 4066 so there's no "creep".
oh yes,
and there's a zero function when in non-motion mode.
3rd sheet
audio.
a 2 stage differential amplifier controls the audio pitch from the last 555 via a homemade
opto resistor isolator.
when set correctly as soon as the detector starts to beep the pitch rises with target intensity.
the volume is constant.
tones at rest is about 450hz, rising to about 1000hz peak.
did this so the "double wobble" of rings,
yeah,
and nails can be much more easily heard.
erm,
there's a battery alarm via a flashing led,
simple elegant solution there I think.......
I used flashing blue, it was the only colour I had......
who care's it's inside the box,
just there to provide a pulsing tone.....
audio output has "soft start" cap arrangement (4.7k and the 4.7uf)
more pleasant on the ears.....
and the top bc337 copies the 8v line (-0.7v) and draws it's currant
from the battery line so as not to load up the 8v line when it beeps.
battery check is via a pretty standard setup with an LM339
via test button,
doesn't use anything unless you press the button,
dam you power hungry led's !!!
4th sheet.
finally the power supply.......
its just a regulator, enough said there then.....
4v and 8v,
4v for the coil,
8v for everything else.
this setup with run quite happily down to 8.1v.
so nearly 1v per cell,
suck them buggers dry.......
the detector only uses about 42ma when not bleeping,
with coil detached the circuitry only takes 7ma,
another key criteria of design.
the coil running on only 4v was a discovery I made a while back,
the depth capability seems to be governed by coil on time more than
the voltage the coil is driven by.
in my experiments 70us on time is a good compromise between power
usage and depth,
and all 12v did for me was increase the power usage.
anywho,
been a long post,
i'll round up with some pics of my finished detector,
if anyone's still reading......
OOPPSS !!!
there's a 10 picture limit on a post,
i'll post the rest asap.....
for the last few weeks I have been working on a new creation,
a pi detector with a few differences.
it's nothing magical,
it's just a pi detector.
it doesn't discriminate,
I've just spent some time doing some things just that little bit differently.
yes,
I could have just made a surf pi, or a barracuda,
and I did breadboard many ideas and variations from these detectors.
was never happy with the "scratchy" audio a lot of diy detectors give,
so that was one point to address.
motion detection is also fine,
and good in slightly "noisier" environments but it does have a slight depth
penalty compared to a "pre-set" level non-motion detector.
but then regular adjustment is required.
swings and roundabouts and all that.
so decided to design and build a detector with motion and non-motion modes,
poss done things a bit quirky,
and a little out of the box,
but,
it's in a box, and it works.
phew,
bonus there then.
little apprehensive posting a self designed project as it will be subject to
many different responses , positive and negative.
anyways,
circuit diagrams below,
i'm not a whizz laying out diagrams, so there all on separate sheets.......
brief basic explanation:
1st sheet.
used 2x555's for the pulse to keep pulse rate and on time separate.
used TS555's for lower power usage.
freq is about 1500hz.
with 70us on time, good compromise.
used a 1N4937G diode between the coil and mosfet to isolate the capacitance
of the IRF840.
the timing is pretty standard (aka surf pi )
2nd sheet.
2 stage pre-amp, been mentioned allot over the last couple of years,
non-inv setup appears to be faster than the inverting setup,
oh,
and a diode across the first stage to speed it up a little more.
used TL072's I like e'm and they seem to fast enough for the job.
rest of the op-amps are TL062's,
low power usage,
and,
unlike the TL072's can go to 0.65v from the rails.
earth field cancelling is something I saw on a post way back from sacho ( Bulgaria )
elegant solution I thought.
didn't go for the sampling intergrator method,
actually quite noisy I found ..........
the detection part of it is what's just that little bit different.
basically,
the input voltage level is held on the balanced 1uf caps,
buffered to the final op-amp used as just an on / off comparator,
whilst the + input has about 10mv removed to keep the comparator off.
the hysteresis is corrected / controlled by the diodes to be as near constant
over the input voltage range.
think that cleared up that section........
I think.....
oh nearly forgot, when the final comparator is on (bleep), the reference caps get isolated
via the 2n7000 and the 4066 so there's no "creep".
oh yes,
and there's a zero function when in non-motion mode.
3rd sheet
audio.
a 2 stage differential amplifier controls the audio pitch from the last 555 via a homemade
opto resistor isolator.
when set correctly as soon as the detector starts to beep the pitch rises with target intensity.
the volume is constant.
tones at rest is about 450hz, rising to about 1000hz peak.
did this so the "double wobble" of rings,
yeah,
and nails can be much more easily heard.
erm,
there's a battery alarm via a flashing led,
simple elegant solution there I think.......
I used flashing blue, it was the only colour I had......
who care's it's inside the box,
just there to provide a pulsing tone.....
audio output has "soft start" cap arrangement (4.7k and the 4.7uf)
more pleasant on the ears.....
and the top bc337 copies the 8v line (-0.7v) and draws it's currant
from the battery line so as not to load up the 8v line when it beeps.
battery check is via a pretty standard setup with an LM339
via test button,
doesn't use anything unless you press the button,
dam you power hungry led's !!!
4th sheet.
finally the power supply.......
its just a regulator, enough said there then.....
4v and 8v,
4v for the coil,
8v for everything else.
this setup with run quite happily down to 8.1v.
so nearly 1v per cell,
suck them buggers dry.......
the detector only uses about 42ma when not bleeping,
with coil detached the circuitry only takes 7ma,
another key criteria of design.
the coil running on only 4v was a discovery I made a while back,
the depth capability seems to be governed by coil on time more than
the voltage the coil is driven by.
in my experiments 70us on time is a good compromise between power
usage and depth,
and all 12v did for me was increase the power usage.
anywho,
been a long post,
i'll round up with some pics of my finished detector,
if anyone's still reading......
OOPPSS !!!
there's a 10 picture limit on a post,
i'll post the rest asap.....
Comment