How about this A capacitive approach Have for each page a aluminium strip cover two plates of which one is connected to ac and the other one to a sense amp (analog input). If you can ense the ac the strip/dot on the page is covering the plates. Peter van Hoof --- Robert Rolf wrote: > Within first two hits on google. >=20 > http://innovexpo.itee.uq.edu.au/2001/projects/s369402/ > http://innovexpo.itee.uq.edu.au/2001/projects/s369402/thesis.pdf > lots of references along with a dead simple (=3DCHEAP) > detection method. >=20 > You really OWE him a royalty for using his solution. > See Fig 4. > It's basically plug and play for your application > (although > the code is for an AVR). So where is my finders fee > ? >=20 > I would also make the coils on BOTH sides of the > PCB, rather > than the single side he uses, to gain sensitivity. > Be aware that PCB inductors are the subject of a USA > patent, > which should expire soon. >=20 > I also found this patent on a more modern use of > planar coil > inductive sensing of objects (1995) > http://www.patentstorm.us/patents/5390109.html >=20 >=20 > Robert >=20 > This is from the thesis (URL 2 above). > "If the original waveform and the delayed version > could somehow be=20 > compared with > each other, this could avoid measuring times on a > nanosecond scale. An=20 > Exclusive-Or > gate would do the job of comparing the waveforms and > where the two=20 > waveforms do > not match, the gate returns a logical 1. A > rectangular wave with a=20 > frequency of twice > the input frequency is thus produced. The duty cycle > of this wave=20 > depends on the > degree of delay between the two inputs. >=20 > To measure the duty cycle of XOR output, the > signal was simply passed through a low pass filter > to obtain the DC=20 > component. The > voltage was now simply measured with an Analogue to > Digital Converter (ADC). > Figure 5 is a diagram of the complete circuit." >=20 > 4.1.2 Eddy Current Sensing > The inductive sensing scheme proposed is based on > eddy currents. The sensing > elements of the chessboard are simply planar coils > formed by a PCB=20 > track. The chess > pieces have aluminium foil on their underside. When > a chess piece sits=20 > on a coil, the > alfoil essentially =91shorts=92 the coil and changes the > properties of the=20 > inductor. The coil > is driven by an alternating current and induces eddy > currents in the=20 > conductive alfoil. > These currents travel in an opposite direction to > the current in the=20 > coil, this reduces the > magnetic flux in the coil and thus its inductance. > The eddy currents=20 > also dissipate > energy, increasing the coil=92s resistance. The > chessboard must somehow=20 > measure these > characteristic changes in the coil. > 4.1.3 Planar Coils > The coils themselves are simply PCB tracks. There > are many advantages to=20 > using PCB > tracks over external coils. A chessboard has 64 > squares and each=20 > square=92s sensing > Chess on a PCB: Inductive Proximity Sensing using a > Microcontroller > 17 > ability should be comparable to the next. By using a > reliable PCB=20 > manufacturing > process, 64 practically identical coils can be > easily and cheaply made.=20 > The planar > nature of the coils means that all of the turns are > as close to the=20 > chess piece as possible, > maximising sensitivity. This concept was the driving > idea of this=20 > thesis, to play =91Chess > on a PCB=92. >=20 >=20 > alan smith wrote: >=20 > > Well....the client was really hooked on hall > effect sensing. So I hooked up a demo circuit, and > got some thin strip magnets (have to be thin) and > found that based on where the circuit board is > mounted, the magnets were too far away and didnt > have enough flux density to them, unless I used a > thicker magnet. That of course, didn't play into > the overall design, as the magnet would end up being > too large > >=20 > > so.... > > =20 > > As Mr Rolf pointed out...there are other > choices.... > > =20 > > The biggest issue with using ambient light to a PV > type sensor is that it would be easy to be covered > or not have enough light and give a false reading. > > =20 > > But, what about doing something with reflective > IR? in that....if there is a hole above an IR LED, > nothing would reflect. But I would need something > that would emit and detect...are there such devices > that are also inexpensive?=20 > > =20 > > I like the idea of the coils, did a search for > chess board piece sensing, came up empty as far as > circuits. =20 > > =20 > > I'd like to look into this....any pointers to > where some information and design data might be? > >=20 > > Robert Rolf wrote: > >=20 > > Your sensor choices seem to be magnetic, optical, > mechanical. > >=20 > > If the number of pages is small, then 11 optical > sensors > > along one of edges to detect page tabs seems to be > a clean > > solution. Use LED carrier modulation and the 11th > sensor > > to remove ambient light concerns. I don't think > you can do this for $3 > > but I'm not into mass manufacturing so it may be > doable. > >=20 > >=20 > >=20 > > You could also look at coils that are HF excited, > and have metallic > > disks on the pages that would change the coil Q or > L. This would > > make the lumps in the pages much less noticeable. > You could binary > > code the sensors because the more iron you throw > above a coil, > > the more the L/Q shifts. And using different > metals would give > > you different detuning as well. An analog mux and > oscillator are > > all you'd need to read your coil array. Look for > 'chess board' > > 'piece' 'sensing' for existing ccts to do this. > >=20 > >=20 > >=20 > >=20 > > =09 > > --------------------------------- > > Yahoo! FareChase - Search multiple travel sites > in one click. =20 > --=20 > http://www.piclist.com PIC/SX FAQ & list archive > View/change your membership options at > http://mailman.mit.edu/mailman/listinfo/piclist >=20 --=20 http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist