Hi Group, Forgive me if this is in the wrong newsgroup - I didn't see an "analog signal processing" group. I have a signal coming in from a photodiode that is situated underneath a pendulum (eventually, this will be a foucault pendulum) in order to detect the pendulum bob when it swings over the center point. Since I have four bright white LEDs facing up to reflect light off of the bottom of the pendulum when it swings, there is a spike of voltage when the pendulum passes over the photodiode. Here comes my question, finally. I'd lke to be able to detect when the pulse happens and send out a TTL high so that I can trigger a magnet driver. Any ideas? Tips? Thank you!
Basic (Analog) Signal Processing Question
Started by ●February 17, 2006
Reply by ●February 17, 20062006-02-17
On Fri, 17 Feb 2006 07:35:54 -0800, yellowbkpk wrote:> Hi Group, > > Forgive me if this is in the wrong newsgroup - I didn't see an "analog > signal processing" group. > > I have a signal coming in from a photodiode that is situated underneath > a pendulum (eventually, this will be a foucault pendulum) in order to > detect the pendulum bob when it swings over the center point. Since I > have four bright white LEDs facing up to reflect light off of the > bottom of the pendulum when it swings, there is a spike of voltage when > the pendulum passes over the photodiode. > > Here comes my question, finally. I'd lke to be able to detect when the > pulse happens and send out a TTL high so that I can trigger a magnet > driver. Any ideas? Tips? > > Thank you!Easier to get a logic 0, by using an NPN transistor to amplify the voltage spike (assuming it is > 0.7V). However, a comparator IC is probably the better alternative. It'll pull the output to ground when the - input is higher than the + input. So, set a voltage reference using a trimmer pot, connect it to the Vin+ input, and connect your sensor to the Vin- input. Then a voltage spike on the Vin- input will pull your output to ground. Use a 'pullup' resistor of 10k or so between output and Vcc, since most comparators are 'open collector', although if you are interfacing with TTL, you may not need the pullup. Add some 'hysteresis', by using a 10k resistor between reference and Vin+, and a 100k resistor between Vout and Vin+. Try sci.electronics.basics for more information. -- Regards, Bob Monsen It is the duty of every citizen according to his best capacities to give validity to his convictions in political affairs. Albert Einstein, 'Treasury for the Free World,' 1946
Reply by ●February 17, 20062006-02-17
yellowbkpk@gmail.com wrote:> Hi Group, > > Forgive me if this is in the wrong newsgroup - I didn't see an "analog > signal processing" group. > > I have a signal coming in from a photodiode that is situated underneath > a pendulum (eventually, this will be a foucault pendulum) in order to > detect the pendulum bob when it swings over the center point. Since I > have four bright white LEDs facing up to reflect light off of the > bottom of the pendulum when it swings, there is a spike of voltage when > the pendulum passes over the photodiode. > > Here comes my question, finally. I'd lke to be able to detect when the > pulse happens and send out a TTL high so that I can trigger a magnet > driver. Any ideas? Tips? > > Thank you!I would think you are going to have ambient light problems. Instead of an optical sensor, maybe you can pick up a signal off the drive coil or connect a neagtive resistance ot the drive coil??? Mark
Reply by ●February 17, 20062006-02-17
Mark wrote:> yellowbkpk@gmail.com wrote: > >>Hi Group, >> >>Forgive me if this is in the wrong newsgroup - I didn't see an "analog >>signal processing" group. >> >>I have a signal coming in from a photodiode that is situated underneath >>a pendulum (eventually, this will be a foucault pendulum) in order to >>detect the pendulum bob when it swings over the center point. Since I >>have four bright white LEDs facing up to reflect light off of the >>bottom of the pendulum when it swings, there is a spike of voltage when >>the pendulum passes over the photodiode. >> >>Here comes my question, finally. I'd lke to be able to detect when the >>pulse happens and send out a TTL high so that I can trigger a magnet >>driver. Any ideas? Tips? >> >>Thank you! > > > I would think you are going to have ambient light problems. > > Instead of an optical sensor, maybe you can pick up a signal off the > drive coil or connect a neagtive resistance ot the drive coil???I once built a bullet-velocity meter whose front end consisted of two photocells six inches apart, looking up at the sky through slits. A bullet passing over the slit within a trapezoid about three inches wide at the box and about a foot wide a 18 inches over it cast a shadow sufficient to be detected. (The time difference between the two shadows and the distance between the slits allowed the speed to be calculated. A seven-segment display showed the velocity. A microprocessor would have been nice, but TTL and CMOS was all we had available at the time.) Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●February 20, 20062006-02-20
yellowbkpk@gmail.com wrote:> Hi Group, > > Forgive me if this is in the wrong newsgroup - I didn't see an "analog > signal processing" group. > > I have a signal coming in from a photodiode that is situated underneath > a pendulum (eventually, this will be a foucault pendulum) in order to > detect the pendulum bob when it swings over the center point. Since I > have four bright white LEDs facing up to reflect light off of the > bottom of the pendulum when it swings, there is a spike of voltage when > the pendulum passes over the photodiode. > > Here comes my question, finally. I'd lke to be able to detect when the > pulse happens and send out a TTL high so that I can trigger a magnet > driver. Any ideas? Tips? > > Thank you! >Its been so....long. I'm not even sure if they still make 555 timers, but it has all the stuff you need in 8 pins.
Reply by ●February 20, 20062006-02-20
Stan Pawlukiewicz wrote:> yellowbkpk@gmail.com wrote: > >> Hi Group, >> >> Forgive me if this is in the wrong newsgroup - I didn't see an "analog >> signal processing" group. >> >> I have a signal coming in from a photodiode that is situated underneath >> a pendulum (eventually, this will be a foucault pendulum) in order to >> detect the pendulum bob when it swings over the center point. Since I >> have four bright white LEDs facing up to reflect light off of the >> bottom of the pendulum when it swings, there is a spike of voltage when >> the pendulum passes over the photodiode. >> >> Here comes my question, finally. I'd lke to be able to detect when the >> pulse happens and send out a TTL high so that I can trigger a magnet >> driver. Any ideas? Tips? >> >> Thank you! >> > Its been so....long. I'm not even sure if they still make 555 timers, > but it has all the stuff you need in 8 pins.Refresh my core memory, but was not the 555 a versatile timer chip. I don't recall it having any Schmidt trigger inputs. In good old days I think I'd attacked OP's problem with a quad 704 in 16 pin package. Then again 3 or 4 CK722's would do nicely. Or if pendulum REALLY fast, 832A's (4CX150's too power hungry). Would any favor 6CW4's? OP, please don't take me too seriously or someone may throw an 026 at me, or even a Vermont Research drum ;} No one ever suggested I resembled a coherer, except not needing routine shocks.
Reply by ●February 20, 20062006-02-20
Richard Owlett wrote:> Stan Pawlukiewicz wrote: > >> yellowbkpk@gmail.com wrote: >> >>> Hi Group, >>> >>> Forgive me if this is in the wrong newsgroup - I didn't see an "analog >>> signal processing" group. >>> >>> I have a signal coming in from a photodiode that is situated underneath >>> a pendulum (eventually, this will be a foucault pendulum) in order to >>> detect the pendulum bob when it swings over the center point. Since I >>> have four bright white LEDs facing up to reflect light off of the >>> bottom of the pendulum when it swings, there is a spike of voltage when >>> the pendulum passes over the photodiode. >>> >>> Here comes my question, finally. I'd lke to be able to detect when the >>> pulse happens and send out a TTL high so that I can trigger a magnet >>> driver. Any ideas? Tips? >>> >>> Thank you! >>> >> Its been so....long. I'm not even sure if they still make 555 timers, >> but it has all the stuff you need in 8 pins. > > > Refresh my core memory, but was not the 555 a versatile timer chip. I > don't recall it having any Schmidt trigger inputs. In good old days I > think I'd attacked OP's problem with a quad 704 in 16 pin package.Its got a pair of comparitors if I recall correctly. You can google on 555 timer for yourself. I remember there was a time when EDN would always have 555 or a 556(dual 555) circuit that someone would submit. I don't have my copy handy, but the "Art of Electronics" book had a section of the 555.> > Then again 3 or 4 CK722's would do nicely. Or if pendulum REALLY fast, > 832A's (4CX150's too power hungry). Would any favor 6CW4's? > > OP, please don't take me too seriously or someone may throw an 026 at > me, or even a Vermont Research drum ;} No one ever suggested I resembled > a coherer, except not needing routine shocks. > >
Reply by ●February 20, 20062006-02-20
Richard Owlett wrote: ...> Refresh my core memory, but was not the 555 a versatile timer chip. I > don't recall it having any Schmidt trigger inputs. In good old days I > think I'd attacked OP's problem with a quad 704 in 16 pin package.OK: http://www.mitedu.freeserve.co.uk/Circuits/Timing/555mono.htm Jerry -- Engineering is the art of making what you want from things you can get. �����������������������������������������������������������������������
Reply by ●February 20, 20062006-02-20
Mark wrote:> yellowbkpk@gmail.com wrote: > >>Hi Group, >> >>Forgive me if this is in the wrong newsgroup - I didn't see an "analog >>signal processing" group. >> >>I have a signal coming in from a photodiode that is situated underneath >>a pendulum (eventually, this will be a foucault pendulum) in order to >>detect the pendulum bob when it swings over the center point. Since I >>have four bright white LEDs facing up to reflect light off of the >>bottom of the pendulum when it swings, there is a spike of voltage when >>the pendulum passes over the photodiode. >> >>Here comes my question, finally. I'd lke to be able to detect when the >>pulse happens and send out a TTL high so that I can trigger a magnet >>driver. Any ideas? Tips? >> >>Thank you! > > > I would think you are going to have ambient light problems. > > Instead of an optical sensor, maybe you can pick up a signal off the > drive coil or connect a neagtive resistance ot the drive coil??? > > > Mark >Yellowbkpk, To expand on Mark's idea, there is a very inexpensive solution that is often used by imitation "pendulum" quartz clocks. The pendulum contains a small magnet and there is a stationary coil which has two windings. One winding is connected between the DC supply and the collector of a transistor. The other winding connects between the base and emitter. The emitter connects to the other end of the supply. When the approaching magnet induces enough voltage into the coil to start a trickle of current in the base, the transistor pulls a short pulse of current. Once started, the process is self-sustaining, and the pulse terminates when the current reaches a maximum that depends on the resistance of the collector coil. A single-winding coil is obviously easier to make, and it is easier to get the polarity right (relative to the magnet polarity.)You could use a comparator handle both the voltage detection and the current drive. The coil would be connected between the supply and the comparator output. You will need to AC-couple from the collector / coil circuit to the appropriate comparator input. Regards, John
Reply by ●February 21, 20062006-02-21
Jerry Avins wrote:> Richard Owlett wrote: > > ... > >> Refresh my core memory, but was not the 555 a versatile timer chip. I >> don't recall it having any Schmidt trigger inputs. In good old days I >> think I'd attacked OP's problem with a quad 704 in 16 pin package. > > > OK: http://www.mitedu.freeserve.co.uk/Circuits/Timing/555mono.htm > > JerryUNCLE I stand corrected.
