Can someone please explain step by step to me how the oscillator works in the picture, i've built it and was just really curious as to how it works
Just guessing but it looks to me like the upper NAND gate triggers the first count then the lower NAND gate triggers the successive countrs, but the upper left led will be always on?
no, the led flashes 10 times a second (10hz) the output from the two schmitt nands is joined toghether, i guess what i was curious about is, how do the nands go from active low to active high and how do they retrigger each other, kinda complicated for me, but i'm sure one of the peeps round here can satisfy my curiosity
i actually removed the upper led and replaced it with a solder join because it looks pretty pointless, i joined the connections in place of it because i wasn't sure if it needed to be connected that way to oscillate
As far as I can see, the 1uF cap charges up through the 1M resistor until the voltage is high enough to trigger the Schmitt, then discharges again until the Schmitt drops out. So you get a sawtooth wave on the Schmitt input, square wave on the output. You can make a similarish oscillator with two inverters, a couple of resistors and a cap, like this one. edit: just looked again, the one below the linked circuit is more like yours.
cheers cpemma, i'll have a proper look at it when i get access to an oscilloscope (when i get back to college on the 1st)
Some info in Nat Semis Application Note 88 (AN-88) on using logic chips for linear stuff. Shows the Wolfstone circuit. Yours is oddball, does it work with the led open-circuit? It might be doing the job of the second resistor.
It looks to be a basic relaxaion oscillator. The transient responce of the 1 uF cap and the 1M resistor set up the basic timing. The two schmitt triggers for a bistable element, and condition the poorly shapped RC curve into a square wave. I belive that the 47uF cap is there to filter, and condition the power (I just saw a problem with a PIC that was fixed by this very configuration) and the LED and resistor appear to be there to pull down the input when the schmitt NAND goes low. It could be done with just a transistor, but the LED is just so much flashier.
At first I was confused by the picture, but just figured it out. The two NAND gates act independantly. The one with the 1M R and cap (pins 1,2, and 3) is the timing circuit. When both inputs are low (they are tied together) the output goes high. The cap charges up and then your inputs go high. Once the inputs go high enough, the outputs go low and the cap and 1M R get discharged till it reaches the threshold for the low input, and so on and so on. I was confused for a second as to why the output of the second NAND gate (pins 4,5, and 6)was tied to the LED, and its inputs ties to the clock signal. If you remove the LED the circuit will keep doing its thing. The LED is their to just show indication. Its inputs are tied to the Astable clock signal, and its output drives the LED. Removing the LED wont do anything, but you DO NOT want to place a short in its place. Leave it open.
oops, Hazer is right. I goofed on seeing VCC there. I also jumped to the conclusion that it is like the bistable element based oscillators I've seen and built in the past.
Yeah. When i first looked at it, I thought the second NAND gate was supposed to be part of the timing circuit. But it confused me as to why it was driving backwards but the 4093 is a quad NAND gate and this schem doesnt show that the gates are powered from another pin not designated in the schem. Then I realized that the second NAND was just taking the clock signal to drive the LED for indication . The second NAND gate does not need to be wired at all for this circuit. Also, putting a short in its place means that all it will do is just send unneeded current through the resistor. I recently used this circuit (the 4017) with a timer (555) to build a 3 LED sequencer for an acrylic sculpture that I gave my GF. I should post the schem I came up with. It flashes 3 Ultra bright LEDS in sequence, but uses caps to fade (bit-tech tutorial guide props) but also added a unique feature. The blue led lights first, then the second output lights both red and blue led, the third lights just the red led, then 4th lights red and yellow led, etc etc . They all fade into each other. The collors are pretty cool. In order to do this, I used 3 2N3904 to drive the LEDs from 5V source, but used resistor and caps to attach to the outputs of the decade counter. In order to light 2 LEDS from one output without crossing outputs, I used 1N4002 rectifiers for each output. 2 rectifiers if the output was to light 2 LEDs. The circuit came together nicely. A little playing around with breadboard to get the right values for timing/fading/brightness, then spent 3 hours of wirewrapping and my patent 'raping old and broken stuff for caps/IC/diodes I could use' and then cutting out a plastic base to house the thing, and walla: cool LED lighted base. Even got the wall wart with plug off of something else. Im trying to get my personal webspace up so I can start adding pics.
