555 Timer Triggering Too Easily

Question

I've got a 555 timer being powered by a 5v rail that is isolated from the 3v rail running a motor driver, which runs a motor.

The 555 is only supposed to fire when it gets the signal from a 38khz infrared detector but it seems to trigger everytime I plug a wire into the 3v rail, or sometimes for no apparent reason at all.

What is going on? Can I give it some buffer room somehow?

Extra info: when the 555 is triggered by the 38khz infrared receiver, it is supposed to tell the driver to spin the motor forward for about 6 seconds. However, the 555 also triggers an infrared LED to try and trigger a photoresistor, and if the light isn't blocked, then the phototransistor is supposed to disable pin 2 of the driver and enable pin 7, thereby reversing the direction the motor will spin for the 6 seconds indicated by the 555. this is so that the device can tell whether to run the motor forward or backwards based on the phototransistor's ability to see light.

a busy cat http://jackvero.com/pictures/schematic.png

Answer 1

The 5V bus needs capacitors across it, near the regulator, near the TSOP4838, and another one near the 555. Without it, there is ripple on the bus. The TSOP4838 will false trigger with more than 100mV ripple. This false trigger will trigger the 555. Switching the load on the 3V bus will cause sag in the 9V bus, which will be passed to the 5V bus via the regulator, before it can compensate. You should prove this by measuring the ripple before you add capacitors.

Answer 2

Also, regardless of your triggering issues, you have the 555 wired incorrectly. You have a timing capacitor (no reference designator) connected directly to pin 7 and through R1 to pin 6.

In this configuration, you are relying on leakage current through pin 6 to charge your capacitor once the 555 is triggered, which is going to give you very unreliable results.

The correct way to do this would be to tie the capacitor directly to both pins 6 and 7, and connect a timing resistor from that junction to the +5V rail.

Answer 3

My guess would be that the output of the 38kHz detector is actually an open drain or open collector, which would leave the trigger pin as high impedance and therefore its voltage level undefined. This can be address by adding a pull-up resistor on the trigger pin plus adding decoupling capacitor as previously mentioned would also help. Finally you could also look into adding a small RC filter on the trigger pin such as that the time constant will not noticeably affect the response of the circuit.