Hello everyone!

I've been having a problem with a prototype whereby one of the 555 timers (bistable) is false triggering. It appears to be due to EMI, because I can reproduce the problem by touching a long unconnected test lead to Vcc, or by rapidly flipping nearby light switches.

I've isolated the problem portion of the circuit, and can reproduce the bug on a breadboard. It looks like this:

The trick is that the voltage divider R1-TR1 is intended to pull the trigger low when ambient temperature rises above some level. TR1 is a 3103 thermistor. This appears to be the source of the problem. Am I using this in a foolish way?

In the test circuit above, warming up TR1 lights the LED, as intended. However, touching a long disconnected test lead to Vcc will also cause the LED to light.

The power supply to the circuit is from mains through a 7809 regulator with massive filter caps, just to eliminate noisy power as a variable.

I know 555's are prone to this sort of thing, and I've scoured the web and these forums for a similar issue. I can't seem to find any fix that works in this specific case, though. Any suggestions would be greatly appreciated! I'm out of ideas on this one.

- Veronica

 

Try attaching a small cap between the trigger (pin 2 ) and ground. Start with 0.1uF and work up or down from there.

 

Try attaching a small cap between the trigger (pin 2 ) and ground. Start with 0.1uF and work up or down from there.

I gave that a shot, and it seems to have worked! It cured the problem on the breadboard, so next I'll hack it into the prototype to see if it does the trick in the real world.

Thanks!

 

Maybe not relavent, but.

7809 regulator with massive filter caps

Do you also have the decoupling/bypass caps suggested by the datasheet
of the 7809?

How about a bypass cap of .1uF connected close to the VCC pin
of the 555 timer? IIRC another cap 100-470uF in parallel to the
.1uF is recommended for proper operation of the 555.

See this:
Decoupling or Bypass Capacitors, Why?

 

When using NE555 always use 10-100uF electrolytic capacitor between pin-8 (Vcc) and pin-1 (GND). Watch the polarity of the cap! If the problem persists try replacing NE555 type with CMOS version such as TLC555. Regular 555 are notorious for generating excessive noise on the supply line.

 

Maybe not relavent, but.

Do you also have the decoupling/bypass caps suggested by the datasheet
of the 7809?

How about a bypass cap of .1uF connected close to the VCC pin
of the 555 timer? IIRC another cap 100-470uF in parallel to the
.1uF is recommended for proper operation of the 555.

See this:
Decoupling or Bypass Capacitors, Why?

Thanks, I appreciate the tips! I do indeed have the bypass caps on the 7809 in addition to the big filters. I also have the bypass on pin 8, as you say. Taking no chances, for sure. I also have a 100uF cap on the power rail of the prototype, within about 5cm of the 555. Theoretically my bases are covered there. Even with all that, the ghost in the machine still lives. JoeyD's idea is showing real promise though.

 

Just to follow-up guys, the .1uF cap to ground on the trigger seems to be doing the trick. I put it in the real device, and it has been operating perfectly for a few hours now. None of my usual tricks to cause the malfunction are working. Looking good!

Thanks a ton for everyone's help! This forum is great.

 

Just to follow-up guys, the .1uF cap to ground on the trigger seems to be doing the trick. I put it in the real device, and it has been operating perfectly for a few hours now. None of my usual tricks to cause the malfunction are working. Looking good!

Thanks a ton for everyone's help! This forum is great.

Good for you. Just out of curiosity...is your thermistor on long wire? If so, using a twisted pair may help as well.

 

Good for you. Just out of curiosity...is your thermistor on long wire? If so, using a twisted pair may help as well.

Fortunately, the thermistor is mounted on very short leads, so that's one less thing to go wrong. Good idea, though!