Timeline for Circuit activated by pulse
Current License: CC BY-SA 3.0
17 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 22, 2017 at 21:03 | vote | accept | George Y. | ||
| Aug 22, 2017 at 0:26 | comment | added | Scott Seidman | You should put up timing diagrams of exactly what you want, or we're just guessing | |
| Aug 22, 2017 at 0:17 | answer | added | Vince Patron | timeline score: 4 | |
| Aug 21, 2017 at 23:53 | comment | added | jonk | What's the voltage rail and what is providing the input that is "low?" (SPST pushbutton switch? Or what, exactly?) | |
| Aug 21, 2017 at 22:02 | comment | added | George Y. | My circuit is DC, there's no AC. Solenoids are also DC-powered. I have flowback diodes there too. The main problem is debugging those issues, which seem to require equipment I don't have. | |
| Aug 21, 2017 at 21:51 | comment | added | rdtsc | "Easier" isn't always "right." In fact, it's pretty much guaranteed to be the opposite - the more thought put into something, the better it will work. I'd suggest studying your micro's watchdog feature, and experiment with it on a breadboard (without anything else attached) to ensure it works as expected. Then study AC Electronics and inductors to get a feel for how to "harden" your design against inductive kickback and other transients. Also research common-mode chokes and ferrite beads (especially the latter if you suspect RF EMI.) | |
| Aug 21, 2017 at 21:19 | comment | added | George Y. | @MarcusMüller I assume the problem is EMI/power issues, but fixing those is way above my head and tools I have. Easier to add a second watchdog, this is a hobby project. | |
| Aug 21, 2017 at 21:15 | comment | added | vicatcu | Or use a secondary micro as an external watchdog, which is what I do on my Wildfire dev board (attiny85 as watchdog, running TinyWDT firmware) | |
| Aug 21, 2017 at 21:13 | comment | added | D.A.S. | consider CM chokes on inductive twisted pairs, improved grounds and shields with C feedthru shunt caps to cure EMI. Any CMOS 1shot design for 5 seconds can work with adequate pulse width to set timer. | |
| Aug 21, 2017 at 21:11 | comment | added | Marcus Müller | then I'd rather fix my watchdog timer code than add another source of bugs (an analog watchdog). It is extremely unlikely your microcontroller's WDT is buggy, compared to the likelihood of a self-build analog watchdog to malfunction, if you ask me. | |
| Aug 21, 2017 at 21:08 | comment | added | George Y. | I have a microcontroller with watchdog and it gets stuck. Ended up watering my lawn for three hours today :) | |
| Aug 21, 2017 at 21:07 | comment | added | Marcus Müller | or: buy a microcontroller with a watchdog. This is fairly common to have in hardware. | |
| Aug 21, 2017 at 21:07 | comment | added | Transistor | Note that with your specification so far that a 0.1 Hz signal, for example, will cause it to disable for a fraction of a second and re-enable on the next clock edge. | |
| Aug 21, 2017 at 21:07 | comment | added | George Y. | Unfortunately the circuit is supposed to protect against microcontroller which got stuck :) so ideally it would be just passive elements. | |
| Aug 21, 2017 at 21:05 | comment | added | Marcus Müller | easy with a microcontroller. Be aware that you need to wait at least 5 whole seconds to be sure there's no further 0.2 Hz signal. If you don't want a single isolated pulse to trigger, you also need to wait 5 seconds to be sure the signal is periodic. It's often not easy to even specify what constitutes desired behaviour. | |
| Aug 21, 2017 at 21:05 | comment | added | D.A.S. | solution is a 5 second retrigerable one shot design | |
| Aug 21, 2017 at 21:02 | history | asked | George Y. | CC BY-SA 3.0 |