Did you think of trying to put a hood over the emitter or the sensor, such that they would be shielded as if nighttime (but still be able to see each other)? I have an old SLR film camera that started producing terrible pix, my father figured out it had a light leak, found and plugged that light leak, and restored it to perfect picture taking operation again. This saga reminds me of that problem / solution. ------------- Max Charleston SC
On Thu, Aug 19, 2021 at 4:42 PM Jim Cathey via Mercedes < mercedes@okiebenz.com> wrote: > Wednesday, August 4, 2021 <>I decided I had a bit of time to look into our > always-irritating Wayne Dalton garage door opener, which usually refuses to > go down in the day. It seems to correlate with light and/or heat, and you > have to bounce it down, sometimes inches at a time, with the remote. Very > irritating! [If you just hold down the wall switch you can force it to > close while ignoring the malfunctioning door safety sensor. Good to know.] > We've already replaced the door sensor components once, not that long ago. > Now it's doing it again. > I got out the oscilloscope and looked at the signals. There are two, 10 > VDC with a several-volt slow square wave (Fluke says 42 Hz) riding on it on > the emitter side, and a negative-going pulse train on the sensor side. (The > emitter supply is probably just dirty DC, with no intentional signal at > all.) The aiming seems correct, as when I flexed the mount on the sensor > side it dropped out about equally on both directions. The sensor pulse > train is about a 10 VDC resting signal, with a 220 µs drop to zero, and > another 200 µs slow rise. The period is about 6.5 ms, and measures (Fluke) > at 151 Hz. (Another opener measures around 170 Hz.) With two pulses on the > screen (negative trigger) I could see that the second pulse wasn't always > there when the door was open to the daylight. With the door down it was > always there. The system is clearly quite sensitive to missing pulses, it > doesn't have to be missing very long for it to trip. > > So, flange up a timer-based pulse train to eliminate the wretched 'safety' > switch? The prior opener didn't have one, just the back-pressure safety > switch, and I never felt the least bit unsafe. We don't have crushable > babies crawling around the garage door while we're driving in and out, and > I suspect that the car itself is a greater threat to whatever than the door > is. > > I think a 555 timer could do the job. IIRC that's a 555, two resistors and > a capacitor for the timing components, and two diodes to ensure a steady > power supply from the door opener. Maybe one more resistor to ensure that > the 555 output doesn't draw too much current. Do I have any 555's left in > the junkbox? [Maybe, but I didn't want to conduct a heavy search. I ordered > a pack of 12 new ones through Amazon, I went for TI rather than no-name > Chinese. Probably cost 2–3× as much, but they're still cheap.] > > Thursday, August 19, 2021 <> > The 555 timers I ordered came, I flanged up an old-school point-to-point > knot of a circuit, using the classic astable form from the datasheet. I > used a 150 kΩ timing resistor (RA), a 6.8 kΩ discharge resistor (RB), and a > 0.1 µF timing capacitor. I left no 555 pins unconnected, there's a > decoupling capacitor on VREF, and RESET is tied inactive. > I had mis-read the wiring on the opener earlier, getting the pushbutton > wiring mixed with the sensor wiring. The emitter and sensor hook to the > same two wires, marked Common and OBS on the side of the opener. Because > power and signal share a pin, I used an inline diode and a 100 µF filter > capacitor to isolate the output pin of the 555 from the power. (When the > [short] low-going pulse of the 555 grounds the OBS pin on the opener, the > filter capacitor keeps the 555 timer powered, and the diode keeps the > filter capacitor from discharging back through the output.) > > So, to summarize this is a 100% standard astable oscillator, right out of > the data sheet, but with a diode/capacitor filter-isolater in the power > feed line, and the output tied to the power feed. The component tolerances > were not great, but I was using junkbox parts. The signal is slower than > the door sensor's, at around a 10 ms cycle rather than the target 6.5 ms, > but the opener doesn't seem to care. I didn't have to tweak the values > after the first build, I just installed it as-is. We'll see if it has > problems later, but initial results are positive. > > Basically a 2-wire circuit blob that hooks to Common and OBS, in place of > the wiring to the door sensor modules. > > > -- Jim > > _______________________________________ > http://www.okiebenz.com > > To search list archives http://www.okiebenz.com/archive/ > > To Unsubscribe or change delivery options go to: > http://mail.okiebenz.com/mailman/listinfo/mercedes_okiebenz.com > > _______________________________________ http://www.okiebenz.com To search list archives http://www.okiebenz.com/archive/ To Unsubscribe or change delivery options go to: http://mail.okiebenz.com/mailman/listinfo/mercedes_okiebenz.com