Some of the circuit does not look right, and if there is only 2.2 volts on the regulated 5 volt line then certainly that is part of the problem. Is C1 leaky?? Or is the IC drawing far too much current?

 

On closer look, the IC has 2.7V on pin 7 (DI) yet only 2.2V on pin 1 (VCC) which means the 5V reg is doing nothing. Backfeed from the IR LED monitor R7/D6 is actually giving the little bit of power. This leads me to believe R4 is open, but OP did not give many part values, or maybe D9 is leaky and stuck at 1.5V instead of 5.6V which seems a bit weird.
I've done reverse-engineering on other obstruction sensors like Liftmaster, Chamberlain, Sears.
I could draw a better schematic if OP is willing to give part values and another pic of the board top/bottom.

 

On closer look, the IC has 2.7V on pin 7 (DI) yet only 2.2V on pin 1 (VCC) which means the 5V reg is doing nothing. Backfeed from the IR LED monitor R7/D6 is actually giving the little bit of power. This leads me to believe R4 is open, but OP did not give many part values, or maybe D9 is leaky and stuck at 1.5V instead of 5.6V which seems a bit weird.
I've done reverse-engineering on other obstruction sensors like Liftmaster, Chamberlain, Sears.
I could draw a better schematic if OP is willing to give part values and another pic of the board top/bottom.

It may be the pass transistor is open, and still, that 12 volts DC supply will not be adequate for evaluation of the rest of the circuit. Also, given that the signal connection is somehow through the power wires, using a regulated power supply is sure to prevent correct operation. The scheme that this arrangement uses to communicate back to the opener circuit is not completely clear to me.

 

I checked the D9. It is a Zener diode 1N5232B. Prairiemystic suggests I can remove Q1 and test D6 for 5.6V. I am wondering if D6 will only connect to D10 and pin 7 of IC, how can I have 5.6V from D6? Is it supposed to have 5.6V from D9 after I remove the Q1?
Or can I remove D9 and connect it with a 1K Ohm resistor on its cathode, and then apply a 12VDC across with the positive 12V on the resistor? I shall have 5.6V reading from the D9 cathode in this test if it is good.
I don't feel Q1 is hot when I apply 12VDC on inputs.

 

R4-3.9K, R9-30K, R1-10K, R2-10K, R5-330 Ohm, R7-2.2k, R8-30k, R6-2,2k, D10 - 1N4148

 

I removed one leg of D9 in the photo.

 

Thanks for the pictures. I drew this schematic based on the board pics. I can't make out all the resistor colors between orange and red in the .jpg's and the camera effects.
Test zener D9 (one leg lifted) with a multimeter on diode-test and then ohms, to see if it is leaky or shorted. It is suspect.
D9 is supplied power through R4, so that's why I was saying pulling Q1 should give voltage (5.6V) across D9 if Q1 is strangely doing nothing, despite having an intact E-B junction. Never power the board with D9 one leg lifted.
What is C2? These age and I would replace it with a low ESR part.

 

[...] The scheme that this arrangement uses to communicate back to the opener circuit is not completely clear to me.

Garage door obstruction sensors are basically a one-wire bus with the three - IR transmitter, IR receiver, and door opener all tied together, in parallel.
It's a bit weird because there is a requirement for safety- fault detection and the system cannot be bypassed or thwarted as well.

Power comes from the door opener which has a pullup resistor to DC power 12-24V, around 220-470R 2-3W.
The door opener's MCU monitors the bus voltage. If (long term) it's high then there is an open circuit no load no sensors connected, if it's low then a short circuit is detected.
The IR transmitter is typically emitting bursts of 38kHz pulses for 0.3msec at a slow frame rate 160Hz (Chamberlain).
The IR receiver is an IR remote control receiver with slow AGC for light curtains/fences. So the demodulated IR is low frequency pulses, PW depends on if an obstruction exists or not. So the garage door opener is always looking for pulses on the bus, else there be a system fault. The receiver actually briefly switches on a transistor to take the bus low (hard) for the pulse duration, but the modules have the capacitor to keep running for that short time.
So the receiver is pulsing the current-loop (modulating the voltage) to send the pulses back to the door opener.
The opener expects to see pulses all the time, or there is a system fault.

 

Since you mention Dafe-T-Beam I assume it's for a Genie garage door opener? You have red and Green LEDs which should be doing a diagnostic. Genie Safety Beams Explained. 12 Volts DC is the normal operating voltage, actually 12 to about 20 volts should work fine. So what are the red and green LEDs doing? Red & Green On is normal. Red & Green Off It's all covered.



So what do you have?

Ron

 

Probably "none of the above" using a 12 VDC power source. At least that is what I see. There is normally communication that goes on through that AC connection to the main unit.

 

The schematic from Prairiemystic is great!
I lifted one leg of D9 and tested it. On the diode test, forward bias is 0.77V, and reverse bias is 1.1V. On the Ohm test, forward bias is 1.4K, and reverse bias is 1.1K.
The C2 is 220uF.

 

Zener D9 should be pretty much open-circuit in the reverse direction (until 5.6V), so I say it is bad according to a multimeter. Look for a replacement as 1N5232B. Is C2 25V or 35V? I will update that schematic.
For no hassles I would probably replace Q1, C2 for the extra $2.

Any chance the problems happened after a lightning storm? These door sensors somehow get zapped by that, it appears to be because they are mounted on earth-grounded rails yet the opener is grounded to the main panel, and a mains strike damages them.

 

The C2 is 16V. It is a good idea to replace Q1 and C2.
Not sure we had a storm lighting. Maybe it is due to water getting into the box and on the board when I washed the garage ground with a water hose.

 

C2 is only 16V? OK- stop powering it at 24VDC or it will pop! I would see if a 25V will fit. The PC board has two footprint sizes.

 

I can buy small quantities of 1N5232B and 220uF Al electrolytic cap at Digikey, but KSP05 has to be bought in bulk for 9,000. Mouser is out of stock. I check OnSemi. this transistor has obsoleted.

 

KSP05 (NPN, 60V, hFE=50) replacements with same E-B-C pinout, I would consider KSP06, MPSA05, MPSA06. The part is not that critical. Digi-Key has all those. The 220uF cap I would get 25V long life Nichicon, Rubycon, Chemi-con etc.

 

Yes. I find KSP06 cut tape at Digi-Key. Thanks!

 

I replaced the 1N5232 diode on the board. The safety system works well now. Thank you for the help!