PIR Motion Detector Night Light MODIFICATION

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Wife bought a couple night lights that have PIR detectors in them. There are two settings; setting 1 turns the light on at dusk via an LDR. In mode 1 the lights are low. When the unit detects motion it changes to a brighter setting. In standby it puts out about 15 lumens and when motion is detected it puts out 40 lumens for 90 seconds after the last MDE (motion detect event). In setting 2 the night light is off until it detects motion after dusk.

What I want: I want to reduce the 40 lumens by at least half and I also want the 90 second after last MDE to be shortened to about 10 seconds. Reason is this unit is going to be in the hallway. Nobody hangs out in the hallway and at night when someone gets up for a drink of water or to use the facility the other person still sleeping in bed has to wait the full 90 seconds before the light goes off - AND it's bright enough to disturb my sleep. So in the picture [edit] Drawing removed - see post #8 for more complete information. [end edit] you see the board with all components possibly identified. U2 has some traces underneath the chip and I can't determine where they go without desoldering it. And I'm not currently equipped to desolder a chip like that without damaging it. To the right of the board is the light board illustrated in traditional schematic diagram form. I could use a little help picking out the time delay (90 sec.) and reducing the current (looks like R10 just needs to be bigger. (and U1 must have a trace underneath that goes to the pin indicated with the dotted line (sort of)).

[edit] Drawing removed - see post #8 for more complete information. [end edit]

The LED boards:
1621298765672.png
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
I'm putting a 1KΩ resistor between the two LED boards (the black jumper wire). Hadn't thought of that originally. I was looking at the 1K SMT resistor on the main board. Expecting good results. However, I still don't know where to start with changing the timing of the board.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Does anybody know what U1 is? I found an audio device in one search but I don't think U1 is a speaker. Also, after having experimented with several resistors in series with the LED's the circuit is now NON-functional. Nothing is happening and I have no idea why.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Thanks for the help ElectricSpidey.

The problem might be with Q2 or with U2. U2 should be pulling Q2 base low to turn it on. LED's are powered with what appears to be about 25VDC. When first turning power on I'm reading 28VDC but it quickly falls into the 25V range and seems to settle down after a minute at 25.4V.

I'm no fan of capacitive voltage drops but the rectifier (DB1) appears to be working properly. No, I haven't scoped it.

Initial attempts at reducing brightness - I used a 1KΩ between the two LED boards. Light diminished very little. Tried 2K, 3.3K, 4.7K then went with 10K. At 10KΩ I achieved the level of brightness desired, down from the original. But then it stopped working when I tried to put tape over the LDR. Testing a night light during the daytime is problematic so I wanted to block all light from reaching it. That's when the problem manifest itself.

Now, at power-up I get the briefest and extremely dim light from the LED's. Haven't a clue what's going on. Searched for a data sheet for U2 but haven't found anything based on the chip markings (as seen in yellow).

Z1 (zener diode) I have no idea what its breakdown voltage is. My guess is that it's regulating the DC voltage at around 25V. Whatever U1 is doing - I'm totally in the dark (no pun intended) (working on a night light here). U1 is either a lamp (based on google search) or an LDO chip. Being that this is powered by mains - not certain why you'd put an LDO in the circuit. Unless it has something to do with the 90 second timing and capacitor (E1, not marked on the solder side) (100µ@16V, just above C3).

ANY assistance would be appreciated (beyond "you broke it").
 

ElectricSpidey

Joined Dec 2, 2017
1,663
Sorry dude I was just messin with you.

Perhaps the LEDs are on constant current and when you increased the resistance you forced the chip to raise the voltage to a level that it cant handle.

Putting in a 1k resistor should have dropped the light output quite a bit...so that might be a clue that it's constant current.

Just a guess.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Good guess. And I didn't mind being messed with. I mess back from time to time as well.

Been looking through different things and I think I found something that somewhat approximates the chip I have. Unfortunately the image is horribly unclear. Yet I think I can make out that pin 1 is VDD and pin 7 is OUT. It's connected to Q2 (and I have to correct the drawing a little bit) Other pins; 2 = PIR; 5 = GND (or VSS, can't tell for sure). But I just can't find a data sheet for AS092 specifically. I can find AS09208, but that takes me to a speaker of some kind. My chip is a chip, not a speaker. So I'm having difficulty finding the exact data sheet for what I have.

Corrected drawing (as far as I can determine thus far): (corrected traces for Q2)
[edit] Drawing removed - see post #8 for more complete information. [end edit]
Not my circuit, but close enough to give me even a clue.
1621522329427.png
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Spent the day deciphering the schematic for this motion detector night light. Updated the "Paintbrush" file as well. Had to do some deconstruction on the dead board. It's not coming back.
Paintbrush:
1623800955902.png

Schematic (IF I got it right)
1623856094077.png
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Now - back to the issues:

1) too bright when motion is detected.
2) stays on too long (want to shorten it significantly)

Before you ask - no - I haven't done any more testing. For what it's worth, the board is non-functional. Messing with 120VAC isn't my favorite pastime, so I suppose if someone has a good suggestion - I'm all ears.

Blue lettering is marked values on components. Note: E1 and E2 refer to Electrolytic capacitors. Other caps with no value - I haven't a clue. You might. I don't. Just what is 3D? Maybe a transistor, maybe a FET. Had a hard time coming up with any information on U1 and U2. If you find something - please link it in.

Thanks.
 
A low-tech solution for excess light:

I used to have a D-Link Wi-Fi router (actually, I still do, it's just sitting in its box out in the garage) that had blue LEDs bright enough to illuminate the room at night, giving the effect of a TV on just down the hall at night.

It took some experimenting with different types and thicknesses of paper to cut the night brightness yet still give enough light to easily see the unit's status during the day.
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Found a few errors in the schematic. Corrected them, those I've found so far that is - - - .

Putting flammables inside lights is not the best solution. Remember, I want to achieve two goals, dimmer light and shorter duration. The first one I blew up (no smoke or light) appeared to be constant current, but I can't be sure without knowing what AS092 is (U2) (not the band either).

1623857414199.png
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
There are two functions one can chose from:
Dim light until detected motion - then full bright.
OFF light until detected motion - then full bright.

Full bright lasts for approximately 90 seconds. Would like to chop that down to around 10 seconds. (it's in the hallway and nobody is hanging out in the hallway). Besides - if motion continues to be detected then the bright period continues until no more motion is detected, then the bright light times out and goes either back to dim or complete dark.
 
Last edited:

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
So correct me if I'm wrong, but Q1 appears to be a regulator. No idea what U1 (7133) is doing. U2 must be some sort of controller and Q2 executes the commands from U2. Output pin 7 must be some form of PWM. But how do I change or reprogram it? I don't even know what it is; can't find any info on it. S1 simply switches from one mode to the other and there doesn't seem to be any feedback from Q2 to U2. How would U2 know how much current is being drawn - that is IF it's a constant current arrangement? Seems I need information on U2.

As a side note I'd like to devalue the sensitivity to light (LDR). Seems when it's well lit its resistance goes low (about 2KΩ) and when dark its resistance goes high (about 22KΩ) based on observations. How would I change R7 to make it so that when the room is darker than currently set for THAT is when the light comes on. This is in a hallway and it doesn't see much natural light. So it comes on in the afternoon when the sun is high. But during morning or late afternoon hours the hallway is better lit from light coming in through the windows that the light doesn't come on. So I'd like to modify that to set it to come on at a darker time of the evening.
 
Good guess. And I didn't mind being messed with. I mess back from time to time as well.

Been looking through different things and I think I found something that somewhat approximates the chip I have. Unfortunately the image is horribly unclear. Yet I think I can make out that pin 1 is VDD and pin 7 is OUT. It's connected to Q2 (and I have to correct the drawing a little bit) Other pins; 2 = PIR; 5 = GND (or VSS, can't tell for sure). But I just can't find a data sheet for AS092 specifically. I can find AS09208, but that takes me to a speaker of some kind. My chip is a chip, not a speaker. So I'm having difficulty finding the exact data sheet for what I have.

Corrected drawing (as far as I can determine thus far): (corrected traces for Q2)
[edit] Drawing removed - see post #8 for more complete information. [end edit]
Not my circuit, but close enough to give me even a clue.
View attachment 239094
Run this through google translate if you need to;
https://chfile.cn.gcimg.net/gcwthird/day_20170822/798af93eb4k4894a4fda7od6dc8d5092.pdf
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
Look at the edge of the component U1 see if there are pins not connected
No. There are three pins on the component. There are three pads on the board. It could not possibly have any other pins. Sorry sister Mary - there will be no hailing this time.
While I haven't tried to run this through a "Translate" app, I did notice in this diagram the 7133 is listed as AS7133, an LDO device (Low Drop Out)(assumed) rated at 3.3 volts on its output.
1623875308205.png

1623875615608.png
It is configured a little different from what is on my PCB, which is probably why I'm seeing some 10VDC on that line.

Oh, wonderful. MY PS is acting up. Will have to troubleshoot that before I can do any more testing as it's a source of an isolated 120VAC through a Ferroresonant transformer and an auto transformer (variable AC).

Anyway - perhaps it's possible to connect to pin 6 and make some adjustments there. But there's nothing that looks like a Constant Current source. When I get my PS up and operational I'll open a new Motion Detect Night Light (MDNL) and take some measurements.
 
If your schematic is correct then U2 should be running on 3.3 volts. The 7133 is a regulator set for 3.3 volts. As for changing the sensitivity or level of the LDR circuit - you'll need to reduce the resistor (R7 56K ohm). Depending on what level you desire - start with standard values. I don't see why they'd want a high precision resistor (R7). A 47K resistor doesn't need to be quite so precise. If that doesn't give you the level you want then you might want to add a resistor between the V+ and the LDR.

As for the functionality or adjustability of U2 - sorry - I can't help you with that. But it doesn't look like CC to me. It might be PWM but you're going to have to test for that. Do you have a scope?
 

Thread Starter

Tonyr1084

Joined Sep 24, 2015
5,992
I've made some rough measurements. When exposed to light the LDR resistance drops to around 2KΩ and when shaded from light it measured around 22KΩ. Lowering R7 means increasing the current and wattage dissipation. Assuming a lighted condition of 2KΩ plus 56KΩ, 58K total at 3.3V means dissipating about 195mW. If I lower that to, as you suggest, 47KΩ or 33KΩ then at 47K (49K) that's 222mW, a 27mW increase. At 33K it jumps to 330mW. Adding resistance before the LDR seems like a better approach. No? If I add 20K before the LDR then the wattage drops and I think that should shift the point where the sensor switches from day to night operation. Again, correct me if I'm wrong.
 
Top