Modifying an LED bulb with built-in PIR sensor to stay on longer

Thread Starter

seanspotatobusiness

Joined Sep 17, 2016
210
I've got this cheap LED light bulb with a built-in PIR sensor. It works great except that it's only on for about 20 seconds before it turns off. I would like to extend this to about three minutes. Unfortunately it doesn't have a potentiometer but I was hoping I could change a capacitor or resistor to lengthen the time. I have a couple of pictures of the PIR circuit. The circuit which converts the mains AC to DC is a separate PCB which is deeper in the darkest recesses of the bulb.

The chip is a BISS0001 and according to the datasheet there is a resistor and a capacitor connected to pin 3 which I can use to adjust the on-time. I used my multimeter to find the resistor (R2!) and its value is 100 K. Will it be likely to work if I change it to 1 M? The datasheet doesn't specify any kind of limits but I assume there will be an upper limit where the current is reduced too much for the chip to function normally? I have a selection of SMD resistors so my preference is to change that and not the capacitor.
 

Attachments

jpanhalt

Joined Jan 18, 2008
11,087
I agree with your concern that there may be an upper limit to the value of that resistor. If we are talking about the same resistor (RR1 on pin 3), the typical circuit shows only 10K. Have you identified RC1 (pin 4)? The datasheet shows that as 10 nF. You might be able to increase that to 100 nF more easily or change both to get the same product.

Unless someone has experience with that chip, I suspect whatever you do will require experimentation.

EDIT: My labels for the components are based on the pinout figure. For the schematic shown later, the labels are R10 and C6, respectively.
 
Last edited:

Thread Starter

seanspotatobusiness

Joined Sep 17, 2016
210
The capacitor is C1, adjacent to the relevant resistor R2 (I'm referring to the markings on the board rather than the datasheet which marks them differently). It's visible in the second image. The in-circuit capacitance measured as 2.7 nF, but I know in-circuit measurements can't necessarily be trusted. I guess it's just a ceramic capacitor?
 

ericgibbs

Joined Jan 29, 2010
18,734
hi,
Its not easy to follow the PCB track on that 2nd image.
Look at this clip and trace from pins #3 and #4 to the Cap.
E
bs1.PNG
 

Thread Starter

seanspotatobusiness

Joined Sep 17, 2016
210
Thanks. I found the resistor and capacitor using my multimeter in continuinity mode with one probe on pin 3/4. They are R2 and C1 on the second image in my OP.
 

ericgibbs

Joined Jan 29, 2010
18,734
hi,
You could leave the existing cap in place and piggy back a larger value cap across, avoids damaging the PCB track trying to lift off the old cap.
E
 

jpanhalt

Joined Jan 18, 2008
11,087
Yes, that capacitor on the PCB looks like a ceramic. Based on the design in the datasheet, I would increase the capacitor first, test, then consider the resistor. That suggestion is based on the ease of increasing the capacitor, as Eric suggests.
 

Thread Starter

seanspotatobusiness

Joined Sep 17, 2016
210
I just got some more information which might mean I can go ahead with a resistor swap. I already had a couple of HC-SR501 PIR modules in my collection and I just measured the maximum resistance of the potentiometers - it's 1 M. They use the same BISS0001 PIR chip. Do you think that this is good evidence that replacing the 100 K resistor with a 1 M one is likely to work after all? It seems like good evidence to me but I might be missing something.
 

jpanhalt

Joined Jan 18, 2008
11,087
Also, too much resistance in that situation won't damage anything. Just be careful removing the chip. The solder is probably lead free and melts at a higher temperature. You do not want to screw up the PCB.
 

ericgibbs

Joined Jan 29, 2010
18,734
hi sean,
As the unit is a cheap item, it may mean that the drive transistor will be OK for say a 20Sec On time, but the transistor may run hot at 180Sec, so check the temperature of the driver.
E
 

DickCappels

Joined Aug 21, 2008
10,140
But only measure things in the circuit when it is completely disconnected from AC power. You probably already knew that, but I just wanted to make sure that it was mentioned.
 

Thread Starter

seanspotatobusiness

Joined Sep 17, 2016
210
I'm not sure which component will be the relevant transistor but I see two candidates, U1 and Q1 and neither exceed 45 °C after a few minutes on. The aluminium PCB to which the LEDs are soldered exceeds 80°C in this time, however. These measurements are taken by taping thermocouples to the surface so the junction temperatures must be higher.
 

ericgibbs

Joined Jan 29, 2010
18,734
hi sean,
At a body Tempr of 45C for the transistors, it will not be a problem, the LED cooling plates always run to hot too touch.
E
 
Top