[Electrical Modifications] Make my touch lamp LESS sensitive?

[iamdarkyoshi]

This has been posted on the EEVBlog forums as well, thought I might as well put it here too.

 

I have reverse engineered my touch lamp's controller PCB, BigClive style.

 

 

 

I have done my best to draw a DaveCAD drawing of the schematic based on the images I had. I used BigClive's trick of taking an image of both sides, and then flipping the image of the backside, layering it on the image of the top, and turning the transparency to about 50%. I love this technique.

 

The problem with this touch lamp is that it does not like LEDs, or lower wattage bulbs. With LEDs, there is enough leakage to actually dimly light the LED lamp no matter what, and the LED lamp makes the touch lamp much more sensitive, and usually will shut itself off with the slightest bit of electrical noise, and will also tend to turn itself on with electrical noise. If I turn my fan to a different speed, it will always cause the lamp to register a touch. What can I do to make it less sensitive to electrical noise and work better with lower wattage bulbs? The COB is very likely the controller that gives three brightness levels on the bulb and does everything else in terms of sensing. Oh, I should add that I live in the US, so the supply is 120V 60hz.

 

Thanks everyone!

[thekeemo]

For the sensitivity if you put a resistor on the end wouldn't that fix it? Or a ferrite bead.

[iamdarkyoshi]

For the sensitivity if you put a resistor on the end wouldn't that fix it? Or a ferrite bead.

The problem is that even without touching it, if I turn on my fan in my room with an LED light in the lamp, the spike from the fan is enough to cause the lamp to turn on/off. I do have a ferrite choke on the cord already. I might add a single stage noise filter with a commonmode choke and capacitor, like what ISNT found in cheap ass PSUs

[thekeemo]

The problem is that even without touching it, if I turn on my fan in my room with an LED light in the lamp, the spike from the fan is enough to cause the lamp to turn on/off. I do have a ferrite choke on the cord already. I might add a single stage noise filter with a commonmode choke and capacitor, like what ISNT found in cheap ass PSUs

just take fitering out of a PSU you have

[iamdarkyoshi]

just take fitering out of a PSU you have

its not like I dont have any

[thekeemo]

its not like I dont have any

you say that turning on your fan causes interference. Maybe filter the output?

[iamdarkyoshi]

you say that turning on your fan causes interference. Maybe filter the output?

turning on like anything causes it to derp. Monitor, fan, PC, other lights, but it is fine wih a 60W bulb in it.

[Kalle Hänninen]

LOL, you certainly didn't give an easy one. Easiest solution is to use power resistor in parallel with the LED.

[iamdarkyoshi]

LOL, you certainly didn't give an easy one. Easiest solution is to use power resistor in parallel with the LED.

which honestly negates the purpose of the LED

[Kalle Hänninen]

which honestly negates the purpose of the LED

Yeah, well.. I don't have a crystal ball. That circuit is its own ecosystem, so tinkering with one part affects every part, and we have no idea what chip it even uses.

[Kalle Hänninen]

What's the breakout voltage of that zener?

[iamdarkyoshi]

What's the breakout voltage of that zener?

Looks to me like it might be ~5v. MOST likely a crude 5V power supply for the chip?

[Kalle Hänninen]

Looks to me like it might be ~5v. MOST likely a crude 5V power supply for the chip?

 

I need to go now and it might take a while till I get back. But knowing form what wattage you started and to what makes it easier to say. Some parts are just more positive than others.

[Kalle Hänninen]

Just had a flash. That might be 555. Google it with capacitive sensor and you might find circuit, or simulate it.

[Kalle Hänninen]

Any luck with EEV?

 

I would start with that 1K resistor. It shouldn't mess with the rest of the circuit, as it's only connected to the IC (555, op-amp, whatever), so adjusting it with potentiometer might do the trick.

 

Got bit confused with that triac earlier. It's been almost ten years from my active days, so I'm not the best, but maybe the only one available. To be honest, that circuit doesn't really justify the effort, and is kinda dangerous when you didn't even understood what zener is, but I'm not my brother's keeper and you die trying, as the saying goes...

[Kalle Hänninen]

Actually... I take it back. Current has nothing to do with it. So even when the skin contact does change voltage reading different across those two pins, it doesn't make any difference. So it's that 470K.

[Kalle Hänninen]

lel.

 

Answer was really simple in the end. Diodes threshold voltage is lower if current is sufficiently low. Also you can power IC using IO's

 

Too little too late, but now it's done.