Photoelectric Switch (Triac) Doesn't Work With LED Lamps

[hollidayp1]

I have several Photoelectric Switches that work well with incandescent lamps (up to 300 watts) but they don't work with LED Lamps. Actually, one of them works with the LED lamps but 10 others do not. The LED lamps are multi-LED lamps with a standard screw base (USA) and range from 1 watt with 19 LEDs, 2 watt with 37 LEDs, and 3 watt with 43 LEDs. A schematic of the Switch is attached.

I'm not 100% sure how this circuit works but here's what I think. It looks like the triac is turned on by the 120k resistor in series with the diac (or diode?) which would basically turn on the lamp (or load) always (like in the dark or light). But then, the .047uf cap in series with the LDR (CDS Cell) would short out the T1 terminal to the gate causing it to turn off in the light when the LDR goes to a low resistance (say 3k ohms). In the dark (low light) the LDR goes to a high resistance (say 300k) and the "short" would be removed allowing the gate to be turned on by the 120k resistor and diac.

I'm not too confident in this analysis because I am not comfortable with the idea of "shorting out" the triac to allow it to turn on. Is this a common trick?

In any case, my objective is to modify these switches so that they work with LED lamps and other non-incandescent (non-resistive) loads. Before I can do that (if possible) I need to be sure about how this circuit works. Thanks for your help. 131122-2

 

[(*steve*)]

See that "Go Advanced" button next to "Post Quick Reply"?

Click on that and you get more tools. The one that looks like a paperclip is the one you want

Beware that there are size limits on the image you can upload.

 

[hollidayp1]

Thanks steve

OK, I did that and it's there as a thumbnail. I wanted to put it in with the post so I could describe it because I saw a lot of other posts with a schematic included full size.

 

[(*steve*)]

OK, I did that and it's there as a thumbnail. I wanted to put it in with the post so I could describe it because I saw a lot of other posts with a schematic included full size.

There's a bit of magic to do that

Press the Attachments tool again (the paperclip), right click on the link it gives you to your image then use the "insert image" tool (it looks like a picture frame) to paste the link to the image.

The other thing is that I (and anyone else subscribed to the thread) won't see that you've done anything in the thread unless you post another message. As such, it is poor form to edit old posts after they've been replied to simply because people won't be aware you've done so.

I suspect the problem is caused because the circuit you have assumes that the globe has a low resistance across its terminals. That's true of an incandescent globe, but not a LED driver.

 

[hollidayp1]

Editing post - will fix it.

OK, I will post again instead of editing the original post. Unfortunately, I edited the original post before I saw your reply. Thanks, and sorry for the inconvenience. I didn't expect anyone to see my post so soon as on most forums you don't see any responses for many days (or a long time) Thanks again.

 

[hollidayp1]

Description of Circuit

Photoelectric Switch (Triac) Doesn't Work With LED Lamps

I have several Photoelectric Switches that work well with incandescent lamps (up to 300 watts) but they don't work with LED Lamps. Actually, one of them works with the LED lamps but 10 others do not. The LED lamps are multi-LED lamps with a standard screw base (USA) and range from 1 watt with 19 LEDs, 2 watt with 37 LEDs, and 3 watt with 43 LEDs. A schematic of the Switch is attached.

I'm not 100% sure how this circuit works but here's what I think. It looks like the triac is turned on by the 120k resistor in series with the diac (or diode?) which would basically turn on the lamp (or load) always (like in the dark or light). But then, the .047uf cap in series with the LDR (CDS Cell) would short out the T1 terminal to the gate causing it to turn off in the light when the LDR goes to a low resistance (say 3k ohms). In the dark (low light) the LDR goes to a high resistance (say 300k) and the "short" would be removed allowing the gate to be turned on by the 120k resistor and diac.

I'm not too confident in this analysis because I am not comfortable with the idea of "shorting out" the triac to allow it to turn on. Is this a common trick?

In any case, my objective is to modify these switches so that they work with LED lamps and other non-incandescent (non-resistive) loads. Before I can do that (if possible) I need to be sure about how this circuit works. Thanks for your help. 131122-2

 

[(*steve*)]

I *think* the solution may be to move the end of the 0.047uF capacitor from one output connection to the other.

However, this places stuff across the mains which means a failure would be rather more spectacular.

See if someone else has a better solution.

 

[CDRIVE]

Am I missing something here? PWM doesn't work with LEDs for the reasons we're all aware of. Why then would Triac phase slicing be any more effective?

Chris

EDIT: OOOPS! I see it's not a dimmer but rather an LDR switch. Regarding my questions: Never mind..

 

[CDRIVE]

I *think* the solution may be to move the end of the 0.047uF capacitor from one output connection to the other.

However, this places stuff across the mains which means a failure would be rather more spectacular.

See if someone else has a better solution.

Steve, I think I do. I wasn't sure what you were describing but I got it now. The LDR / Cap circuit shunts the gate current when the LDR's resistance goes low, IE, sees light. Following your suggestion it would seem logical that if two or more (more would be preferable) LED lamp clusters were powered by his (as is) circuit it may cure his problem without modifying anything.

Chris

 

[hollidayp1]

Reply to CDDRIVE

..... Following your suggestion it would seem logical that if two or more (more would be preferable) LED lamp clusters were powered by his (as is) circuit it may cure his problem without modifying anything.

Chris

Thanks Chris. I have tried putting 3 or four of the lamp clusters on the output to no avail and with varying results. To clarify again about the LED Lamps - The 1 watt LED lamps have 19 LEDs with some internal circuitry to allow them to operate at 120 vac. The same is true for the 2 watt and 3 watt lamps.

I have tried 10 of the switches with various combinations of the LED lamps with varying results. One of the switches works perfectly with one 1 watt LED or two 1 watt LEDs (I haven't tried 3 or more yet). However the same switch does not work with the 2 watt or 3 watt LEDs. Some of the 2 watt or 3 watt LEDs don't turn on at all and some of them light up about half power and the LDR light or dark has no effect at all.

I have not opened up the one working switch because I don't want to damage the one unit that works with these LED lamps. (The switches are in a case that must be cut open to get to the circuitry).

Since the LED lamps that work with the one particular switch do not work with any of the other switches, I have concluded that the switch is the culprit and it should be able to be modified to work consistently with the LED lamps.

A further conclusion is that it is the unique characteristics of the particular triac in the working switch that allows it to work with these LED lamp clusters.

 

[(*steve*)]

Here is my recommended modification:

I'm not totally happy about the capacitor used in this circuit, but if it works now, it should work during a test.

 

[CDRIVE]

OK, if you look at the circuit you posted you will note that the LDR along with the .47uF cap shunts current around the Triac gate to neutral through the lamp load. The manufacturer shunts current is through the lamp load to insure that the LDR module will not draw current when no lamp is screwed into the lamp socket. Otherwise they would have routed the LDR and .47uF cap directly to the neutral. The circuit is designed to reduce this shunt current when the LDR sees light and its resistance increases. When this happens the Gate of the Triac will see sufficient voltage and current to fire it because less current is being shunted around the gate.

I believe you have a few of alternatives here. ..

(1) Lower the 120K to ~ 100K (not my favorite solution)
or
(2) Decrease the .47uF to ~.2uF
or
(3) Place a 100K in series with the LDR. (I like this one best)

These values are ballpark figures and total guesswork based on your current values. I'd be more than receptive for other opinions. I always get a bit queasy when advising modifications of any mains powered appliance. Please read my disclaimer.

A bold prediction: One year from now there will be no issues finding LDR switches to control LED lighting.
A bolder prediction: maybe in 6 months?

Chris

 

[(*steve*)]

The manufacturer shunts current is through the lamp load to insure that the LDR module will not draw current when no lamp is screwed into the lamp socket.

It also limits the fault current to that of the lamp.

 

[CDRIVE]

It also limits the fault current to that of the lamp.

Yes Steve, I'm assuming that you mean if the cap fails (shorts or leaks) the current is routed harmlessly through the load.

Chris