I have an RGB LED strip with 58 LED's on it. Resistors are found on every third pair, except for the last one which has its own set on it. Putting a 12V 5A power supply on it, I measured the following:

Input voltage: 12.24V
Red Current - 290 mA
Green Current - 380 mA
Blue Current - 350 mA

The multimeter I used couldn't get more precise than 0.xx amps, so those numbers might be slightly off. Keeping that in mind, I have three questions:

1) How do I measure voltage drop of each color of the LED strip? If I measured in series with the whole strip, the LED's wouldn't light at all, but I would still get what appeared to be a useful reading (on the order of I think 10.50 volts left on the red channel). If I measured in parallel, they would light, but my reading would be that 12.24V number.

2) As long as the voltage supply meets or exceeds the voltage drop per color, is it safe (and equally bright) to drive it at lower than 12V?

3) Would a constant-current driver work when a series resistor is already utilized?

Thanks in advance for any input!

 

I think you need to take some clear photos with different angles, I guess the led circuit as below, you may check the connections if you can see them clearly, and draw how their difference.

 

Resistors are found on every third pair

What constitutes a pair? 2 red? 1 red + 1 green? .....?

If I measured in series with the whole strip, the LED's wouldn't light at all

If you have the meter set to read Volts it shouldn't stop the lights working .
Can you post a schematic showing the parallel/series arrangement of the LEDs, and the measurement points?

 

If you have the meter set to read Volts it shouldn't stop the lights working .

I didn't think so either, that's what confused the living daylights out of me. There appeared to be a voltage drop, but nothing lit, which baffled me. I would've thought it's just be dimmer if the input voltage was too low, not that nothing would light. It didn't seem to cause any damage by doing that though either.

Okay, so here's a quick schematic I drew up of the board itself. I have to make a correction about the solo LED, they actually appear to have two sets of 2 LED's each with higher resistor values, not just one solo LED (which actually makes more sense IMO).


I was able to use the continuity mode of my multimeter to verify that the higher resistor value was indeed for the red bulb (although it didn't provide a direct reading, it did illuminate the one LED slightly when I touched the probe to the one "391" resistor. Wasn't able to replicate this for the other two colors, but since they're the same resistor value for both, I'm assuming blue and green share the same voltage drop.

As for voltage testing, I put six labels on the left side of the schematic that indicate where I probed the connections for voltage readings. Should be fairly self-explanatory (ones ending in "-" are where the black/COM lead was connected to, ones ending in "+" was where the red/V lead was connected). One thing to note: I only grounded one color at a time (there's only four pins on the connector, so either it's a common-anode harness or circuitboard).

Pictures of the unit itself:


Mixing the last two questions in the first post -- if a series resistor is already used, can constant-current drivers be used at a lower voltage? Or would I need to supply an equal or higher voltage altogether? I've thought about maybe bumping up lower voltages to 12V, and if the supply voltage exceeds that, to then limit it to an exact 12V. Could this potentially work? Or is there a better approach I can take here?

 

I can't relate the pics to your schematic . You show a 2-wire connection for the whole array whereas the fifth pic shows 4 wires (common +R+G+B?). Are the three colours selectively switched/controlled?

 

Oops, didn't catch that. My bad, schematic completely ignored the separate grounds. Just updated it to include everything from the individual grounds to the cable (see below). Does this help to clear things up?


EDIT: Shoot, forgot the test points in the new schematic. Everything is tested on the "END1" side of things, since that's where the wire terminates.

 

That's clearer.

How do I measure voltage drop of each color of the LED strip?

With the meter set to 'Volts', put the red probe on End1-pin 4 (i.e. the +12V rail) and put the black probe where the cathode of the bottom-most LED connects to the series resistor.

 

That's clearer.

With the meter set to 'Volts', put the red probe on End1-pin 4 (i.e. the +12V rail) and put the black probe where the cathode of the bottom-most LED connects to the series resistor.

So that's what I was doing wrong. Thank you very much for this! Seems to have cleared up a ton by being able to measure it this way. Got some very interesting results actually, measuring with one of the "three LED in series" pairs:

Before resistor:
Red - 10.10 V
Green - 6.24 V
Blue - 6.04 V

After resistor:
Red - 4.24 V
Green - 3.11 V
Blue - 3.34 V

Calculating for a power supply of 12.24V, we get roughly:
Red - 2.14 V / 3 led's = 0.7133... Vf
Green - 6 V / 3 led's = 2 Vf
Blue - 6.2 V / 3 led's = 2.066... Vf

Here's the interesting thing: if I'm interpreting the after resistor measurements correctly (and I may not be), then it appears that as long as I have (12.24 V - 3.11 V = 9.13 V) or more, I should still be able to run these units at full brightness. Maybe the seller was wrong about it being a strict 12V after all. That would explain why no one has seemed to report any "bad" units in the field yet.

Am I wrong in thinking that I can actually run these at 9.13V with identical results? Or does this mean that using a constant-current driver may be doable after all?

 

Running the LEDs at reduced voltage will result in reduced current and hence reduced brightness.

 

Running the LEDs at reduced voltage will result in reduced current and hence reduced brightness.

In that case, it sounds like I'll need something like this:


"END1" being the same part as in the previous schematic (which connects via ribbon cable to "END2" and then on to the LED circuit). The above is seriously oversimplified, but it does bring up two questions:

1) How do I go about selecting a constant-current driver? Will need PWM and "disable" functions accessible by an external Arduino (how doesn't matter).

2) Once that's done, how do I bump up voltages below 12V to at or above 12V, but remove transient voltage spikes also?