# How do I interpret an RGB LED max amperage given as a range e.g. 30mA-70mA?

I have some LED's I got in a mixed package and the RGB LEDs instead of a single max amperage just list a range. One example is I have a common cathode RGB LED with ratings listed as 1.9-3.4v and 30mA-75mA. I've been interpreting the V range as the lowest forward V of the red to the max Vf of the green and blue elements.

Should I assume that like with the Vf the red element likely has the lower 30mA rating while green and blue have the 75mA rating, or do I assume that 30mA is the lower end of the functional range of the all 3 elements and 75mA is the max? Nothing burns out when running I run it briefly at 60mA to all colors and no issues.

Mostly I'm wondering what the standard practice is for interpreting these values when they are listed this way.

Edit/New info Based on a great link from anrieff I've tried playing around some with the bulb in question. At below a total draw of 30mA the color balance of the bulb starts to shift from a cool white to the green. This brings up another issue, since the draw at 30mA would be 10mA per color element, does that mean the upper 70mA limit should be treated as a total instead of an upper limit per color element. i.e. ~23mA per color element for a total of 70mA instead of the 70mA per color I was originally thinking?

• See this, it may be the same thing. – anrieff Feb 28 at 18:04
• Please share the datasheet if available, might improve the explanation. – Huisman Feb 28 at 18:11
• Huisman: One of the issues here is I have not datasheet or part number etc. I've tried to find a matching bulb with a spec sheet that has info but so far no luck. These particular bulbs came in a verity pack from my local electronics store. – CyF Feb 28 at 18:21
• anrieff, thanks for the link, I'll try playing around with the mA along those lines. – CyF Feb 28 at 18:24
• anfief, there is definitely a point where the color balance starts to shift. I can see it noticeably at around a 30mA total draw (~10mA per color) which brings up another question I'm editing in above. – CyF Feb 28 at 18:35

Without seeing the datasheet, it's not possible to confidently give a single reason, but here are a few possibilities:

• Each color within the RGB package almost certainly have different $$\V_f\$$ and $$\I\$$ ratings. The rating you are looking at may just be trying to encompass these values, as you have already assumed.
• It's possible that the current ratings are based on duty cycle. Many LED datasheets will provide a constant-current value (e.g. 20mA) as well as various ratings based on lower duty cycles (e.g. 50mA at 50%).

You'll need to find the datasheet or ask the manufacturer to be sure.

• Thanks for the answer. I've been sticking to getting bulbs from other locations for this very issue, hard to find a sheet when you don't know the manufacturer or any other info. I'll see if I can get in contact with the company that put the pack together in This case. – CyF Feb 28 at 18:44
• @CyF: Bulb, noun, a) the glass housing, in which a partial vacuum has been established, that contains the filament of an incandescent electric lamp. b) an incandescent or fluorescent electric lamp. An LED is not a bulb. You will cause some confusion when you use the wrong terms. – Transistor Feb 28 at 18:58

Note the current derating with rising ambient ( such as local self-heating) depends on the package and board heatsink.

I be correct to interpret the below ΔT/ΔW as (100-25'C)/20mA*3.2V= 1.17'C/mW junction temp rise above per LED for 3 LEDs to be 390'C/W rise.

The RGB LEDs that give an "Absolute Maximum" current spec usually have fine print to indicate the conditions like 10% duty cycle 2KHz or 0.1 ms. These devices may be limited by the wirebond whisker wire size that could be smaller for Red ( as they have lower Rs in the LED and lower Vf) thus the lower current for Red , I think is fusing considerations of the gold wirebond.

But overall it is the thermal resistance and power dissipation that you must pay attentionto including the need for a copper "common node" ( * ) heat radiator on the PCB. ~1 sqin/W if SMD.
(*) either Common Anode or Common Cathode

e.g.

never operate near the ABS MAX. You can measure junction temps using pulse Vf with 1~100% If due to NTC temp coefficient with pulse low If cycles, if you care about junction temps.

• Thanks again for the detailed info. I've usually not been going above 80%-90% of the max listed values. I've then been using the max though in my wattage calculations to make sure I have some extra breathing room. – CyF Feb 28 at 19:16
• You ought to use rated current not MaX unless using 2kHZ PWM – Sunnyskyguy EE75 Feb 28 at 19:21