Do LED limiting resistors (or LED strips) get dimmer/fail?

Hi All

Just thought I'd ask here before climbing up into my attic to check out the wiring etc!

I have recently (3 weeks ago) installed some night time LED lighting in two rooms of the house - top stairs hallway and bathroom. Each 5V LED strip is activated by its own independant PIR sensor, both are powered independently from the Arduino and activated by an optically coupled relay module.

Worked perfectly when initially installed, but at the moment what I am seeing is a definitive and pretty significant drop in light output from one set of LED's as soon as the other set is activated. They are powered by a converted PC ATX PSU, which is able to dish out plenty enough amperage, so I am wondering why this occurs?

I have a 270Ω limiting resistor in each LED circuit, but am curious as to why this seems to be a 'getting progressively worse' issue and am keen to learn why this is so.

No big deal as the LED strips are low cost and I already have a replacement set ready to install, but I am curious. Any advice?

John, you have posted often enough to know that you should provide details when you ask questions. What type of led strip, PSU etc. Model numbers, links to data sheets or at least pages giving some details of the specs.

Are you saying both strips are at full, original brightness if the other strip is off? If so, that points to a problem somewhere else - wiring, PSU etc. Measure the supply voltage at the strips when each is on separately and together. Measure the voltage at the PSU in each condition.

PaulRB:
John, you have posted often enough to know that you should provide details when you ask questions. What type of led strip, PSU etc. Model numbers, links to data sheets or at least pages giving some details of the specs.

Are you saying both strips are at full, original brightness if the other strip is off? If so, that points to a problem somewhere else - wiring, PSU etc. Measure the supply voltage at the strips when each is on separately and together. Measure the voltage at the PSU in each condition.

Sure Paul, and as you've seen, when I need detailed and more specific answers I normally do post additional info. For this issue though, I did mention it's no big deal and was just curious as to why this might be happening. You are right though, if I wanted specific and detailed answers I would have posted more info, but here... for those who are experienced and know electonics well, I'd have thought the info provided would suffice to suggest what might be happening.

The LED strips are simply plain, 3x1.5V battery powered LED strips, cheaply obtained from eBay, that in my circuit are powered by 5V from a converted PC ATX PSU. There's a 270Ω resistor on each LED circuit, a PIR triggering a relay on each circuit and as I mentioned, the brightness of one set diminishes when the other set is activated.

Surely that's enough info for any experienced electronics-head to visualize the setup and maybe just suggest what might be happening without giving a warts and all answer. That's all I was after, Paul :smiley:

Subsequent to my question though, I have removed the problematic LED strip from my circuit to power it up and test it separately and it eems to be operating fine! On direct 5V power it is brighter than it is when activated at night by the Arduino. Weird though, as it's the very same PSU I am using to do the test that powers the circuit! What's also weird is that the other set of LED's, connected to the same 5V PSU in my set up, are fine! :astonished:

Odd eh?

I'll connect the spare set of LED's I have and see if the same thing happens - at least that will identify if it's the problematic LED strip playing up 'just because', or if it's something else!

No, the info in your original post want enough for even an experienced electronics expert.

When you said "5V led strip", my best guess was that you were using ws2812b strips. These are individually addressable rgb 5V strips. A low value resistor is normally used on the data line.

How much current do your strips draw? Have you measured the voltages as I suggested? What is the wattage of those 270R resistors?

Resistors are one of the most reliable components and they usually last "forever". If they are "overpowered" they can burn-up, and it's standard practice to de-rate them. i.e. If the resistor is going to dissipate 1 Watt, it's common to use a 2W resistor.

...I've never heard of using a series resistor with an LED strip. Usually, the LED strip has resistors built-in and you simply provide 12V or 5V, or whatever the strip is rated for.

Computer power supplies often have trouble supplying 5v when the 12v rail is not under load - they're designed to supply something that uses both 5v and 12v (with a heavy load on 12v), and don't always do well with loading on 5v but not 12v (because they cut corners for cost savings - based on assumptions about the type of load they are expecting).

Better to use phone chargers and butchered USB cables (or purpose built 5v supplies) to power 5v projects.

PaulRB:
No, the info in your original post want enough for even an experienced electronics expert.

When you said "5V led strip", my best guess was that you were using ws2812b strips. These are individually addressable rgb 5V strips. A low value resistor is normally used on the data line.

How much current do your strips draw? Have you measured the voltages as I suggested? What is the wattage of those 270R resistors?

I guess we disagree then Paul. I am no electronics expert by a long long way, but I may have had enough info in an example like this where the OP was asking for a simple pointer, rather than a detailed electronic engineering illustrative answer, to suggest what I might have thought to be the issue. We're all different I guess!

Oh and if I was using addressable LED strips, given that they are far more complex than simple LED's or LED strips, quite simply, I'd have said so, Paul! Another current project of mine is to add coloured LED strips to the stair hand rail in my night lighting rig and I am finding writing the code for these complicated enough - God help me when I move on to addressable ones!!! eeeek!

As yet I've not measured the current, neither have I done any testing or measurements in-situ in the circuit, I've simply removed the LED strip and tested that on its own, and it appears to be OK. I'll do as you suggest though, and will measure voltage and current tomorrow to see if that gives me any pointers. I'll report findings here.

Re the wattage of the resistors, I'm afraid my inexperience and lack of knowledge again rears its head here as I am not sure. They came with an Arduino starter kit I purchased a few weeks ago and the wattage isn't given. So unless there's a way similar to coloured resistance markings, I have no idea how to ascertain the power rating of them! I'll do some googling and see if I can find out how to obtain the power rating. Personally though I'd say it is pretty low, as they are physically quite small resistors, and thin too, as are the wires on them!

DrAzzy:
Computer power supplies often have trouble supplying 5v when the 12v rail is not under load - they're designed to supply something that uses both 5v and 12v (with a heavy load on 12v), and don't always do well with loading on 5v but not 12v (because they cut corners for cost savings - based on assumptions about the type of load they are expecting).

Better to use phone chargers and butchered USB cables (or purpose built 5v supplies) to power 5v projects.

That's an interesting point re ATX converted PSU's, I wasn't aware of that, thanks! I do have a small load permanently across the 12V supply - currently (excuse the pun!!) using a 20W car headlamp bulb!

I might try switching the PSU to a USB one to see if that changes anything.

Or get something that's meant to just put out 5V.