LED strips pull MUCH less than expected power draw

Going a bit mad trying to figure out if the LED strip wattage rating is a joke or perhaps a max/burst rating, if I am losing power somewhere in my wiring, if my power supplies are crap or if this is just the way LEDs are...

Built a car port for car. Projects from garage moved to carport, garage somehow remains full. Car remains outside on driveway. Wife is not amused. Need more light for working under carport in evening.

I went with LED strips - I purchased 20m (4x 5m) lengths from here:
https://www.led-shop.com.au/shop/smd2835-non-waterproof-led-flexible-strip-120leds-m/
Yes, more expensive than eBay but I wanted some advice. This mob is local for me, they answered my pre-purchase questions and so I am happy to spend a bit extra to support a local business.

With that out of the way, I have wired 2x 10m lengths as I am going to put each half of the carport on its own switch. From the webpage, LED strips are rated @ 14W/m, so 10m should be 140W.
Therefore @ 12v, I need a PSU to provide at least 11.67A for 10m and double that for 20m.

I have a couple of 12v 30A power supplies which should be sufficient.

I tested the first (new) PSU and I was only seeing amp drawer of 5.56A constant at just over 12v.
Where is the other 15A?
I tested the other much older and abused PSU (same 30A output rating @ 12v) and for whatever reason the LEDs pull 6.65A constant again just over 12v.
I tried 2 strips in series and 2 in parallel but not too much difference.
I tried 1 strip by itself and it pulls just over half the current as with 2.

What is going on?

Instead of using my very cheap and old multimeter to measure current, I put my equally old but trusty in-line RC wattmeter and confirmed those numbers.

Why am I only pulling around 25% of the expected current draw.

The other option is to confirm light output. If I'm getting around the rated output of 1500 lu/meter then the low amp draw turns into a win. I downloaded a light meter for my phone. With the phone approximately 500mm away from the LED strip the meter reads between 290-300lux.
Putting those numbers into a few different lux-lumen calculators gives about ~240 lumen assuming 120 degree view angle.That's under 25% of 1500 but I have no idea how properly calibrated the light meter in my phone is. Still, pretty underwhelming and I think that confirms that the lights aren't as bright as they should be.

Can anyone give any advice on what I might test next?

 

So generally these strips have a number of LED's in series with a limiting resistor. Check the value of the limiting resistor by reading and measuring if possible (you might need to try in both polarities as the leakage could be high on a full strip).

Now, power the strip with your power supply. Measure the voltage at the cut point at the LED's closest to the supply and also the voltage across the two terminals of the resistor. This will tell you how much each cell should be running current wise, and therefore power wise. You can measure the length of the cell and extrapolate how many cells are in the strip. If this comes out to be very different to the main measurement, the discrepancy is a voltage drop issue across the length of the strip - entirely possible for a 10 metre strip.

If the cell measurement is low to begin with and the voltage at the first cell is ok, it is possible they have fitted an incorrect combination of LEDs and resistor compared to their specifications. However LED efficiency can vary kind of wildly and it is not necessarily the case that your lower power consumption actually equals significantly lower light output. Having said that for this kind of strip I would be dubious that you might have been provided the highest efficiency LED's.

Their rating by the way is about 110 lm/W to begin with. Which if you consider that the image on their site already has 3 LED's in series, presumably about 3 V each in the 12V strip, means the chip LED performance should be about 150 lm/W to achieve the specification which is a upper-mid range performing LED - and that hasn't accounted for binning performance spread of the emitters yet. Nonetheless it's pretty unlikely you could be 4 times down in power input and expect specification light output, based on these figures.

 

Thanks for the reply and instructions mtma.
I measured the first 2 resistors with the multi meter set on the "200" ohms setting and I got a reading of 150.5 on both. The resistors have 151 written on them.

I then turned on and measured voltage across components as you said. The first component is an LED, then a resistor, then the 3 LEDs in series.
#1 LED 3.00v
#2 Res 2.94v
#3 LED 2.95v
#4 LED 2.95v
#5 LED 2.95v
#6 Res 2.92v

I then measured the next 3 resistors skipping the LEDs
Res #3 2.85v
Res #4 2.85v
Res #5 2.88v

The last resistor in the 10m series string had .35v
The last LED in the same string had 2.72v.

I didn't quite understand what you meant with the cell measurement? Is a cell each lot of 3 LEDs between resistors?

 

I think you are getting quite a bit of voltage drop from one end of the string to the other.

A good way to measure the current across each LED section individually is to just measure the voltage across the resistor and divide by the resistance (150). The first one is getting 20mA, the last one is only 2mA!

You aren't going to be able to run the full strip with just one power feed, you're going to need to run multiple power feeds with beefy wire along the length of the strips. Start off by powering both ends at the same time and see what the drop is at the middle.

 

I didn't know you could do that. I thought powering the strip at both ends would let the smoke out (short something).
I just tried powering 1 strip of 10m in series - amp draw was 2.76A. I took 5m off and the amp draw dropped to 2.52A.

If I am to power both ends, do I just attach another set of wires to the far end of the LED strip back to the power supply so they both turn on at the same time?
If I wanted to, where do I attach the multimeter to read amp draw as there are 2 inputs and only 1 load?

 

Yep, just wire the other end up, no smoke will come out. If you want to measure current, you’ll just need to join the branch after the meter.

Measure the voltage across the resistors down the chain, that’ll be your best indicator to show the power loss down the strip.

 

Yes, because these elements are all in series and this grouping is pretty much an independent set that's parallel to many other sets along the strip - I'm just referring to that as a cell but you could call it a group or section, whatever. The website picture shows that there are also cut markings between each one.

But as dakiller's already guided you towards, you definitely have a big voltage drop issue. It's no surprise really as there is quite a bit of loading and length across the strip. The copper on the strip itself will just be a thin layer which doesn't help at all.

The back of the envelope is still low but you will be a lot closer. They might have added on some margin in the consumption figure in case they get that guy who buys a PSU exactly rated then proceeds to have incessant problems - although that light rating is probably still an optimistic theoretical number.

 

Well, you broke the First Rule of those 5 metre stick-on LED strips - NEVER daisychain them together, unless you want to run a risk of not only not having enough power go through the end, but also run a risk of cooking the copper tracking on the first bit of the first one.

Look at the specs and Do The Maths:

- 14 watts per metre. 10 metres equates to 140 watts (let's ignore losses at this point for the sake of argument)

- 140 watts at 12 volts means just under 12 amps. Now have a good close look at just how much copper on those strips is supposed to handle that...

Short Answer to your problem there - never join the strips together, and if you do run the power from each end like dakiller suggests. Which as a bonus also gives you the option of putting On/Off switches at each end, rather than having to walk all the way down to one end to turn it off...

 

Thanks for the help everyone. Yep, as suggested I had better results connecting the 10m sections (2x 5m) in parallel which was draws just over 4A. Additional connections the two loose ends (3 connections to PSU) pulls just over 6.5A for the 10m. The strips get warm but not hot at each connection point. The brightness is better connected this way - not heaps, perhaps 20% but definitely a visible improvement over connecting in series, even when powering both ends of the 10m.

 

For that sort of length, your supplier should have sold you 24v strips.

 

Looking at the specs when you select the 24 volt option, ummm but No - still rated at 14 watts per metre...

You'd expect the current draw to be appropriately less if you simply ask Mr Ohms, but if that's still correct specifications as indicated then No.

If that is correct, what you CAN do though if you're keen/mug enough to muck about with the necessaries is run them on appropriate AC.

But from having mucked about with these for over a decade myself personally I wouldn't. On saying that though have done it for a couple of icky installs, that even after pointing out "Not A Terribly Good Idea" the Customer still signed off on (before you ask, usually came down to cost - cheap-arse customers didn't what to pay for quality DC supplies and had low-level AC handy...).

 

Any chance they accidentally supplied 24v strips?
Else as the others have said, the copper is an inadequate dance floor for the number of angry pixies trying to dance.

If at all possible try to run a single metre and see if the results improve more.

You could jam more volts in.. but might break and/or burn something.

 

Um, what on the site is saying that consuming 14W per meter isn't going to follow P = EI ?

Seems like the 24V option should be a different strip, not the same strip at 24V ?

 

They are 12v strips, 3*3v leds and 3v on the dropper resistor makes perfect sense. The issue is voltage drop along the strip, you need to power both ends of a 5m strip length with minimum 18ga wire and you’ll be pretty close to on spec.

 

Hopefully OP doesn't mind me hijacking but while we are on the subject, I too am also playing around with a few led projects and one I was looking at was LEDs behind your screen that MATCH the video you're watching. As shown here.

In the video it appears that he is powering a 12v led strip and a raspberry pi with just a 5v power supply. Would this work?

 

So WS2812 LED strips (also other strips like it) aren't your normal LED strips. The LED's themselves are 5V, if there's a 12V version it will be (should be) using some kind of regulator or supply.

They're digitally controlled and manage the dimming and hence current draw, across the tricolour emitters themselves.

Being so different, it's worth having your own thread if you have further questions.

 

Measure the 12v supply at 1m intervals when fed from one end, you will see a large voltage drop. The solution here is to solder in additional supply points along the length, supply both ends of the 5m strip as well as in the midpoint with 12v with decent gauge wire.

 

Eh? Where have I said they're changing the Laws of Physics, Captain? I simply indicated that higher voltage for same power draw will draw less current...

Correct, sorta. They usually either change the size of the individual limiting resistors, or change the LED connection setup to run at the higher volts (as in run more series LEDs per segment).

Can't recall seeing a 24 volt one that had a bigger sticky strip than the usual 12 volt ones.

 

Nup, or I'd be surprised if they even got a glimmer of light down that low - the 12 volt LED strips run 3 LEDs in series per segment plus a limiting resistor, thus at 5 volts total each LED would initially be trying to run at less than a third of the supply volts.

I've got some high efficiency 5mm LEDs 'round here somewhere that would glow at just under 2 volts, but the ones used on LED strips would barely show any life at all down that low.

 

According to the video, he's running a 5v LED strip.