1
\$\begingroup\$

I am getting an odd reading from an array of LEDs, which I could do with a little help understanding. I have a very specific project, but to abstract this question from that, let’s assume the following:

I have 24 LEDs in 4 sets, each of 6 LEDs with a 68ohm resistor. They are 1.8v LEDs using 20mA and I have a 12v supply.

schematic

simulate this circuit – Schematic created using CircuitLab

I understand how to calculate the required resistor and current for single LEDs but I used the rather excellent http://led.linear1.org/led.wiz to do the math for me on this. It concludes my array will use 80mA.

However, when I use my meter, it tells me the array is only using 8mA?!

I am planning to run this from my Arduino Micro, which can easily deliver 8mA from a pin but certainly not 80mA, so the difference is crucial to me.

Initially I thought I was reading the meter wrong (it’s new) and I was out by a factor of 10 but the only other thing of a known(ish) current I could benchmark my meter with at 12v was the Arduino itself, which is drawing 37mA, which to me sounds about right (and 370mA would seem too high for a Micro), meaning my meter is probably correct?

I know you can get variations in LEDs or whatever, but I have 3 of these arrays and they are all reading the same ~8mA. What am I missing or should this array be pulling only 8mA? How should a parallel array of in-series LEDs be calculated?

Thanks for any help.

\$\endgroup\$
7
  • \$\begingroup\$ What are you using for the 12V source? \$\endgroup\$
    – vofa
    Jan 23, 2018 at 21:54
  • \$\begingroup\$ A 12v wall-wart, so it is a pretty stable 12v (unlike a car battery). \$\endgroup\$
    – HeMan
    Jan 23, 2018 at 21:56
  • \$\begingroup\$ Does it give a current rating? And are you certain you used 68 Ohm (Blue Grey Black) resistors? If you used 680 Ohm (Blue Grey Brown) accidentally, the numbers work out correctly. \$\endgroup\$
    – vofa
    Jan 23, 2018 at 21:58
  • \$\begingroup\$ Yes, the PSU is rated at 2A. I checked with the meter and it provides a stable 12.13v. And although I doubted myself at first I checked the resistors; blue, grey, black and I double checked with the meter (in case of dodgy eyes) and they are 68(ish)Ohms. \$\endgroup\$
    – HeMan
    Jan 23, 2018 at 23:14
  • \$\begingroup\$ Measure the voltage across one of your resistors. What is the voltage? (It is likely that there is almost no voltage across any of the resistors. But it is worth checking to be sure.) \$\endgroup\$
    – jonk
    Jan 24, 2018 at 0:37

2 Answers 2

3
\$\begingroup\$

How much current flows through each branch of the circuit?

All branches are in parallel with the 12V source, so there is 12V across each branch. $$12V = 6*V_{LED} + I_{BRANCH}*68\Omega$$ \$V_{LED}\$ is about 1.8V, so $$1.2V = I_{BRANCH}*68 \Omega$$ $$\frac{1.2V}{68\Omega}=I_{BRANCH}\approx18mA$$

You have four identical branches each conducting the same magnitude of current, so the total current is about 4*18mA = 72mA. The tool you used gives a reasonable estimate.

If you accidentally used \$680\Omega\$ (Blue Grey Brown) resistors instead of \$68\Omega\$ (Blue Grey Black) resistors, the current will be 10% of the calculation above, giving around 8mA.

\$\endgroup\$
3
  • \$\begingroup\$ I checked checked the resistors, they are definitely 68Ohms. Two other things occurred, neither of which I would have thought matter, but... 1) I am running all three arrays (72 LEDs) and a smaller one (12 more LEDs) on a common supply, and then measuring with my meter in series of one array at a time. They read 8mA but I notice this goes up to 11mA is I run and measure just one array at a time. 2) I have quite a lot of glue from a heat-gun holding stuff in place but if this conducted at all, which I don't think it does, it would raise the mA not lower them? \$\endgroup\$
    – HeMan
    Jan 23, 2018 at 23:17
  • 1
    \$\begingroup\$ @HeMan - Try building and measuring a single branch (6 LEDs + resistor) before doing so on the whole system. If you still get nonsense measurements on a single branch, that indicates you haven't built it properly or are not measuring it properly. Hot glue isn't conductive and won't affect your circuit. \$\endgroup\$
    – vofa
    Jan 23, 2018 at 23:38
  • \$\begingroup\$ OK, unfortunately I only have 4 yellow LEDs left, but I have many red, which have the same specs (admittedly, that's only what it said on Amazon though!) so I made a single run of 6 red and a 68Ohm resistor. I get 0.26v across the resistor, 1.92v across the LEDs (give or take 0.02v) and 12.12v across the supply. And yet, I still only get 5.23mA for the circuit?! \$\endgroup\$
    – HeMan
    Jan 24, 2018 at 20:48
1
\$\begingroup\$

I think you are ignoring the tolerance on these LEDs for Vf.

I have never seen any 20 mA LEDs rated for 1.8V max, only typical.

If in fact they were Vf=2.1 @ 20mA then this would explain why you are only getting 8mA.

What you could do is measure the Vf on a single LED and then measure Vss=12.?V to confirm this is true. Then 8mA*68R=0.5V (approx) so 11.5V/6 = 1.92V @ 8mA which makes sense.

Solution.

Reduce string to 5 LEDs and recompute R for actual Vf . My estimate is 2.1V @ 20mA based on your accurate results. Then (12V-(5LEDs*2.1V))/20mA= 1.5V/20mA = 75 ohms ( so 68 will be slightly more than 20 mA)

Or change it from 6S4P to 5S5P using 25 LEDs

Conclusion.

Pay more attention to tolerances.
All diode Vf variation is due to ESR variation and not the threshold at 100uA or even 1mA.

So if you want more accurate LED current in your designs, specify sources with tighter tolerances on Vf or select on batches , measure vf and choose Rs to meet your tolerance on current. Each batch of 200 LEDs or so ought to be within 1% if they are not mixed up or unsorted or worse untested by cheap and dirty distributors.

\$\endgroup\$
3
  • \$\begingroup\$ You're right, the maths works assuming 1.92v per LED, which is roughly what I am reading. So the question is, can I assume my array will only ever draw ~8mA or so (maybe a bit more as they warm up) and therefore I am OK running them direct from a I\O pin rated to a max of 20mA? I guess the question is, are you best running these things based on the actual readings or based on the specs? \$\endgroup\$
    – HeMan
    Jan 24, 2018 at 20:57
  • \$\begingroup\$ Thankyou for validating my calculations. You must factor the stability of the supply voltage as the biggest variable when the LED string nearly equal the supply voltage. You can easily short out the 68 ohm and drive direct as long as Vcc doesn't rise . Actual readings are more accurate than specs. But you would be wiser to follow my answer solution. But if you trust the PS V+ to be stable, shunt the 68R with an DMM mA mode and measure the current. It should be close to 20mA. \$\endgroup\$ Jan 24, 2018 at 21:26
  • \$\begingroup\$ IO drivers are usually RdsOn of 25 to 50 Ohms so you can reduce 6S string Rs to 0 Ohms in this case. \$\endgroup\$ Jan 24, 2018 at 21:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.