# 16 Volt on a 12 Volt LM317 Regulated Power Supply

I recently purchased a 1,2-12V regulated power supply with an LM317 chip and accordingly, a 220V to 12V transformer.

Because I did not need the added functionality of the "logic pen" facilitated by an included IC, I unsoldered it only to find out that without it the power supply could be cranked up to 16 Volts (with no other modifications).

My first question is: How are 16 Volts possible if the transofmer is rated to output only 12 Volts?

My second question is: When trying to drive 96 standard 3,5V LEDs connected in parallel, I noticed that they were rather dimly lit, even when the power supply was set to max. Having found out that the power supply can only deliver a max of 2 Watts and accordingly 2/3,5 = ~571mA I was wondering what could be done to increase the maximum current it can handle seeing as the LM317 can handle a max of 1,5A.

Below is a schematic of the power supply (with the only difference that C1 & C4 are 1000uF instead of 680uF)

Powering 96 led connected in parallel is a very bad design in the electrical perspective. Totally inefficient because it demands very high current. Try to group it in 12 parallel groups with 8 LEDs in series in each group. In this way you may power all these LEDs using 16 volt. Each series string will require 20mA and total current demand of the 12 groups will be 240mA at max.

Why 16v when the trafo is rated to only 12v? Because 12v is the RMS rate of the trafo ac output, but when it is rectified to DC and smoothed by a cap, the cap will be charged to the peak of the RMS output, which is 12x1.414=17 volt.

Why the 1.5A capable ic snubs at 560mA? Because the ultimate current capacity is limited by the capability of the trafo. Your trafo must be rated to supply only 500mA at max.

The last thing is, this question must be posted in the electrical engineering site, not here.

How are 16 Volts possible if the transformer is rated to output only 12 Volts?

The rating usually is expressed as rms voltage, not peak to peak: those 12V are actually $12 \cdot \sqrt{2}$ Volts, so approximately 17V. If you subtract the voltage drop caused by the diode brigde (usually 1.2V - 1.4V) you get the 16 Volts you measured (more or less, of course, the mains voltage is not precisely 220Vrms).

When trying to drive 96 standard 3,5V LEDs connected in parallel, I noticed that they were rather dimly lit, even when the power supply was set to max.

You usually can't simply parallel the LEDs and expect them to work properly, since their I-V curve is very sensible to each LED's characteristics. Two (almost) identical LEDs may sink quite a different current with the same voltage applied, due to the exponential I(V) dependence.

Anyway, as the datasheet says, "It includes current limiting, thermal overload Output Safe-Area Compensation protection, and safe operating area protection". The problem might be this self-protection operation. Does it get really hot when powering the LEDs? In this case, thermal dissipation might be necessary.