Problems with dimming LEDs

Hello All!

I’m working on a project which controls some LEDs as well as a 20ohm resistor heater. Whenever the heater is toggled on I get a dimming in the LEDs. The power supply is a 75w switching laptop power supply feeding 18v and the schematic is attached. While not reflected in the image - all tip120 transistors actually have a 330ohm resistor between them and the ATTINY. I’ve tried this on a NANO as well and I still get the dimming.

The LED panels are 18v and pull ~150ma per panel. There are three panels total (LED1 consists of two panels in parallel).

Thinking that the power supply was somehow the cause I went ahead and built a power supply along with dedicated switching regulators per panel to bring the output 18v (PS output is 24v with just the LEDs and Tiny/Nano on and 20v with the heater and everything else enabled) and I’m still seeing the dimming. I’ve even removed everything but one led panel and just toggled the heater on and off every two seconds in my program and the dimming is still there :(. I checked the output voltage on the LED output pin just to see if maybe there was voltage droop when the heater is toggled on but the output voltage is steady on all of the control pins.

Any help with this is very very much appreciated.

The LED panels are 18v...

...Thinking that the power supply was somehow the cause I went ahead and built a power supply along with dedicated switching regulators per panel to bring the output 18v (PS output is 24v with just the LEDs and Tiny/Nano on and 20v with the heater and everything else enabled

I'm pretty sure your 18V to the LEDs is not holding constant. (Depending on how your meter works, you may not see the variation or get an accurate reading on the meter... This is one case where the "averaging" of an analog meter is helpful. An oscillocope would be ideal if you have access to one.)

Try connecting the LED panels directly to the 18V supply, and if they dim the 18V supply is dropping (or "switching") below 18V.

I whipped up a basic breadboard with a button to feed the heater with the laptop PSU and your absolutely right... Whenever I push the button to heat the resistors I get dimming. Its not as bad as through the full project but then I'm only using one LED panel. I guess I need to make the heater draw less current. I earlier did use a quicky 18vac 3.5a power supply rectified and smoothed with two 2200uf caps but I was only using one of the two secondaries on the transformer. If I were to use the other secondary just to handle the load for the heater (and maybe a small 1w air pump) and left the LEDs and MCU on the first secondary would that prevent the dimming as well? I have to build the power supply anyway.

Thanks for the help DVDdoug!

What does the led panel rating of 18v actually mean? What is the turn-on voltage for the panel?

Note in the attached image that near their threshold turn-on voltage LED’s are extremely sensitive to minor changes in voltage. I suspect you may need to operate at a somewhat higher voltage with some current control, perhaps just a resistor. Another approach uses 2 power sources, isolating the LED’s to their own supply.

Joe

ledCurrent.jpg

Each panel is a 4 x 10 row of leds (white 3.2v, 20ma) with a resistor of 300ohm or 330ohm per row. They were assembled a few years ago and used to just run off 16v for a now disassembled project and I've been using them to sort out why I'm getting the dimming. I just assumed that the switching power supply in a laptop would be able to reliably supply at least close to the rated wattage especially given how sensitive laptops are (especially when the batteries are not present).

The toroidal transformer I have does have two secondaries. Would using the second of the secondaries thats just rectified and not filtered as its just for the heating element be enough separation to prevent the dimming?