Hi guys! I wired on my machine some LED strip ~20cm. I power them with cables that are soldered directly on the Vin of the ramps board. I don't know why but when I heat my printer or start printing my LEDs flick continuously. My printer has a 12V 25A mean well power supply so I don't think this is the problem. If my printer is in stand by mode, they don't flick. Could be this a problem of voltage input variation? I mean, in stand by mode yes, my power supply deliver 12V to ramps but when I print or heat up I thing there is a voltage drop. How can I solve this problem?
Edited 1 time(s). Last edit at 04/08/2017 12:03PM by DOYT.
I had this problem as well. It is due to the construction of the LED stripes which have always 3 LEDs and a current limiting resistor in series. Given that each LED needs about 3.5V, this means that there is only about 12V-3*3.5V = 1.5V over the current limiting resistor. Thus, even very small variations on the voltage (due to PWM changes in the load) will cause large variations in the current passing through the LEDs, causing the flickering.
O_lampe's suggestions certainly help a bit, so does also increasing the 12V input by 1 or 2V (if your PS allows this). But for me the only way to really solve the problem completely was to use a a cheap buck/boost (step up/down) DC-DC converter to provide a stable voltage for the LEDs. It costs about $1.50, search for "buck boost dc converter xl6009" on eBay.
I have actually added to the buck/boost converter a little prototype board with 3x button+led+connector and one of those little 3-digit voltmeters (~1$), so I now use this to control the main light, the light mounted at the hotend and the fan cooling the stepper drivers:
Hehe.... I see. You are smart. What I did was to create a 12V stabilizer with LM7812, some condensers and a diode. And connect the LED's at the output of this stabilizer. The input on the stabilizer is 13.5 because I increase the voltage on my power supply and the output is 12V. But......surprise.... the problem persist.
Ok. So I did some tests. I mount two more strips of LEDs. One on the + and - sign of the power supply and one along the supply wires that power the ramps board( I strain the wires with some needles and connect one strip of LEDs there) Now...... something strange happend. During the preheat and after, the LED strips from the Vin from ramps, and the one that was connected with needles on the power cable start to flick but the other one that was connected directly on the + and - of the power supply was not. So my question is: How can a meter of power supply cable can affect this behavior? Because my conclusion in order to solve this problem is to sold my printer LEDs straight on the + and - of my power supply. If I connect them along the wires or at the Vin of the ramps board they will flick. Or another method is to have separate wires from power supply just for LED. Why is this happening?
It's simple: The wires are a resistors on which the voltage drops proportional to the current according to Ohm's law. so the voltage remaining on the LEDs that are connected through the same wire as the rest of the printer will diminish as the current used by the printer increases. So, as o_lampe said in his post, using thicker and/or shorter wires (to decrease their resistance) will reduce the flickering. If you have a voltmeter, try to measure the voltage drop along the wire when the printer is drawing full current...
Anyway, the good news from your experiment is that you obviously have a good power supply with a very stable output.
So from 13.55V is decreasng to 12.98 when it's drawing full current. Yeah, it's a Mean Well 25A. Well... maybe if I'll try with a performant stabilizer like you have it will be good enough to get rid of the flickers. Because as i said, I tried to improve a stabilizer with an LM7812 and some 33 microF capacitors.
I solved this issue by providing a separate power line for lighting directly from the power supply. Switch-mode power supplies can only regulate voltage at their terminals so it can vary a lot at the end of the run.
If the mean well is in good working order it should solve the issue.