Maker Pro
Maker Pro

LED Light Strip Flicker Mitigation w/ DC-DC Boost Controller

I wonder if I could get help with a problem we're having with LED light strips that I installed to replace four (4) dual fluorescent tube light fixtures in our Motor Coach/Recreational Vehicle (RV). The issue is that the water pump, and other sources, are spiking the DC voltage available at the fixtures (see attached O-Scope screenshots*) causing the LEDs to flicker. I’m curious if this problem can be mitigated at the lights by conditioning & filtering the Vdc to the LED strips, using an active DC/DC step-up (boost) switching regulator? The active device performance needs be fast and robust enough to mitigate the spikes shown in the attachments. However, any selected device/design must operate with fundamental & harmonic frequencies outside the Amateur Radio spectrum allocations on HF, VHF & UHF.

I realize that I’m attacking the effect and not the cause (the water pump, nom. 8-10A draw). However the pump is not the only potential source of voltage variations in the RV e.g. exhaust fans, 12V chest refrigerator/freezer, and perhaps a damaged Inverter/Charger (which is scheduled to be replaced).

The input DC at the LED strips is not good (10.75v) due to RV wiring/switches voltage drops. It is supposedly within the LED’s 9-12 Vdc input range, but the LEDs/strips are obviously sensitive to any variations in voltage. The LED strip pairs (2) have a combined current draw of 1.25A per fixture.

I’m looking for ideas and suggestions on device & component selection to achieve the desired results. The end game is to come up with a solution to boost the voltage up to 12Vdc and “filter” the spikes that I can replicate on each of the four (4) LED Fixtures as a DIY home project. Any help would be appreciated as this whole field of semiconductor power management is new to me.

* Variations in waveforms are attributed to variations in water flow demand. (Using a JYE DSO138 DIY Scope Kit I assembled)
1106161748b-crp-small.jpg 1106161846.crp-small.jpg
 

(*steve*)

¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd
Moderator
Are the LEDs bright enough at present?

Are they driven from a constant voltage or a constant current source?

One simple thing to try would be to place an appropriately large cap across the LED. It has the major advantage of not generating rfi :) You might also consider decoupling the capacitor/LEDs from the line with a schottky diode.
 
Are the LEDs bright enough at present?

Are they driven from a constant voltage or a constant current source?

One simple thing to try would be to place an appropriately large cap across the LED. It has the major advantage of not generating rfi :) You might also consider decoupling the capacitor/LEDs from the line with a schottky diode.

The LED strips are sufficiently bright with the voltage at the fixture (but they will brighten up some if the voltage approaches 12 Vdc). The voltage source for the lights is the 12V Battery system in the RV thru disconnect relays, CB panel and several wire assemblies w/ connectors (being it is a Winnebago industries product the wiring schematics and installation diagrams are available).

I've canabalized the flourescent tube ballast PCB and used it as a wire board to try various things. On the PCB is a 1N5822 Low Drop Power Schottky rectifier (ST uElectronics). I’ve tried (parallel) 1000uf + 4700uf caps before and after the diode. I did notice a reduction in noise at the steady state DC voltage but It did not significantly mitigate the flickering. Perhaps that combined capacitance (5700uf) is insufficient, but I could see only a slight change in the pulse waveforms on the DSO with those caps after the diode. [I even tried, at the suggestion of another, to use the secondary winding of a 12V transformer as an inductor with the caps. Unfortunately that configuration only further lowered the voltage at the LED strips making the flickering no less.]

Therefore, I’ve migrated to the “unoptimized” approach (but a learning opportunity) to investigate whether an individual fixture DC-DC boost solution could be developed with the “response” needed to mitigate the problem (regardless of the source of the spikes). That plus we’re replacing the RV's entire 2000W Inverter/Charger system which has been pointed to by others as potentially part of the problem (evidenced by random shifts in brighness of other 12Vdc lights).
 
Top