I am designing a custom board that requires 5V@3amp PS and I have found a few that are listed as regulated but does that mean they won’t output more than 5V for sure? Even if that is the case is it cheap insurance or something to cause to problems? This source will power the WS2812B LEDS, VIN on Photon and logic shifter through a MBR120 Schottky diode. I have the Photon, 24xWS2812B (calc at full white is 1.44 amp even though I have no code that would go full white on all LEDs), HTU21 500 uA, logic level shifter 10 uA and possibly some indoor air quality sensor. I am counting the overage of using the full 60 mah towards the oversizing the PS so it isn’t running at max all the time. Here is the circuit and board for those interested. 5.5mm barrel connector and right-angle smd switches are on the bottom layer hidden under the leds and such.
I am using 3v3 to power the HTU21 and one CA RGB LED (indicator led)…is that ok?
Thoughts and feedback are very welcome!
@LukeUSMC, if you expect 1.44 amps through the LEDs then an MBR120 won’t do the job (max 1A). You’ll need a 2A or 3A rated diode to allow for a wide margin. The quality of the output of a power supply is related to the quality of the power supply itself! Perhaps you can post the model you propose to use.
Powering the HTU21 and an RGB LED via the 3V3 output is fine with the current limiting resistors you have. I recommend 0.1uF decoupling caps near the Vdd lines of the HTU21 and the Photon. I also HIGHLY recommend a large electrolytic buffering cap (470uF) between Vdd and GND to moderate any large current/voltage fluctuations caused by the LEDs switching.
1 Like
@LukeUSMC, if you expect 1.44 amps through the LEDs then an MBR120 won’t do the job (max 1A). You’ll need a 2A or 3A rated diode to allow for a wide margin. The quality of the output of a power supply is related to the quality of the power supply itself! Perhaps you can post the model you propose to use.
Powering the HTU21 and an RGB LED via the 3V3 output is fine with the current limiting resistors you have. I recommend 0.1uF decoupling caps near the Vdd lines of the HTU21 and the Photon. I also HIGHLY recommend a large electrolytic buffering cap (470uF) between Vdd and GND to moderate any large current/voltage fluctuations caused by the LEDs switching.
Thanks for the feedback! I’ll make those tweaks for sure. Should I do the ferrite bead, electrolytic cap .10uF and 100uF (I think those are the values) on 3v3 so it is “filtered”?
As for the PS I was looking at: https://www.amazon.com/Power-Adapter-2-1mm-Regulated-Supply/dp/B006QYRVU6
OR
https://www.adafruit.com/products/276
The second one is a 2A but I calculated the 1.4amp for neopixels was at 60mah per pixel but like I said before, I don’t have any code that goes white on all pixels. In fact no white on any pixels.
@LukeUSMC, if you for sure won’t exceed 2A then I would go for the Adafruit one, if only because they carry it.
@peekay123 so does that mean I wouldn’t/shouldn’t put a LDO Regulator on VDD “IN”? The PS will regulate everything? I just don’t want to burn up pixels over a $2.00 part.
@LukeUSMC, no need for an LDO with a good supply. As per Adafruit's description:
This switching supply gives a clean regulated 5V output at up to 2000mA
I use that supply to power a 16x32 RGB panel. You can also get the larger, higher current supply which I use to power a 32x32 RGB panel:
1 Like
@peekay123 is there a “safety factor” in voltage rating I should be looking for on the 470 uF cap? Do I double the supply voltage or is a 6.3V enough overhead? Also other than a tantalum (too expensive) are there alternatives that I could/should at over the can style caps?
@LukeUSMC, doubling the voltage is always good but not necessary on a regulated DC output. I am assuming you don’t want to use a “large” electrolytic? Is it size or cost that is concerning you?
No real concern on the can style other than my OCD. aside from my OCD, all smd components is my goal which I know I can do in aluminum electrolytic. Just trying to explore all options. Cheaper is generally better as long as it doesn’t sacrifice reliability/lifecycle.
@LukeUSMC, you could put a number of 100uF ceramic SMD caps in parallel to acheive a simlar affect.
@peekay123 those would be in addition to the 100uF ceramic caps I already have spread throughout the ring?
I’ve not seen capacitors added within strings of these WS2812 LEDs, why do you have them there? Are they actually embedded on the strings and I never noticed?
@LukeUSMC, you should change the 1uF caps you have at each LED to 0.1uF caps which are more suitable decouplers. The large cap I recommended is for the main power entry to the board so as to handle current spikes when a lot of LEDs are switching. Even then, with a good power supply rated above max current requirements, the large cap is optional.
This is one with caps like my design. I read their UberGuide and see the large cap that @peekay123 recommends, they called it a “smoothing cap” and I guess the smaller caps are focused on decoupling.
For round one of the design I referenced this shield…been using one in my latest proto design actually but that doesn’t mean that it is done perfectly so if someone like @peekay123 recommends that I make some modifications…I am not of a knowledge level to doubt!
EDIT: I looked at the datashert and the typical circuit design calls for a 104 cap which is the cap code for a 0.1uF cap…just as recommended. I’ll be updating the values on the schematic and already added a large cap for “smoothing”.
1 Like
Ok…I found a tantalum cap I that wasn’t crazy expensive so went that way. Here is the latest:
@LukeUSMC, looking good! What is the trace size of the power traced going around the neopixels?
With power “trace” visible…
1 Like
I’m pretty proud of this so far…may add a few sensors for indoor air quality and send it off for printing. I wish there was a legit pollen sensor that didn’t cost hundreds to thousands. Need lasers to get level but may go with dust and the nasty gases that could end up in a kid’s room from furniture, toys, stuffed animals etc.
1 Like
