I've a project that uses a substantial number of LEDs sop I decided to power them separately rather than through the Trinket M0 that I am using. However when I use the power pack to power the LEDs, they simply don't come on. If I use the 3.3v line to power them, they do, but I get substantial dimming in later LEDs and as I also have an OLED screen as part of the project, when I connect that to the 3.3v line, it ovelroads it and resets the Trinket.
I've used the USB pin to power them directly from the USB which supposedly gives 5v, and does reduce the dimming, but this means I'm power the Trinket from the USB port whereas I'd much rather use the BAT pin, again from the 5v 10A power pack.
I measured the voltages across the LEDs power and GND pins and they were not what I was expecting.
3.3V line from Trinket M0 gives 3.0v
The USB out (supposed to be 5v) gives only 3.3v
The 5v 10A power pack delivers 5.5v
All my grounds are united (first thing I checked) so I was wondering if the WS2812LEDS have some sort of overvoltage protection?
I also have a breadboard power module, YWRObot 545043. This will accept the 5v 10A power pack as DC power...and using that works - but it's only delivering 3.3v at the pins to the LEDs again.
I know you'll want a diagram...bear with me, working on it...
OK, diagram one attached - this way, it works if I power the Trinket through USB.
I would have posted a third, but there's no fritzing file for this YWrObot thing...it sits on the power rails on the breadboard. has a barrel jack that I plug the 5v10A power pack into, and outputs power to the rails. Although I've got the jumpers on the YWRobot set to 5v, it only seems to be outputting 3.3v...but it does mean it works - however, in the final project I don't have room for anything so bulky...
Is there a smaller form factor that will enable my 10A power supply to work?
GreyArea:
I would have posted a third, but there's no fritzing file for this YWrObot thing..
That's OK, Fritzings are quite useless anyway. Hard to figure out connections, impossible to read pin annotations (if they exist in the first place), etc.
Can't see anything obviously wrong in the "not working" image but if that trinket puts out 3.3V signals that may very well be your problem as your LEDs expect a 5V signal, especially if powered at 5.5V that's probably simply too low a level for them to work with.
wvmarle:
... but if that trinket puts out 3.3V signals that may very well be your problem as your LEDs expect a 5V signal, especially if powered at 5.5V that's probably simply too low a level for them to work with.
That’s what I did not understand...the 3.3v output from the Trinket WILL power the LEDs, but the 5.5v (measured) output from the power pack WON’T, unless I put it through this YwRobt thing, which seems to reduce the voltage to 3.3v.
raschemmel:
You shouldn't be trying to power the led string from a USB.
I don't see where you mentioned the load current of your led string . What is it ?
Currently, 96 WS2812b LEDs, three channels, 20mA per channel when set to 255, so 96 x 60 mA, = just short of 6A.
I know I shouldn’t power the LEDs from USB since at best it’s 500 mA...but my problem remains that the USB method is working (save some dimming late in the string) whereas the “proper” way (external 5v DC 10A power supply) is not...and I don’t know why.
GreyArea:
That’s what I did not understand...the 3.3v output from the Trinket WILL power the LEDs, but the 5.5v (measured) output from the power pack WON’T, unless I put it through this YwRobt thing, which seems to reduce the voltage to 3.3v.
He's talking about the data line. It's HIGH and LOW threshold depend on Vcc. And 3V3 on the dataline probably does not read HIGH for the LED. It's out of spec.
The datasheet for WS2812B LEDs says >0.7*Vcc for a high signal. So if Vcc = 5.5V (which, by the way, could be out of spec: absolute maximum rating is 5.3V, though a few lines later the same datasheet suggests to use 4.5-5.5V!) you need 3.85V for a high signal. This means that the 3.3V signal for the trinket is not recognised by the LEDs and they stay off.
If powering at 3.5V (the minimum rated voltage) you need just 2.45V for a high signal.
Ah, NOW I get ya...the data for the LEDS has to meet a percentage threshhold of the voltage they receive? I didn't know that.
Which is best practice...to step the 3.3v up to 5v for the data, or to reduce the voltage? I kinda feel Id rather not do the latter given the dimming i see at 3.3v when all 96 leds are connected...
5V is a good supply voltage for those LEDs. Mind that 96 LEDs can consume up to 5A of current if they're all set to white at full brightness, your 5V supply must be able to supply that current. You will also have to supply power to the strip at both ends (run a power cable parallel to your strips), if possible a power point halfway is even better.
Use a level shifter to step up the 3.3V to 5V. The data connection is of course only at the beginning of the strip.
THanks WVMarle. I already have power and earth connected every eight LEDs or so, which is probably overkill but it's convenient to do it that way given how they are arranged. Yes, I've just got the one data in point...thanks for the help...off to find level shifters!
