Do I really need a 5v signal for the Data pin on WS2812b's? (Teensy 3.2/Audio shield)

Status
Not open for further replies.
propa

propa

Well-known member
I'm trying to get an electret microphone working with the audio board to control LEDs, I haven't soldered any pins to these new boards just yet, for fear that maybe the audio board and 3.2 isn't the best method.

My original idea was to stack a 3.2, prop and audio shield together, found out you can't stack the audio on prop, so went to just the audio board on 3.2, but 3.2 doesn't have any buffered 5v pins for leds, so I can't control any WS2812b LEDs without additional circuitry. Or can I?

But I'm wondering if the data pin for WS2812s really need exactly 5v and potentially have a margin of error built in. The reason why I say this is because I've done some work with a WeMos ESP8266 chip and I'm pretty sure it's 3volts as well, I've powered the LED strip from the end and fed the gnd/power from other end of the strip into the WeMos to power it, fed the (3v) data signal out of the WeMos into the DataIn of the strip and it worked fine. So i'm wondering if you really need a 5v Data pin signal for 3pin LEDs?

Has anyone got anything to add before I solder anything unretrievable or add needless connections.

My naive intuition would assume it will be fine, or did it only work on the WeMos because the pins are 5v tolerant? I assumed 5v tolerant meant the pins could take up to 5v, not make 5v from thin air.

Has anyone else directly connected to the data pin of a WS2812b with a Teensy 3.2 and had success?
 
Most of the WS2812B strips made after 2015 work with 3.3V signal input. The oldest ones from 2012-2013 also work (most of those had 6 pins on the LED body, 2 were no-connect). But from 2013-2015, something changed in the WS2812 which didn't work with 3.3V signals.

3.3V is still just slightly below the Vdd * 0.7 spec. Of course, the published specs on these products are notoriously unreliable. Still, a full 5V signal should be considered the "proper" approach. If you're building just 1 project and things are working for you, I'd go with "if it ain't broke, don't fix it". But if you're making something that will be copied over and over like a professional product or an online tutorial, using a 3.3V signal is a liability for the future.

A full 5V signal also gives you more margin to tolerate noise or interference. Again, probably not a concern if you're making just 1 project that will be used in a fairly benign environment. But if you're building something that will be taken to an event like the huge Maker Faire in San Mateo, consider it may end up in a room next to dozens of other poorly designed maker projects which emit a tremendous amount of electrical interference. Noise margin makes the difference between something that's rugged versus only works when conditions are just right.
 
PaulStoffregen said:
Most of the WS2812B strips made after 2015 work with 3.3V signal input. The oldest ones from 2012-2013 also work (most of those had 6 pins on the LED body, 2 were no-connect). But from 2013-2015, something changed in the WS2812 which didn't work with 3.3V signals.

3.3V is still just slightly below the Vdd * 0.7 spec. Of course, the published specs on these products are notoriously unreliable. Still, a full 5V signal should be considered the "proper" approach. If you're building just 1 project and things are working for you, I'd go with "if it ain't broke, don't fix it". But if you're making something that will be copied over and over like a professional product or an online tutorial, using a 3.3V signal is a liability for the future.

A full 5V signal also gives you more margin to tolerate noise or interference. Again, probably not a concern if you're making just 1 project that will be used in a fairly benign environment. But if you're building something that will be taken to an event like the huge Maker Faire in San Mateo, consider it may end up in a room next to dozens of other poorly designed maker projects which emit a tremendous amount of electrical interference. Noise margin makes the difference between something that's rugged versus only works when conditions are just right.
Click to expand...

Excellent, thank you for the information, it seems to work fine (for now), Audio board with a mic attached controlling 20 odd WS2812s via fft, haven't had a problem! Maybe if I scale up using more LEDs I will go to level shifting for peace of mind.

I think I may be pushing it if I power the Teensy from the USB, I think theoretically I could power the strip from VIN, USB is giving 5v 0.5A, or if using a usb phone charger 1 Amp or so. Is it a bad idea to put that amount of current through the VIN pin?

EDIT: https://forum.pjrc.com/threads/27859-Max-current-through-VUSB-to-PCB There's a fuse, I may give it a crack without worrying so much!
 
Status
Not open for further replies.
Back
Top