Shop OBEX P1 Docs P2 Docs Learn Events
PropBOE Power Supply Problem — Parallax Forums

PropBOE Power Supply Problem

RobertDyerRobertDyer Posts: 23
edited 2013-07-26 14:25 in Propeller 1
Recently I have been having a very elusive problem with some smart, serial LEDs (Adafruit NeoPixel Stick w/ 8-WS2812 LEDs). I have the stick connected to one of the servo ports (pin14), and powering the stick with the 5V from that servo connector. I have the power selection jumper connected to the 5V side not the VIN side.

When powering the BOE with one of the Parallax 7.5V wall warts the LED strip just won't behave, regardless of whether or not the USB cable is connected. When I go back to using only the USB connector for power, the LEDs work fine. It appears the I/O pin, or the DC supply pin, get so noisy that the LEDs flicker and act in other inconsistent ways.

I've measured the output of the wall wart, and it is 7.6VDC with only a couple mV of ripple (unloaded). I don't have a scope to check the 5VDC voltage output for ripple, but my DVM only shows about 20mVAC (loaded), though there's no way to tell what the noise waveform may look like. I thought it might be a ground loop problem between the wall wart and my PC, so I reversed the AC connection at the outlet with no change. Next I tried using batteries connected to the barrel connector (no ground loop) - still a problem. ANYTHING connected to the barrel connector seems to give me problems. It appears to me it's an issue with the power supply circuit.

Could I have burned out a component of the DC supply that is making the switcher VERY noisy? I tried connecting a fairly large (2200uF) cap across the 5V supply, but that didn't help at all. Is this a problem that anyone else has reported? If so, can you suggest a fix for me?

By the way, I've had the opposite problem before when using a servo, where the servo appeared to load the USB supply enough that the BOE wouldn't work. I had to use the wall wart to make the servo not mess up the CPU.
Anything you could suggest would be helpful. As it stands right now, I have to be very cautious how I develop with this board.

Thanks.

Robert

Comments

  • David CarrierDavid Carrier Posts: 294
    edited 2013-07-25 10:04
    Robert,
    The WS2812 LEDs require an input-high voltage of at least 70% of VDD. If you run it off of 5 VDC, it will require at least a 3.5 volt signal, which is a higher voltage than the Propeller microcontroller will generate. If you run it at 4.5 VDC, it will work with a 3.15 volt signal, which would work. USB ports can supply power anywhere between 4.5 and 5.5 VDC. The power supply on your USB port may be running a little low, allowing the WS2812 LEDs to work with the Propeller microcontroller.

    Try buffering the Propeller's I/O pin using a 5 volt buffer. It should condition the signal to a voltage range that will work with the
    WS2812 LEDs, no matter which power supply you are using.

    — David Carrier
    Parallax Inc.
  • RobertDyerRobertDyer Posts: 23
    edited 2013-07-26 05:07
    Thanks David, Yours is an interesting idea. I'm guessing this means there haven't been many reports of problems with the PropBOE power supply. - Others on this forum have mentioned that we're running the WS2812s past the ragged edge, though they generally seem to function at 5V. I checked the USB supply at the power pin of my LED strip and it's 4.91V. So not sure this is the problem for my particular instance. I'll keep looking. Any other suggestions would be much appreciated. I've considered pulling the output pin up to 5V and then playing instead with the direction pin of the Prop. I thought I might leave the output set to low, and then switch the direction from output to input so the pin would be alternately driven low and then pulled up to 5V. Is this done by others? Robert
  • David CarrierDavid Carrier Posts: 294
    edited 2013-07-26 14:25
    Robert,
    I was thinking of suggesting just that. If you use a 3.9 k ohm or greater pull-up resistor, it will keep the current through the ESD protection diode within the maximum ratings, and the 0.3 volt drop from the ESD protection diode will allow the pull-up resistor to pull the line to 3.6 volts, enough to register as high on the WS2812. It isn't an ideal solution, as the voltage is right at the threshold, and you will have to modify your driver object, but it should work in a pinch.

    — David Carrier
    Parallax Inc.
Sign In or Register to comment.