Teensy3 + RGB strips wiring

Status
Not open for further replies.
Enje said:
i.
V- psu -> teensy gnd -> led gnd
V+ psu -> led +5V
teensy pin16 -> 220 ohm -> led DIN
usb power -> teensy microUSB
Click to expand...


Did you connect pin 15 and 16 together ? The picture doesn't show any of the connections.


Enje said:
ii.
V- psu -> teensy gnd -> led gnd
V+ psu -> teensy Vin -> led +5V
teensy pin16 -> 220 ohm -> led DIN
(this section should require cutting something something, but check on the pictures if I have done that, because I simply never really understood what to cut)
Click to expand...


It's not easy to see it without a magnifying glass but there is actually two pads under the teensy (VUSB and VIN) that are soldered together with a very very thin connection. It's this little connection that you have to cut/remove.
 
Senescence said:
It's not easy to see it without a magnifying glass but there is actually two pads under the teensy (VUSB and VIN) that are soldered together with a very very thin connection. It's this little connection that you have to cut/remove.
Click to expand...
That is why I included some images of the back of my Teensy, because I was not sure if I had done it or not, or where to cut, and what to cut if I had not :rolleyes:
Better yet, how do I test if I have done it correctly anyway?
Would the Leds fail to light up when I am only using USB power then?

Oh yeah, I didn't use OctoWS2811 lib for this test, because I only wanted to figure out how to purely use external power.
Once that part is taken care of, I am sure I'll go over to OctoWS2811 at one point :cool:
 
Hi guys,
I tried to build a 28x16 pixel array following the octoWS2811 tutorial but something went wrong. So I came back to a smaller setup with a single strip led, as documented by Paul Stoffregen, but my setup doesn't work. I built one circuit with Teensy 3.0 powered by USB cable and other one with +5V connected to VIN pin (VUSB and VIN cut apart). I loaded on the Teensy the Basic Test example but it doesn't work.

Please look at the attached pics. Any suggestion for a dummy? :eek:

IMG_0110.jpgIMG_0113.jpgIMG_0117.jpgIMG_0118.jpg
 
Last edited:
Alecmd, I don't know much about this, but when I had a few problems I started from the beginning again. Might be worth going through the build step by step. Load the Blink sketch and see if this works by getting the built in Led to flash. If this works after you unplug your usb cable, you know you are powering the Teensy from you non-USB supply. Check the voltage with your meter to see what you are getting. Load basic test for the octows2811 and hook up just a few leds to pin 2. ensure that you share a common ground for the Leds and the Teensy. The wiring in the pic looks right, so it might be worth checking your sketch, or upload the original basic test sketch without alterations (it still worked for me even with the incorrect number of leds declared. I found that I got flickering if I did not use a resistor (100 Ohm) and needed to have the CPU speed at 48MHz. It sounds a bit patronising, i know, but I found that I had to go back and check the most basic aspects.
 
Looks like it ought to work, from those photos.

This won't solve your problem, but maybe it will help at least confirm if BasicTest is running. Try adding some code to pulse the LED each time show() is called. For example, in BasicTest.ino

Code: 
void colorWipe(int color, int wait)
{
  for (int i=0; i < leds.numPixels(); i++) {
    leds.setPixel(i, color);
    digitalWrite(13, HIGH);  // orange LED on
    leds.show();
    delay(5);                      // always light the LED at least 5ms
    digitalWrite(13, LOW);   // orange LED off
    delayMicroseconds(wait);
  }
}

This can at least give you some confirmation BasicTest is really running.
 
Something went horribly wrong :confused:

I did get it to work using this wiring:

V- psu -> teensy gnd -> led gnd
V+ psu -> teensy Vin -> led +5V
teensy pin16 -> 150 ohm -> led DIN

Basically I ended up with trying a smaller resistor from the teensy to the led DIN.
I went from no light at 220 ohm,
to much flicker with 180 ohm,
to smooth and correct light with 150 ohm.

Could I have calculated this beforehand somehow?
My psu gives 5.10-5.13v dc.

In any case, I got it to run and was happy with that. I let it rest for a while, and when I came back to try it again I plugged in the psu and the light ran smooth like no tomorrow.
Suddenly I see a small puff of smoke, and the light stop.
After looking at the teensy this is what presents itself:
http://ultraimg.com/images/DAafF.jpg

What happened :confused:
Could I have prevented it?

What can I do to make sure it doesn't happen when I try it with the other teensy I have laying around?

(I do have some 1N5817RL diodes, some 74hct244, and some 74lvc245 somewhere. Could I perhaps use any of those to prevent this? If so, how would the wiring be?)
 
Enje said:
to smooth and correct light with 150 ohm.

Could I have calculated this beforehand somehow?
Click to expand...

Not easily. The ideal resistor depends on the type of wires you use and their spacing to each other. If you use coax or cat5 or cat6 cable, the impedance is specified because those cables are made with specific plastics and the spacing the wires is carefully controlled. Most coax cables are 50 of 75 ohms. Twisted pairs in ethernet cables are 100 ohms.

With just ordinary wires that don't have specially made plastic insulation and whose position relative to the GND wire, there's no much you can do. In theory, it could be calculated somehow, but in practice just trying resistors in the 50 to 300 ohm range is much easier.


Suddenly I see a small puff of smoke, and the light stop.

What happened :confused:
Click to expand...

The +5V line touched the AGND pin.

The part that burned is the inductor between +5V and AGND, so this is certainly what happened.


Could I have prevented it?
Click to expand...

Less exposed metal on the +5v wire and/or better position of the wire.


The burned part is an inductor. If you have a magnifier, bright light, and fine tip on a soldering iron, you be able to solder a wire in its place and maybe get the board to work again. It's ok to replace with a wire. The inductor just helps isolate the analog stuff from high frequency noise that might (or might not) be present on the digital wires. Using only a wire will allow the potential source of noise to get into the ADC, but that's a relatively minor issue.

Whether the board can work again depends on what voltage the chip was exposed to as the part burned. If 5V got into the AGND pin on the chip, then the chip probably got destroyed. But if the part burned fast, the chip work again when the part is replaced by a wire.
 
Last edited:
Senescence -

Did you have any luck figuring this out? I took a bit of a break from my huge LED project, but after starting back up recently, one of my 32 strips is exhibiting this exact behavior! Actually, I recently changed wiring a bit and now have 24 strips of 252 LEDs each instead of 32 strips of 168 each, so I'm wondering if for some strange reason, the first LED is seeing too much current on longer strip lengths. This really shouldn't be the case, AFAIK, but I was curious if you were still having the problem, and if so - have you tried powering up a very short strip segment instead?
 
PaulStoffregen said:
Whether the board can work again depends on what voltage the chip was exposed to as the part burned. If 5V got into the AGND pin on the chip, then the chip probably got destroyed. But if the part burned fast, the chip work again when the part is replaced by a wire.
Click to expand...

The board took a visit to the 'doctor' :p
They did everything they possibly could, but the sad news was a tough blow for the 'family' to bear.


R.I.P. big brother. You have always been an inspiration in your ongoing effort to try and keep your fellow siblings safe; by undertaking the most dangerous missions all on your own.
May you be remembered as the loving and protecting brother you always were, and never fade from our conscience.
ioQyydV.jpg
In loving memory of Bob, a proud and quintessential member of the Teensy-clan.
 
I use RGB amplifiers to provide 12v power to my LED strips and an Arduino simultaneously:

41kWe6ryxeL._SX450_.jpg

http://www.amazon.com/TaoTronics-Repeater-Amplifier-Strips-Connectors/dp/B0081CZJGO

More info here:
http://www.instructables.com/id/LED-Strip-Controller-w-LED-Amp-Arduino/

You can control a 3x RGB strips...or 3x 16 foot single color 60 led/m strips off this controller.
Each color requires a PWM channel.

Optoisolated.

I really like the pass through 12v power terminals on the output. I use a big deep cycle battery pack to run Arduino + lights off-grid.
 
Status
Not open for further replies.
Back
Top