Hi All,
I have an apparent problem with the USB connectivity of the Teensy 3.1 Boards.
I am using them to drive a large number (1920) of WS2812B pixels, using the OctoWS2811 Library.
My hardware setup is as follows;
1) 1 x Teensy 3.1 Board, buffered with a 74HC245 for 5V compliance.
2) 1920 WS2812B LED Chips, arranged in a set of 8 x 240 LEDs, where each set of 240 is arranged in a set of 4, 60 LED strips, in a "zig-zag" fashion
3) 2 x 320W 5V SMPS PSU.
I have had this setup working nicely using both OctoWS2811, *and* the NeoPixel libraries. This is the 2nd time however that I have observed a NTeensy 3.1 fall into an "unrecognized" state. I assumed that the first time I had this happen, I must have shorted something without noticing. However, having had this occur twice , with the 2nd Teensy fully enclosed and protected from shorts, I'm not so sure any more!
Observations;
1) I can see all voltage rails running at their correct levels.
2) I can observe the oscillator package running at 16MHz, at approx. 1Vpp.
3) I can observe USB traffic on both D+ and D- (scoped at the series resistors adjacent to the micro-USB connector, and on the D+/D- pads on the underside of the Teensy PCB.
Troubleshooting:
1) Reset button and driving of it's connected pin on the PCB has no effect.
2) Driving the reset pad, located on the underside of the board has no effect.
3) Windows 7/8.1 both report unrecognized device (both machines have been sucessfully used for teensy development before).
4) Running "lsusb" on my rPi setup (used for FadeCandy) shows no devices listed in relation to the Teensy.
Final Comments / Thoughts
1) Both times I have seen these boards fail, it has been during a session of intense code uploading / serial activity.
2) I am very careful to sequence power correctly; I am only powering half of my array at the moment, but I ensure that the load is manually disconnected before switching off the supply, etc. I would also have hoped that the 74HC245 would have provided *some* protection against serious spiking...
3) Although this is a hobby of mine, I am an embedded electronics engineer by day, with a basis in R&D; Thus, I am confident in my asessment of the current situation!
4) *once* last night, I was able to load the "Blink" sketch, which appeared to bring everything back to life. This is not possible any more it seems!
I suspect that the Teensy is still actually operational (ie, *not* physically damaged). Paul, is there any way for me to dig a little deeper into the operation of the board? I'm used to working with Atmel SAM devices, using a full debugger and IDE setup. I'm finding troubleshooting a little difficult with the limited arduino IDE, and more importantly, I have a client waiting for this project, having already had to wait for me to re-build this controller once already!
*Any* thoughts and feedback are greatly appreciated!
Many Thanks,
Tom Fleet
I have an apparent problem with the USB connectivity of the Teensy 3.1 Boards.
I am using them to drive a large number (1920) of WS2812B pixels, using the OctoWS2811 Library.
My hardware setup is as follows;
1) 1 x Teensy 3.1 Board, buffered with a 74HC245 for 5V compliance.
2) 1920 WS2812B LED Chips, arranged in a set of 8 x 240 LEDs, where each set of 240 is arranged in a set of 4, 60 LED strips, in a "zig-zag" fashion
3) 2 x 320W 5V SMPS PSU.
I have had this setup working nicely using both OctoWS2811, *and* the NeoPixel libraries. This is the 2nd time however that I have observed a NTeensy 3.1 fall into an "unrecognized" state. I assumed that the first time I had this happen, I must have shorted something without noticing. However, having had this occur twice , with the 2nd Teensy fully enclosed and protected from shorts, I'm not so sure any more!
Observations;
1) I can see all voltage rails running at their correct levels.
2) I can observe the oscillator package running at 16MHz, at approx. 1Vpp.
3) I can observe USB traffic on both D+ and D- (scoped at the series resistors adjacent to the micro-USB connector, and on the D+/D- pads on the underside of the Teensy PCB.
Troubleshooting:
1) Reset button and driving of it's connected pin on the PCB has no effect.
2) Driving the reset pad, located on the underside of the board has no effect.
3) Windows 7/8.1 both report unrecognized device (both machines have been sucessfully used for teensy development before).
4) Running "lsusb" on my rPi setup (used for FadeCandy) shows no devices listed in relation to the Teensy.
Final Comments / Thoughts
1) Both times I have seen these boards fail, it has been during a session of intense code uploading / serial activity.
2) I am very careful to sequence power correctly; I am only powering half of my array at the moment, but I ensure that the load is manually disconnected before switching off the supply, etc. I would also have hoped that the 74HC245 would have provided *some* protection against serious spiking...
3) Although this is a hobby of mine, I am an embedded electronics engineer by day, with a basis in R&D; Thus, I am confident in my asessment of the current situation!
4) *once* last night, I was able to load the "Blink" sketch, which appeared to bring everything back to life. This is not possible any more it seems!
I suspect that the Teensy is still actually operational (ie, *not* physically damaged). Paul, is there any way for me to dig a little deeper into the operation of the board? I'm used to working with Atmel SAM devices, using a full debugger and IDE setup. I'm finding troubleshooting a little difficult with the limited arduino IDE, and more importantly, I have a client waiting for this project, having already had to wait for me to re-build this controller once already!
*Any* thoughts and feedback are greatly appreciated!
Many Thanks,
Tom Fleet