Teensy 3.1 dead... or almost

[luckyg3]

I just started using a teensy 3.1 from Adafruit yesterday and tried a few sketches just to test things out. After loading one of them it did not appear to be functioning. I went back and tried a simple 'blink' sketch and it would not even detect that the teensy was connected. I went through the PJRC troubleshooting page and several forum posts for testing a dead teensy. I checked voltage on several suggested pins as noted in the following post:

https://forum.pjrc.com/threads/28539-Ho ... teensy-3-1

Program was only showing about 0.1V regardless of pressing the pushbutton or not. And reset was zero volts as well... not 3.3V as expected if the push button is not pushed.

So at that point I was thinking it was dead.

I tried loading 'blink' again today and it did actually load... but the led is flickering (not a steady blink) and sometimes staying off for several seconds (as if not powered). I've tried both USB power and an independent 5V supply with no difference in results. When the USB cable is used, the computer acts as if the unit keeps connecting/disconnecting though yesterday it did not appear to connect at all. I've checked for 3.3V on the output pins and I don't see any change between when the LED is flickering and when it stays off. I believe I saw a low voltage on the program pin briefly, but other times I checked it showed 3.3V which it did not yesterday.

Any thoughts? It's definitely not working properly with just a simple 'blink' sketch.

[adafruit_support_mike]

It does sound like something is failing.

When did you get it?

[luckyg3]

I ordered it on May 21st.

[stevech]

Flickering LED clue.
Loading a known good program is always the first step with Teensy - as it uses USB to download and if the downloaded program is hosed up, you get unpredictable action. The PROGRAM button on the T3 invokes the second microprocessor on the board. That little micro copies bootloader code to the main ARM processor which then executes that to commence the download.

USB cable? Some I've encoutered have wire gauge so small they can't pass much of any current without excessive I * R drop.
A DVM (meter) might not show a pulsing voltage.
Bad USB hub?

if you ever get stuck for help, the teensy user forum will save you!

[adafruit_support_mike]

Do try another USB cable.. there are a lot of flaky ones out there these days.

If that doesn't do any good, send a note to [email protected] with a link to this thread and your order number and the folks there will get you a replacement.

[luckyg3]

Stevech, you're saying the program upload itself might be messed up by a flaky cable, correct? Obviously power isn't the issue because it works the same on a stand alone PS. I'll try another cable when I get a chance...

[luckyg3]

Another question based on your hunch... is there a way of measuring a USB cable's individual wires to calculate how good it is?

[adafruit_support_mike]

There are, but you need some equipment to do it.

To check a cable's current capacity, hook the output end to a low-resistance load like an automotive tail lamp, and the input end to a power supply with adjustable current. Set the voltage to 5v and the current to 100mA, and measure the voltage between the input and output ends of the positive line. Increase the current to 200mA, 300mA, 400mA, and 500mA measure the cable loss each time. It doesn't hurt to put an ammeter into the line so you can verify the amount of current at each setting.

To check the data lines, connect the output ends to an oscilloscope and the input ends to a pulse generator (an Aruino can do the job). Set the scope's trigger voltage to 3.5v, send a 1MHz square wave into the input side of the cable, and see what shows up on the screen. If the pulses come through cleanly, raise the frequency of the input wave in steps until you get to 12MHz. If the output gets bad enough that the 'scope can't trigger reliably, the cable won't be able to support data transmission at that rate.

[stevech]

Another question based on your hunch... is there a way of measuring a USB cable's individual wires to calculate how good it is?

No. I'm saying that some USB cables have small gauge wire and the I * R drop can be large enough to cause the 5V USB to fall too low for brief times, as the current demand changes in time. At the 1/10th of a second pulse level.

None the less, go to the Teensy user forum and bring this up. If the users' consensus is bad Teensy, and the board is new, PJRC/Paul will get you a new one, direct or via the seller. He's good about such things.

[stevech]

Another question based on your hunch... is there a way of measuring a USB cable's individual wires to calculate how good it is?

The crummy ones have a cable that is much more flexible or small in diameter - due to the wire gauge. No easy way to test. Best to stock up on varied USB cables so you can troubleshoot by substitution.
And don't use a USB hub if you have such problems.