Soo... my new Arduino DUE just arrived! Super happy? Nooo.
After reading this [http://arduino.cc/en/Products/Counterfeit] page I was sure, I bought an fake arduino board.
I decided to run a simple test:
[20 uSec/Div ~QuantAsylum QA100]
delayMicroseconds(10); > took 14,0us
delayMicroseconds(50); > took 54,5us
My question is: is that a normal result?
The board was powered via the usb directly connected to my computer.
Is there any other test that i can run to see if this board is at least equal in the hardware?
Not sure how you are going to determine that this is a counterfeit by doing this test. Did the board pass the sniff test based on the link you provided? One good test is did it come in a small box, the size of the Due, with some stickers and a long fan-fold warranty paper? That is a give away that it probably not if purchased new without these items.
Regardless, on the signal quality. It appears that maybe your scope probes is not calibrated properly. (pre-compensation)
Have you connected them to the cal port and and adjusted? Adjusted probe on the 1KHz port?
What is the load on the output pin? Into the LED Pin13?
What is the vertical scale?
Yeap, no box with stickers.
Follow the images of the arduino clone.
I alreary have an genuine arduino. So, it was no too difficult to see the differences in the silk.
In a few hours I will re-calibrate my scope, change the pin and try again. [and re-post here]
Is this delayMicroseconds the best function for trying to spot problems in the hardware?
I need to know if this clone will be reliable.
What did the same sketch produce on the "Real" Due that you have? Have you done an A - B comparison?
I doubt it would be any different. 10-50 us cycle time is in the 100KHz and less. They would have really screwed something up to really alter the signal at these freqs. SAM3 through the trace to connector/LED. The diff in time is the the SW prop delay assuming SW timing.
I would just use it and if you suspect something then you have the Authentic Due to compare too.
I should mention the pulse width (while high) bounces between 12.4 and 13.7 us, with most of the pulses being 12.4. Here's another zoomed in view, where I turned up the waveform intensity so you can see the less-commonly-occurring version of the pulse.
If you have a slow scope (which doesn't capture thousands of waveforms per second and render with variable intensity), you can view this by enabling display persistence.
EDIT: those curved edges on your waveform are likely from mis-calibrated scope probes. Connect your probe to the calibration output and adjust the little set screw on the probe until you get a square waveform. Then use the probe on your counterfeit Due.
In your picture of the component side of the pseudo-Due,
it looks like the crystal (for the RTC?) is missing, normally
located about a half-inch to the left of the big uP chip.