Stepped rising edge waveform on Arria 10 1.8v output

Hi Like,

I can confirm that this is not PCB / track loss, or a power problem. I've performed a whole load of other tests and communicated these to our Intel supplier and so will give a quick update here.

One of them measured the issue under the pin of the FPGA (track length < 1mm + a via), where it was still seen - this is the "a10_1.8v_at_pin" image from my first post, with the poor rising edge.

My latest test is to compare that same pin using the 50 Ohm Uncalibrated vs. Calibrated options. These two 'scope traces are attached and make it very clear that it's NOT a PCB or power issue. The Calibrated option is spot-on with a perfect rising edge. Uncalibrated gives poor results at the pin and are unusable on longer tracks.

From what I can see, Quartus is setting the wrong pull-up current strengths for the 10AX032 device we're using.

Please take a look at the attached 'scope shots and let me know what you think.

a10_1.8v_at_pin = Copy of previous test. Uncalibrated, toggling 1.8v under the FPGA pin, no track connected (just <1mm track + via).

a10_1.8v_at_pin_calibrated = exactly the same, but using the Calibrated option and a 100 Ohm RZQ resistor & OCT block.

Thanks,

Richard