Hardware PWM IP Updates

[TJLW]

@jaredrholley Suggested there was an issue when testing the PWM IP in software. A simple PWM test that bridges a PWM Writer port to a PWM Reader port did not behave as expected. The supplied duty cycle does not match what is read on the input port.

This could be a timing issue introduced by the ZynqRobotController system or could be an issue inherent to the IP themselves. Pre-synthesis tests have been performed and the testbenches are included within the packaged custom AXI IP.

Currently uncertain how to proceed:

• Test hardware IP in standalone fashion, update with more registers for controlling/getting status of frequency. Also allow reading back supplied duty cycles.

or

• Dive into the timing issues Vivado is throwing. Simulations work so maybe a smaller SoC system (only Zynq7, AXI, and PWM IP) would indicate if it is the current setup or the IP introducing the error.

[TJLW]

Update for this. The PWM generator had some issues that were attempted to be fixed. However, in these efforts, the AXI Timer IP was discovered to ave a PWM capability. This method has been successful with most things though there are apparent current issues as some ESCs (namely the VXL8S. A dedicated PWM driver breakout board produces the same PWM but the IP does not get any response. I need access to a GOOD oscope to look deeper. For now the AXI timer will be replacing our PWM generator function. Upload coming soon.