Changing Voltage on PS SPI when routed to EMIO
Posted: Thu Jan 07, 2016 3:13 pmHello,
I am using SPI to talk to another board with the Zynq. In order to go to the right pins, I routed the SPI to the EMIO. I took care of the issue with multi master mode by tying SS_0 high, and made the bank voltage where the SPI were being routed to LVCMOS33. Despite all this, I am not seeing 3.3V logic when looking at the signals on an oscilloscope digital logic probe. The voltage is always 2.5V.
I read it was an issue with the PS-PL voltage level shifters not being enabled, however when adding the recommendations on page 47 of the TRM to the FSBL through the fsbl_hooks.c and making sure the operations and registers were correct, the problem persisted.
I am driving the SPI master using the Cadence SPI driver and sending simple signals. Any help on this matter is greatly appreciated. There was another post with a similar issue on the Xilinx forums, but they found the problem to be that their pins were being shorted, https://forums.xilinx.com/t5/Zynq-All-P ... 1086#M7355, which I am confident is not the case here.
I am using SPI to talk to another board with the Zynq. In order to go to the right pins, I routed the SPI to the EMIO. I took care of the issue with multi master mode by tying SS_0 high, and made the bank voltage where the SPI were being routed to LVCMOS33. Despite all this, I am not seeing 3.3V logic when looking at the signals on an oscilloscope digital logic probe. The voltage is always 2.5V.
I read it was an issue with the PS-PL voltage level shifters not being enabled, however when adding the recommendations on page 47 of the TRM to the FSBL through the fsbl_hooks.c and making sure the operations and registers were correct, the problem persisted.
I am driving the SPI master using the Cadence SPI driver and sending simple signals. Any help on this matter is greatly appreciated. There was another post with a similar issue on the Xilinx forums, but they found the problem to be that their pins were being shorted, https://forums.xilinx.com/t5/Zynq-All-P ... 1086#M7355, which I am confident is not the case here.