Honored Contributor
Honored ContributorDave,
This is an excellent development. To FvM's question and your response, the FT232H can emulate the USB Blaster as long as you program the VID/PID of the device. A 3.125 MB/s transfer rate at the FPGA is great. Another thing that I would add is that in case of doing hardware in the loop simulations, this hardware solution opens up more flexibility. For the hw circuits that you describe with a hub+ dual FT232H + CPLD, you could employ one FT232H as a receiver and another one as a transmitter. This would allow the immediate data flow after computation in the FPGA back to the host. Best, James --- Quote Start --- Hi Frank, A UM232H FT232H operating in asynchronous FIFO mode can transfer data at 7.1MB/s PC-to-FPGA (140ns per byte) and 10MB/s FPGA-to-PC (100ns per byte) (SignalTap II traces at 100MHz clock). A UM245R FT245R operating in asynchronous FIFO mode transfers data at between 1MB/s and 2MB/s. If you program an FT232H with the VID/PID of an USB-Blaster, then Quartus recognizes it as such. I'm just working on the USB-Blaster logic now. If that works, then the FT232H solution should be a few times faster than the FT245-based designs. The maximum JTAG speed is limited by the JTAG clock, eg., a Cyclone IV can have up to 25MHz (40ns) period, so a maximum data rate of 25Mbps, or 3.125MB/s over the FTDI FIFO interface. The FT245 cannot max out the JTAG rate, but the FT232H could. After I get the USB-Blaster logic working, I'll test the synchronous FIFO mode. I'm interested in that mode for communications. Cheers, Dave --- Quote End ---
