Help with Lattice MachXO3L SMA/DSI breakout board

[liu3hao]

Hi all!

I'm working with a Lattice MachXO3L SMA/DSI breakout board. After trying to use the Reveal Analyser and uploading the code to the FPGA, the breakout board is no longer recognised by my computer. When using the Lattice Programmer, and searching for device on the FTUSB-0 port it says there's "No board with FTDI USB Host Chip detected."

I've done some research and found out from the Lattice MachXO3L Starter Kit User Guide:

An alternate failure mode can occur when the user design assigns an output signal to the FPGA package pin C8
which is connected the oscillator (X1) output signal "12 MHz". This can occur unintentionally when the Placer randomly
assigns unconstrained outputs. In this case, the contention squelches the FTDI device (U1) clock input, rendering
it unable to communicate. To eliminate the contention, remove the resistor R23. This will restore the FTDI
device operation and allow the erasure of the offending FPGA image. Resistor R23 should be reinstalled if an
external clock source is desired.

However this may not apply for the MachXO3L SMA/DSI breakout board as there is no R23 on the FTDI (referring to the MachXO3L SMA/DSI breakout board user guide).

Is there any fix or solution for this?

Thanks for any help or advice!

[kakemoms]

I have this exact same problem, and in my instance I also tried to measure the 12MHz clock that was supposed to come from the X1 crystal. I found no clock  , so it basically means that my board is unable to communicate (through USB) since the FTDI chip has no clock (and thus isn't running). I have a second starter kit board so I wont bother with repairs for now. One should still be able to program the board with JTAG, so if you have one of those you can try to use that and see if it works.

The R23 is basically a shortcut between the 12MHz clock and C8. It seems that older versions of the board had a 0 ohm resistor on R23, so that the MachXO3 could use the 12MHz clock as an external clock source. Since the software connects non-specified outputs to random pins, it could happen that the C8 was connected to a non-specified output and thus rendered the 12MHz clock dead. I think that is the reason they removed the R23 resistor from the later boards.

[kjertil]

RESOLVED!
I had the same problem so i removed R23 - et voilà' - the 12 MHz oscillator is running again! Thanks for the tip!
Next step could be to add in a resistor with the right value so that your FPGA can't pull this line too low for the oscillator to die again.
*EDIT*
A 330 ohm resistor seems to do the trick, the PLL locks to the signal coming in on pin C8.