MAX10 on custom board fails to configure: "Error (209014): CONF_DONE pin failed to go high..."

[DEMandell]

Greetings,

We are using a MAX10 FPGA, specifically 10M02SCM153, on our own development board, and most of the boards I have tried fail to configure, giving the error "Error (209014): CONF_DONE pin failed to go high..." I'm using Quartus Prime version 23.1std.1 with a USB-Blaster II and have tried both Windows and Linux with exactly the same results. One board I have been able both to configure (using .sof file) and flash (using .pof file) repeatedly, however, right now although this one board can be programmed most of the time, no other boards I have will configure. The other circuitry on the boards works, so I know it isn't a power supply problem.

We followed the guidelines in the Pin Configuration Guide document when preparing the schematic, and the one board that works can be programmed at up to 6 MHz so I'm fairly confident it's not a signal integrity problem inherent in the board's design. I have also examined the JTAG signal waveforms on an oscilloscope and they look fine.

I found one discussion from 2019 (link to post) in which @Abe stated, "The DEV_OE pin needs to be pulled low as well as the DEV_CLRn pin." Is this in fact the case? The PCG does not say this, nor does any other reference I was able to find. In our design we don't use the DEV_OE, and left it open, which the PCG says is valid: "... or leave the DEV_OE pin unconnected provided that the Enable device-wide output enable (DEV_OE) option is disabled and not used as a user I/O pin." We are using the DEV_CLRn pin, but pulling this pin low during JTAG configuration makes no difference.

It seems as if the JTAG interface is generally working since I am able to Auto Detect the device and I can successfully Erase the flash, and Examine it too, and when it tries to configure a device, it is able to read the JTAG ID code, but no other functions work on four out of five boards I have here, giving the same error.

If the cause of this problem is something in the circuitry that we need to fix, we do have a board spin coming up, but it's critical that we get it right because we need these boards for an upcoming high-profile set of demonstrations.

All assistance gratefully accepted.

Thanks.

[DEMandell · Accepted solution]

Greetings again,

Mystery solved! It turns out there is nothing wrong with our board or the programming procedure.

It is powered from a "lump-in-line" 24V power supply, and we have various versions of these from a variety of vendors. It just so happened that the one I was using to test some of these boards generates a high-frequency noise EMF between its negative output terminal and mains ground such that it will induce several volts (!) of HF noise due to currents flowing from its negative supply terminal back through the ground wiring in the USB Blaster II to the computer's ground connection, wreaking havoc on the JTAG signals. I have no idea how this power supply design ever pass EMI compliance but it has all of the usual logos (FCC, CE, etc.) on it!

Anyway, using a different power supply module solved the problem completely and all boards can now be programmed. I foresee a number of power supply modules hitting the e-waste bin very soon, or being relegated to powering LED lighing only.