MAX10 - 10M02DCV36I7G - BOARD - High current draw or shortcircuit ?

[Altera_Forum]

Hi all, 

 Target MAX10: 10M02DCV36I7G Quartus: 15.1  We are trying to design a very minimalist and small MAX10 FPGA board using the 10M02DCV36I7G. The schematic (attached) of the board is very simple and only implement the following features :  

• Single 3.3V supply power : VCCIO3_5_6 = VCCIO_1_2_8 = VCCA = VCC_ONE and are all connected to a single 3.3V LDO regulator 
• We aim to implement a SPI compatible bus. Considering the small amount of IO on this chip we decided to use the JTAG Pins Sharing features 
• The FPGA is clocked with a 50MHz Resonator 
• And a small led is connected to one of the digital I/O 
• Other IO are connected to board pins or left unconnected 

  the problem :  When we power the board the board draw a large amount of current (>400mA) and the voltage drops. So there must be a short-circuit somewhere but we are unable to locate it. We tried several approaches:  

• Double check the schematic (cf. schematic.pdf (https://www.alteraforum.com/forum/attachment.php?attachmentid=13795&stc=1&d=1498232560)) and 
• Verify FPGA schematic pin assignment using the pin table (https://www.altera.com/content/dam/altera-www/global/en_us/pdfs/literature/dp/max-10/10m02dc.xls) and pins connections guidelines for max10. (https://www.altera.com/content/dam/altera-www/global/en_us/pdfs/literature/dp/max-10/pcg-01018.pdf) 
• We already used a very similar schematic for a 10M02 EQFP144 that have been validated with working board. 
• Re-Checked pin order between our PCB Footprint and Quartus Prime Pin Planner view founding no differences. (cf. pinnout-comparison.pdf (https://www.alteraforum.com/forum/attachment.php?attachmentid=13794&stc=1&d=1498232555)) 
• Asked the manufacturer which sort of test they performed: Their answer “all the board passed assembly optical inspection (aoi) and solder paste inspection (spi)”. So apparently it shouldn’t be a soldering issue. 
• Unsuccessful JTAG chain test with the board connected to a 450mA – 3.3V power supply. 

   We probably missed something and are now out of ideas.   Thanks for your help. Tim

[Altera_Forum]

What do you have unused pins set to in Quartus (Device settings -> Device & Pin Options)?

[Altera_Forum]

Hi sstrell, 

 Thank you for taking the time to answer.     --- Quote Start ---  What do you have unused pins set to in Quartus (Device settings -> Device & Pin Options)?  --- Quote End ---    Currently unused Pins are defined on de default value : "As input tri-stated with weak pull-up".  Best Regards Tim

[Altera_Forum]

This sounds familiar: 

 https://www.alteraforum.com/forum/showthread.php?t=55817  Sorry, I can't help you. But if you figure out the problem please post back here as it may help others who are seeing the same thing.

[Altera_Forum]

--- Quote Start ---   Target MAX10: 10M02DCV36I7G [*]Single 3.3V supply power : VCCIO3_5_6 = VCCIO_1_2_8 = VCCA = VCC_ONE and are all connected to a single 3.3V LDO regulator   --- Quote End ---    10M02DCV36I7G is not a single supply device - and as your schematic shows, there is no VCC_ONE because this part needs separate supplies for core (1.2V) analogue (2.5V) and I/O (could be the same as one of the other supplies, but you appear to need 3V3).  I don't think that the single-supply parts are available in the WLCSP36 package size.    --- Quote Start ---   [*]Asked the manufacturer which sort of test they performed: Their answer “all the board passed assembly optical inspection (aoi) and solder paste inspection (spi)”. So apparently it shouldn’t be a soldering issue.   --- Quote End ---    AOI isn't enough to prove you have no soldering faults - you'd need X-Ray inspection for that on a BGA.  But probably in this case you've simply blown up the device by applying 3V3 to the core.

[Altera_Forum]

Hi Arg, 

   --- Quote Start ---  10M02DCV36I7G is not a single supply device - and as your schematic shows, there is no VCC_ONE because this part needs separate supplies for core (1.2V) analogue (2.5V) and I/O (could be the same as one of the other supplies, but you appear to need 3V3).  I don't think that the single-supply parts are available in the WLCSP36 package size. [...] But probably in this case you've simply blown up the device by applying 3V3 to the core.   --- Quote End ---    You were right the WLCSP36 package is only available in dual supply with maximum 1.5V on the core. So we blew up the core with the 3.3V. I have checked the user guide and according to the naming convention the dc after 10M02 means dual supply. Our previous design was working since we use a 10M02sc with single supply capabilities.  As I cannot dissociate VCCIO and VCC on the PCBs, I made some trials on a fresh board powered by a single 1.2V supply and there is no more short-circuit. My USB blaster isn't compatible with 1.2V I/O levels so I need to add a level shifter circuit prior to test if JTAG is properly operating.  Thanks for your help. Best regards, Tim

[Altera_Forum]

--- Quote Start ---    As I cannot dissociate VCCIO and VCC on the PCBs, I made some trials on a fresh board powered by a single 1.2V supply and there is no more short-circuit. My USB blaster isn't compatible with 1.2V I/O levels so I need to add a level shifter circuit prior to test if JTAG is properly operating.   --- Quote End ---    I think you are still in trouble if you can't split VCCA from the other supplies. 1.2V is OK as VccIO, but VccA is still required at 2.5V.  I've long been irritated by the need to provide the 2.5V nominally for the PLLs even if I'm not using the PLL; I haven't tried it on MAX10, but certainly on older Altera families (Cyclone 2/3) it won't boot without the 2.5V being at the right voltage, even if you don't use the PLL after boot. Possibly the 2.5V also runs the internal oscillator that clocks the boot process.

[Altera_Forum]

--- Quote Start ---  I think you are still in trouble if you can't split VCCA from the other supplies. 1.2V is OK as VccIO, but VccA is still required at 2.5V.  I've long been irritated by the need to provide the 2.5V nominally for the PLLs even if I'm not using the PLL; I haven't tried it on MAX10, but certainly on older Altera families (Cyclone 2/3) it won't boot without the 2.5V being at the right voltage, even if you don't use the PLL after boot. Possibly the 2.5V also runs the internal oscillator that clocks the boot process.  --- Quote End ---    Yes re-read the documentation and it is mentioned that VccA is monitored by the por circuit (https://www.altera.com/content/dam/altera-www/global/en_us/pdfs/literature/hb/max-10/ug_m10_pwr.pdf) then the MAX10 will not boot.   We choosed the WLCSP36 for its ultra small size, but in our configuration (with 3.3V IO needed) it requires 3 different supplies. We should probably consider to move to a bigger package with single supply. Maybe the M153 is a good trade-offwhich is only 8x8 mm.  Best regards. Tim