07-01-2016 08:19 AM
While probing on a board during power-up configuration (Spartan-6 FPGA configured by XCF32P using Master SelectMap),
the INIT_B signal is seen to transition from +3.3V to +1.8V at the end of configuration (when CONFIG_DONE goes from 0V to +3.3V). It was expected that INIT_B would stay at +3.3V since it has a 4.75K pullup on it per Xilinx recommendations. This is on a
board that has had problems with internal shorts, so I'm trying to rule that out for management. I believe what is happening is
a weak pulldown is engaged on the INIT_B at end of configuration (answer record AR# 35002 seems to hint at this), but was hoping someone could confirm this. Per my calculations, the weak pulldown would need to be about 2.5K, but I couldn't find spec for this in Spartan-6 datasheet.
The following related forum seems to ask same kind of question, but I didn't see a conclusive answer in it:
Spartan 6 (SC6SLX16-3) INIT_B goes to 1.9V during serial slave mode config
07-01-2016 08:52 AM
If your design doesn't use the INIT_B pin after configuration it takes on the same defaults as other Unused IO. Unless you change the bitgen settings, the default is weak pull-down. ds162 Table 4 lists pulldown current for a 3.3V bank as 200 to 550 uA. So with a 4.7K pullup to 3.3V, the output voltage should go to somewhere between 2.36V (@ 200uA) and 0.715V (@550uA). So basically what you're seeing is normal. If you don't want this behavior you can add the INIT_B pin to your design so it doesn't get pulled down, or you can change the bitgen settings to apply weak pullup istead of pulldown to unused IO.