On Fri, 03 Sep 2004 10:05:33 -0700, Markus Meng wrote:

> Hi all,
>
> I just wonder if someone else did make the same measurements.
>
> We do have a system with a 9572XL having inputs with an external 47KOhm
> resistor. The design has done by another person. I prefer pull-up
> resistors in the 3.3K...10K max value ranges.
>
> We did have strange behavior until I did measure the voltage on these
> inputs pins. After configuration through JTAG we do have a clean 3.3VDC
> pulled up input value ...
>
> Bringing one of these inputs to GND and releasing it again, shows with
> the oscilloscope, that the voltage on these pins doese rise until
> 0.9Volt but not higher. Thought the external pull-up resistor should
> bring these inputs - only - in the design back to 3.3VDC again. No way,
> once these pins have been brought to GND, the release of it give us a
> voltage rise up to 0.9VDC but not more.
>
> This explains a lot, why even the simplest logic did give us a real
> blues ...
>
> However reducing to the external pull-up resistors to below 10K, we
> found everything is working as expected. Checking the datasheet a see a
> max input current for the inputs of +-10 uA. Multiplying this value with
> 47K gives us 0.47V max ...
>
> It seems that this max value is somehow not the max value .... (On these
> inputs, there is no additional cap ...)
>
> Any idea?
>
> Markus

XC9572XL have bus hold circuitry on inputs. Unless you overpower the bus
hold circuit with your pullup resistor, you will have funny I/O levels.

Not sure but maybe you can program input for pullups instead of buss
hold...


Peter Wallace