7 Oct
2007
7 Oct
'07
11:33 p.m.
No matter that it's a [bad] idea to [apply] logic
signals to an
unpowered chip.
I recently built a ROM reader with a parallel-port interface.
But it had one bizarre problem: the last byte of the ROM always read as
FF. I finally tracked it down. I'd connected Vcc, and I had tied the
enable signals on the ROM together, to the ground pin - but hadn't
connected the enable-and-grounds anywhere. As long as at least one
input was low, it powered the chip through the protection diodes (thank
the TTL, and TTL-alike, output current-sinking capacity when low),
which was why the rest of the chip read correctly - but as soon as all
inputs were high, the outputs perforce all showed high too. (The time
between address line setting and data line reading was in the
microseconds, plenty of time for ROM power-up.)
/~\ The ASCII der Mouse
\ / Ribbon Campaign
X Against HTML mouse at rodents.montreal.qc.ca
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B