'What exactly are pullups? [SOT]'
Clyde Smith-Stubbs wrote:
> Pullups are more common that pulldowns mainly because in The Old Days
> when TTL logic ruled the input to a TTL device naturally floated at a
> high level, and required only a small pullup for noise tolerance, but
> a pulldown had to sink quite a large current to get the input low.
But that no longer applies. The *other* and probably major reason for
pullups rather than pulldowns, is something that hails from the days of
the oldest "computers", that is controllers; the terrific machines that
were - the telephone exchanges.
Where signals (data) are conveyed any particular distance, and this
may vary from a few inches on a PCB to many metres along cables, the
consequences of a short circuit (ground) need to be taken into account.
Where the load is returned to ground, a short on the control line causes
an overload on the power source and switching device.
Where however the load is returned to the supply, a short on the
control line causes spurious activation of the load, but generally no
overload and no damage (smoke).
In addition, open-collector (or relay) switching is frequently used to
interface loads powered by different supply voltages, for example a 12V
relay controlled by an O/C buffer from a PIC or similar device operating
from a 5V logic supply.
This is far easier if switches return to ground and loads to their
respective supply. Indeed, with isolation diodes, a single O/C output
can control multiple loads such as relays from different supplies, or a
relay and a LED indicator.
It is somewhat more likely that this was the design concept underlying
the design of TTL logic, than merely the ease of making NPN transistors
(though I will concede the latter to avoid an argument).
More... (looser matching)
- Last day of these posts
- In 1999
, 2000 only
- New search...