Resistors and Pulling down or up.

droffset · Posted: Sun Mar 14, 2010 5:21 pm Post subject: Resistors and Pulling down or up.

Musing about the next system I want to build and came up with a query for you folks:

I've gotten into the habit of using pull-down resistors on CMOS inputs because it was suggested to me. But why is a resistor necessary, can one not just use a jumper to connect to ground?

I see people mostly use 100k, is that an arbitrary value or is anything out of the parts bin usable?

Thanks

droffset · Posted: Sun Mar 14, 2010 6:58 pm Post subject:

Still want to hear from people, but couldn't resist the Goog.
http://www.seattlerobotics.org/encoder/mar97/basics.html
http://en.wikipedia.org/wiki/Pull-up_resistor

Psyingo · Joined: Jun 11, 2009 Posts: 248 Location: Canada

it is my understanding that the pull up/down is used when there is another signal that is coming in, so that when the signal isnt there, it will default. like when you are patching modules, leaving the input unconnected would make it float.

whether or not you pull up or down is determined by whatever you want the input to default to.

if the input is going to permanently remain high or low, tying it to either supply or ground is fine.

Uncle Krunkus · Posted: Mon Mar 15, 2010 1:07 am Post subject:

It may not be much of an issue when being pulled down (sinking) but when an input is being pulled up the resistor stops the input from sourcing significant amounts of current.

blue hell

electri-fire

electri-fire · Posted: Mon Mar 15, 2010 6:33 am Post subject:

electri-fire · Posted: Mon Mar 15, 2010 6:42 am Post subject:

quote: "Some disadvantages of pull-up resistors are the extra power consumed when current is drawn through the resistor, and the reduced speed of a pull-up compared to an active current source."

Ermm... now it gets complicated. This would mean: a higher resistor limits current draw.

Ah well, as we're using mostly audio frequencies speed would be less of an issue compared with normally used clockspeeds in digital circuits.

blue hell · Posted: Mon Mar 15, 2010 6:45 am Post subject:

Uncle Krunkus · Posted: Tue Mar 16, 2010 3:24 am Post subject:

Yeah, definitely listen to Jan on this one.
CMOS = No resistor needed.
I was just floating something I remember hearing, and Jan is right, it was in reference to TTL. Sorry to have misled anyone.