Re: [uCsimm] Are the uCsimm I/O pins 5V Tolerant?

From: John Ananny (
Date: Fri Feb 04 2000 - 13:57:49 EST

A few more thoughts.

A simple, conservative interface between a 5V LS buffer output and the 3.3V
HC EZ input is a small inline resistor, which effectively raises the output
impedance of the LS buffer. The only reason not to hook the two pins
together directly is in response to Tom's concern that the loaded
output-high voltage of the LS device might be a bit above 3.3V. Adding a
small resistor (100 to 200 ohms, say) gives the extra 0.2V or so somewhere
else to appear. The exact resistor size depends on the input-high current
drawn by the processor. Unfortunately, given that the data sheet for the
processor completely bites, I can't size the resistor except by throwing
darts. Or else by hooking it up and measuring. :-)

About the note on TTL/CMOS levels copied below... Don't confuse levels
with voltages. A 5V LS (TTL) device has an input-high threshold of 2V or
thereabouts, while a 5V HC (CMOS) device typically has an input-high
threshold of about 4V. (This, by the way, is the reason behind my earlier
mail describing how 3.3V HC can drive 5V LS directly; since HC outputs come
almost to the rail, a 3.3V HC device more than meets the 2.5V or so safely
needed to guarantee a valid TTL input-high.) These switching thresholds
are the TTL/CMOS levels referred to in the note below. Nothing in the note
suggests actual voltages -- think of levels, very roughly speaking, as
defining the proportion of the supply voltage needed to guarantee a valid
logic-low or logic-high.

Anyway, without a better data sheet, this is only so much
speculation. Prototype with an LS buffer plus 100 ohm output resistor, and
you'll be just fine. Volume designers need to find a cheaper solution.


> "Note: All pins, except EXTAL, support TTL levels. When EXTAL is used as
an input
> clock, it needs a CMOS level. To ensure proper low-power operation, all
> should be driven to CMOS level. More power is consumed when you use a TTL
> level to drive those inputs."
> I think this is the info you require. So 5V is OK, but will cause system
> to consume more power when going into low power modes.

John Ananny
Interval Research Corporation (650) 842-6039
This message resent by the list server

This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:34 EST