# CMOS logic level range

Discussion in 'Electronic Design' started by Robert Harris, Mar 8, 2007.

1. ### Robert HarrisGuest

I am designing a circuit using all CMOS 4000 series IC's.

What is the minimum and maximum voltage that will work as logic level
"1"?

IOW how far off 5V can it go either way?

Thank you,

Rob Harris

2. ### D from BCGuest

Isn't this a datasheet thing?
D from BC

3. ### DotinhoGuest

level 1 is 4.5V to Vcc 0.5V, okay?

4. ### Jim ThompsonGuest

The "official" specification is usually...

"0" <= 0.3*VDD

"1" >= 0.7*VDD

...Jim Thompson

5. ### JoergGuest

The not-so-official specification is usually...

"0" <= 0.3*VDD

"1" >= 0.7*VDD

"linear" = 0.5*VDD

))

6. ### Jim ThompsonGuest

Sane people don't do "linear" with off-the-shelf CMOS ;-)

...Jim Thompson

7. ### JoergGuest

Good. So if someone wants to sue me I could then claim immunity based on
the grounds of insanity ;-)

8. ### John LarkinGuest

That will certainly work, but if any logic level is very far off the
rails (a volt maybe? less?) the chips will draw excess current and run

John

9. ### Rich GriseGuest

So, I guess the HCU04 isn't "off the shelf"? ;-)

and, BTW, I had thought that 0 <= 1/3 Vdd (.33) and 1 >= 2/3 Vdd. (.67)

Thanks,
Rich

10. ### JoergGuest

That doesn't include taxes ;-)

11. ### jasenGuest

typically cmos high is 2/3 of VCC and above, read a few datasheets.
with a 15V supply 5V would be low

Bye.
Jasen

12. ### LionelGuest

Depends on the logic family. Traditionally (eg the old CD4000 series
chips), you can power them from anywhere from 3V to 15V, & a logic '1'
is anything above 1/2 the supply voltage, '0' is anything below that.
For newer CMOS logic families, they specified for all sorts of weird
voltage ranges. You can find out for sure from the data sheet.

But as a general rule of thumb for any CMOS logic chip powered by 5V,
you can assume that any signal above 2.5V will be treated as a '1', &
that anything below about 1V will be considered a '0'.

Again, if you're doing anything critical, you need to look up the data
sheet.

13. ### Anthony FremontGuest

I believe 2.5V is wayyyyy too low for a Vcc of 5V unless you're warming up
the parts. ;-) I'd keep it at .7Vcc (.8 *Vcc for Schmitt trigger inputs)
to be safe. IOW, 3.5V is the minimum.

14. ### LionelGuest

You've obviously grown up with TTL compatible CMOS parts. Go check
the 4000 series spec's.

15. ### Jim ThompsonGuest

Lionel, YOU need to re-read the data sheets.

...Jim Thompson

16. ### Anthony FremontGuest

You must think they all are. ;-) I'm not trying to be rude, but maybe you
should look again. Vih is usually at least 3.5V on CMOS, anything below
that and you are gambling. 2.5V is waaaayyyyyyy toooooooo looooooowwww for
CMOS inputs. ;-) PIC chips want 4V minimun on 5V Vcc (.8 * Vcc).

17. ### Anthony FremontGuest

I will add that for the PICs, the pins that have TTL input buffers (most of
them) only require 2.0V (4.5V < Vcc < 5.5V). The official formula is (1.25
* Vcc) + .8V, but they make a special case for 5V operation. But I believe
this is a special case and not consistant with normal CMOS specifications.
The Schmitt trigger inputs want .8 * Vcc.

18. ### LionelGuest

I do? I'll admit that it's been at least 15 years since I last looked
at a 4000 series data sheet...

<goes & checks>

Well, according to Motorola
(<http://www.onsemi.com/pub/Collateral/MC14001B-D.PDF>),
at a Vcc of 5V, (@25c), a logic '0' = 0-2.25V, & a logic '1' = 2.75V -
5V. That'd make the swicthing point 50% of Vcc, wouldn't it?
Seems to me that I was dead on the money.

Maybe your memory's getting a little rusty too, eh Jim? ;^)

19. ### Anthony FremontGuest

That's the Typical spec and certainly wouldn't be something to recommend as
a rule of thumb IMO. I'm going to stick with the value in the column next
to that one that says (Min 3.5V).

20. ### LionelGuest

Sure, but they're designed to switch at Vcc/2.