[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 3.3V supply



Andrew,

Good points about the JEDEC spec.  I am thinking of large scale devices,
such as a Marvel/Galileo PCI controller chip, and a PMC-Sierra RM5261
MIPS processor chip both using the Narrow Range (which is actually +/-
4.5% tolerance).

I haven't looked at other PCI controller chips-- maybe Marvel/Galileo is
a odd-ball.  

Comments from other PCI controller chip users?

regards,

Ted


"Ingraham, Andrew" wrote:
> 
> > If anyone knows why the 9.1% tolerance was specified in the 2.2 spec,
> > please let me know.  I find it hard to believe that the spec
> > developers
> > didn't do their homework at the time and discover 5% tolerance devices
> > would be common.
> 
> "At the time" would be circa 1990.
> 
> The JEDEC standard for 3.3V integrated circuits (JESD8-A, 1994)
> specifies a VDD range of 3.0 to 3.6V, with no mention of anything
> narrower.
> 
> Perhaps the question should be, why did IC manufacturers later start to
> deviate from existing standards (not just PCI but JEDEC) and specify a
> narrower range?
> 
> The newer JESD8-B now includes 3.15 to 3.45V as its "Narrow Range," with
> 3.0 - 3.6V as the "Normal Range" and 2.7 - 3.6V as "Extended Range."
> 
> Just how "common" are devices with a 5% supply voltage tolerance?  I
> cracked open a T.I. data book of 3.3V parts and they take +/-0.3V just
> fine.  I checked some others and they, too, specify 3.0 - 3.6V, if not
> wider.  At what point did the exception become the rule?
> 
> Regards,
> Andy