EmbeddedRelated.com
Forums

Vdd (core voltage) decoupling requirements

Started by Oliver Betz November 8, 2006
Hello Jonathan,

please use a more readable quoting style, more info at
http://web.presby.edu/~nnqadmin/nnq/nquote.html or
http://www.netmeister.org/news/learn2quote2.html

Jonathan Masters wrote:

[...]

> At this rate has anyone seen failures as a consequence of not meeting
> this requirement? If so, what sort of symptoms have been seen?

We are using 2*100nF on a DT128 design, hundreds (from different mask
revisions) are in use, no failure.

The two C32 I had tested heavily in the last days worked fine even
with 22nF on Vdd. Another C32 is running with 100nF, and a fourth
with 220nF. But that's no significant quantity.

Regarding the symptoms: I guess you will get random errors, so it
will be hard to assign.

Oliver
--
Oliver Betz, Muenchen
--- In 6..., "Oliver Betz" wrote:
>
> Hello Jonathan,
>
> please use a more readable quoting style, more info at
> http://web.presby.edu/~nnqadmin/nnq/nquote.html or
> http://www.netmeister.org/news/learn2quote2.html
>
> Jonathan Masters wrote:

Concerning quoting, the absolute worst is relying on Microsoft's
Hotmail or Outlook to quote in a way which only works for top-posting
(which is also not good). I ditched Hotmail, partly because I could
not find any way to change the quoting scheme to precede lines with
'>'. Anybody know how to do that in Hotmail and in MS Outlook?
> The two C32 I had tested heavily in the last days worked fine even
> with 22nF on Vdd. Another C32 is running with 100nF, and a fourth
> with 220nF. But that's no significant quantity.
>
> Regarding the symptoms: I guess you will get random errors, so it
> will be hard to assign.

At first I mentioned my 0.1uF because I thought it would be
significantly more than 400nF. I guess I was thinking of pF (400e-12).

I wouldn't think the two 0.1uF have been trouble for me either. I have
had crashes (CPU runs astray), but think it is from noise such as EST,
often through Gnd. I can shock myself on the grounded bolt on the 7805
heatsink, and my prototype crashes. With all the possible causes
building up so fast, it's hard to troubleshoot.