Re: Windows XP startup question

krw wrote:
On Sun, 27 Dec 2009 13:05:46 -0800, Joerg <invalid@xxxxxxxxxxxxxxx>
wrote:

Jim Thompson wrote:
On Sun, 27 Dec 2009 12:18:04 -0800, Joerg <invalid@xxxxxxxxxxxxxxx>
wrote:

Jim Thompson wrote:
On Sun, 27 Dec 2009 10:52:30 -0800, Joerg <invalid@xxxxxxxxxxxxxxx>
wrote:

Jim Thompson wrote:
On Fri, 25 Dec 2009 12:50:16 -0800, Joerg <invalid@xxxxxxxxxxxxxxx>
wrote:

Jim Thompson wrote:
On Fri, 25 Dec 2009 12:16:32 -0800, Joerg <invalid@xxxxxxxxxxxxxxx>
wrote:

[snip]
Personally, the most I'd invest in a case like this would be a telephone DTMF interface. Call the cabin, push star-something, wait for confirmation tone, hang up, pack skis, head up there.
I haven't crashed a computer in so long I've forgotten how ;-)

WHERE is it that you surf that you keep getting crashes?

I don't, but others do. Mine sometimes crashes on Acrobat but this is not very surprising, it was IMHO never a particularly robust software.
Full Acrobat or the Reader? What version?

Reader, 8 and 9. They get worse the younger they are, just like a lot of other sofware does.
I stopped at Acrobat v7... primarily used for reading, creating from
scanner. I use v4 for "printing to", and PDF's that need links and/or
forms (I like to make "hierarchical" PDF schematics for design
reviews... works better than PowerPoint).

Come to think of it, I've _never_ had Reader on any of my machines.

I don't need the full version because I have a virtual PDF printer so can get PDF output from any program. Plus the scanner already produces PDF (if you set it to PDF format).

Anyhow, I would not want to trust a Windows-based PC with any kind of home automation. Just like I won't trust X10 with that.

[snip]

Yep. I'd use your DTMF interface... battery powered, with a
charger... discrete chips and components... latchable relays... beep,
beep, beep reporting that mission was accomplished... NO uP ;-)

It can be a uC though. While it took that industry seemingly forever to figure out how a good BOR works now there are some that are reliable enough. Some, not all. Texas has a nice app note on how to do that, with the code and all.
Can't you crash a uP without a brown-out/power-drop-out?

That entirely depends on the code quality :-)
I know. You're perfect ;-) ...

Nope, only one guy was and was killed at age 33.


... I like digital solutions simple enough
that even I, an old Analog guy, can understand, and proof every
possible combination.

Even if you don't, what's the big deal of generating a regular pulse that resets the watchdog? uC code hangs -> watchdog triggers -> big master reset happens.

What happens with flaky power? You need something to wake up the dog
reliably.


A good BOR. Which is where I spend a lot of diagnostic review time because I don't trust some of the commercial solutions.


I can't even count the apps anymore where I found that the watchdog had been disabled. Maybe because it was "inconvenient".

A separate circuit and NMI.

I recently designed a robust BOR (with an additional power monitor to
cover another chip)... only took ~90 active devices ;-)

I built my first reliable BOR circuit when I was 18 or 19, out of salvaged transistors. Because it was needed in a high voltage design where uncontrolled brown-out could cause tube grid voltages to become suboptimal and lead to a serious *KABLOUIE*. That's why I never understood why large swaths of the electronics industry could not get that licked for decades.
Part of my chip design procedure includes testing for VERY soft power
supply rise and fall times... looking for any possibility of latching
before every thing reaches normal quiescent bias. Particularly
critical with split supply chips.... but those I usually design so
that minus supply must pass P/BOR before stuff powered from the plus
supply is allowed to bias up (avoids crazy substrate behaviors).

I do the same with board level stuff. "Oh, it's going to be fine." ... "Should we quickly test it anyhow? Only takes a minute." ... "Sure." ... crank, crank, crank ... *PHUT*

[...]

Or quickly cycle power to confuse the POR timing.


No, it was immune, it would just go into the big freeze for the max power-on delay time.

--
Regards, Joerg

http://www.analogconsultants.com/

"gmail" domain blocked because of excessive spam.
Use another domain or send PM.
.

Relevant Pages

  • Re: Windows XP startup question
    ... DTMF interface. ... I don't need the full version because I have a virtual PDF printer so ... Because it was needed in a high voltage design ... Part of my chip design procedure includes testing for VERY soft power ...
    (sci.electronics.design)
  • Re: Windows XP startup question
    ... DTMF interface. ... I don't need the full version because I have a virtual PDF printer so ... Because it was needed in a high voltage design ... crank, crank, crank ... ...
    (sci.electronics.design)
  • Re: Windows XP startup question
    ... DTMF interface. ... Call the cabin, push star-something, wait for confirmation tone, hang up, pack skis, head up there. ... I don't need the full version because I have a virtual PDF printer so can get PDF output from any program. ... crank, crank, crank ... ...
    (sci.electronics.design)
  • Re: Windows XP startup question
    ... DTMF interface. ... I don't need the full version because I have a virtual PDF printer so ... I'd use your DTMF interface... ... Because it was needed in a high voltage design ...
    (sci.electronics.design)
  • Re: PowerPoint to CMYK PDF for Press? How?
    ... Sometimes the content is more important than the design or aesthetics. ... just providing a valid PDF file is sufficient. ... A high-res, first generation jpeg ... I think it depends on what you are printing. ...
    (comp.publish.prepress)