Spiro:

My MacAfee SW continues to alert me when you send e-mail to the PIC list.
It says:
---------------------------
McAfee SiteAdvisor Warning

	This e-mail message contains potentially unsafe links to these
sites:
 	trexon.com
---------------------------
FYI....

Rich
=20

-----Original Message-----
From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf Of
Spehro Pefhany
Sent: Friday, December 10, 2010 12:07 PM
To: Microcontroller discussion list - Public.
Subject: Re: [EE]: My first circuit using comparators and seeking feedback.

At 11:07 AM 10/12/2010, you wrote:
> > Sure. A few brief comments...
> >
> >         if the engine is cranked for 30 seconds then the voltage
> >         might be low enough during that period to kill the MOSFET
>
>I was not aware of this.  To be sure I understand, if the supply
>voltage across the MOSFET in series with the load drops then there is
>a risk of destroying the MOSFET. I am actually using a MTP3055E.

No, if the gate-to-source voltage drops much below 10V (but more than
a small fraction of a volt). Your original circuit does not
necessarily do this, but it has other problems (slow turn-off),
and the cure will likely lead you down this path.

> > 2]      Your time delay is 100% controlled by leakage currents.. 30
seconds
> >         implies about 150nA. Stability is dubious under hash conditions=
..
>
>This has now been rectified by using an additional comparator which
>has the non-inverting section fed by a cap in parallel with a 1Mohm
>resistor.
>
> >
> > 3]      A PIC circuit could be lower power.. your zener reference alone
>
>Yes, the reference is not good, but I figured it did not need to be
>just repeatable which I suspect it is.  Again for me the commonly used
>LM2931 is a long way away.
>
> >
> > 4]  I'd tend to divide down the reference rather
> >         than the input..
>
>I have seen this approach but wasn't intuitive.  To me it was easy to
>think of fixing a reference then setting thresholds relative to it.
>
> >
> > 5]      Vehicle electrical systems have transients that kill
electronics.
> >         Some protection against something like +/-60V spikes would be a
> >         good idea.
>
>I had considered this but not completely sure how to resolve it.  I
>was considering an inductor in line with the 12V in, and a cap across
>the 12volts in.  I can get 30V zeners and could use 2 in series to get
>the 60V but are very low power.

I'd wager that 60V will kill the LM339. You should endeavor to
limit the transients to the ratings of the parts, 36V in the case
of an LM339.

A 22V 3,000 W TVS (basically a very solid zener diode) will have
a clamping voltage of <36V while clamping 85A. That's not bad,
but it is not guaranteed to survive the double battery test. A
24V 3,000 W TVS will clamp to 38.9V at 77A, above the rating of
the LM339, but will survive a double battery.

Depending on the circuit, a tighter constraint may be the Vgs(max)
of your MOSFET, typically between 10V and 20V maximum, depending
on the type.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the
reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.co=
m
Embedded software/hardware/analog  Info for designers:  http://www.speff.co=
m



--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.