I am a big advocate of anti-brownout features.  The simple circuit =
described below, i.e. PNP to resistor divider, no Zener required, has =
worked for me in a number of low cost apps, hooked into the 12C508 MCLR =
line on a PIC.  It takes 3 parts each costing a penny or less, and has =
survived all the tests I can throw at it.  The knee is soft, but it =
still seems to work OK. =20

I would not use brownout on a battery powered toy where failure is =
unlikely to cause anything other than a temper tantrum.  =20

Other than that, I would use a brownout circuit on anything.  I can't =
think of a situation where it is not justified even on very cost =
sensitive designs.  Most newer MCHIP parts have internal brownouts so =
this is less and less of a problem. =20

-- Lawrence Lile, P.E.
Electrical and Electronic Solutions
Project Solutions Companies
www.projsolco.com

> -----Original Message-----
> From: Russell McMahon [mailto:apptech@PARADISE.NET.NZ]
> Sent: Wednesday, August 18, 2004 2:05 AM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: [PIC:] ATtiny11 - Reset & Brownout
>=20
> Summary:
>=20
> Discusses provision of low cost antibrownout and reset for processors =
with
> no external reset pin and inadequate on chip reset/anti-brownout.
>=20
> How do other people plan to do this?
>=20
> If they don't, why not?
>=20
> __________________________
>=20
> I put this under [AVR:] a day or so ago here and on tiny 11 list and =
am
> amazed that nobody responded. The problem is wider than for AVR's so =
I'll
> revise it and post again here and see what happens.
>=20
> Bottom end processors with no on chip antibrownout circuitry and the
> ability
> to use the reset line as an I/O line instead are deficient in the =
reset
> area
> when used with the reset pin disabled.
>=20
> A good example is the AVR ATtiny11, but there are PIC and other =
examples.
> The ATtiny11 does not have internal anti-brownout protection.
> If an external reset generator is used it can be as capable as one =
wishes
> for brownout etc. However, when the external reset pin is used instead
> as an I/O pin it seems that the processor is at great risk of running =
off
> and hiding if there is a power supply glitch or eg battery voltage =
falls
> below operating point etc.
>=20
> The Attiny11 has an on chip "power on reset" (POR) circuit but this =
resets
> at typically 0.8 of Vcc nominal and as low as 0.4 of Vcc nominal.
> As the tiny11 has Vcc min =3D 4.5v and nominally 5v I assume that this =
means
> POR on a falling supply resets at 2 to 4 volts. Even the upper level =
=3D 4
> volts is BELOW the guaranteed 4.5v.
>=20
> The tiny11 is priced such that even cheap reset generators tend to be =
more
> expensive than the processor.
>=20
> Rather than "waste" a pin where every pin can be critical you need a =
cheap
> reset/brownout circuit
> that actually pulses the power supply line to ensure internal reset =
and
> restart. A pretty drastic solution.
>=20
> How are others proposing to address this?
>=20
> Consider 2 simple (and cheap) circuits. Yet to try, but should work.
> Aim is to maximise effectiveness while not destroying the tiny11's
> amazingly
> low price advantage.
>=20
> The simplest is extremely standard.
> PNP.
> Emitter to Vcc.
> Base to centre of resistive divider from vcc to ground.
> Vdd to processor from collector.
> Set resistive divider to turn transistor on only whenVcc is above eg =
4v5.
> Add a zener in base to ground resistor to sharpen response.
>=20
> Result:    Soft knee.
> No guaranteed response to short low magnitude glitch.
>=20
> 2nd cct
> Add another pnp
> Emitter to Vcc
> Resistor, base to Vcc
> Collector to base of first transistor
> Capacitor from 2nd base to 1st collector.
>=20
> 2nd transistor adds regenerative action that will (with design) turn =
1st
> transistor completely off for a period controlled by capacitor.
> Redrawing this circuit reveals it to (almost) be a cross coupled
> monostable
> multivibrator.
>=20
> Extras:
> Add a resistor in series with cap to control cap discharge time.
> Add a resistor from 2nd base to ground to increase sensitivity to =
glitches
> (bias near turn on).
> etc
>=20
> Starts getting high parts count.
> All low cost parts.
> Liable to be far cheaper than an IC solution.
>=20
> Without something like this the tiny11, and similarly equipped =
processors,
> are able to get lost on low supply or supply noise.
>=20
>=20
>=20
>         Russell McMahon
>=20
> --
> http://www.piclist.com hint: To leave the PICList
> mailto:piclist-unsubscribe-request@mitvma.mit.edu
>=20
> ---
> Incoming mail is certified Virus Free.
> Checked by AVG anti-virus system (http://www.grisoft.com).
> Version: 6.0.735 / Virus Database: 489 - Release Date: 8/6/2004
>=20

---
Outgoing mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.735 / Virus Database: 489 - Release Date: 8/6/2004
=20

--
http://www.piclist.com hint: To leave the PICList
mailto:piclist-unsubscribe-request@mitvma.mit.edu