I guess you checked that the zener is the right way around?
More likely, the low voltage zener doesn't have a real nice well defined
zener point so it's actually starting to conduct at a much lower voltage. If
the gain of the BS328 is 100 then at 2V you'll only need about 0.5uA of base
current to turn the transistor on - and 0.6v across the 33k resistor, about
23uA total.
Try using forward biased junctions (e.g. LEDs) or just using the resistor
divider method if you don't need too much accuracy. 12K to V+ and 56k to
grnd might be close.

Richard
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Swichtec Power Systems Limited,   | Phone: (++64) 3 343-3314
39 Princess St, PO Box 11-188,      | Fax:   (++64) 3 348-8871
Christchurch, New Zealand.          | Email: rprosser@swichtec.co.nz
Visit us at http://www.swichtec.co.nz
+++++++++++++++++++++++++++++++++++++++++++++++++++++++



> -----Original Message-----
> From: John Duncan [mailto:j.duncan@BOM.GOV.AU]
> Sent: Wednesday, May 26, 1999 10:35 AM
> To: PICLIST@MITVMA.MIT.EDU
> Subject: Brownout cct
>
>
> Hello,
>
> I am building a project with a 16F84 and thought it would be
> a good idea to
> incorporate a brownout circuit to reset the part in the event of a
> brownout.  I want to use the circuit featured in the spec
> sheet for the
> 16C84 with a zener diode and a PNP transistor and a few
> resistors.  I have
> a lot of BC328 sitting about so I thought that I would use a
> few.  Trouble
> is that the circuit does not work for me.  Vdd is 5 volts I
> am using a 3.3V
> zener.  I thought that the transistor would trigger when Vdd
> got to 4V, but
> /MCLR does not go low till Vdd is approx 2V.  Has anyone had
> success with
> this circuit and point me in the right direction.
>
> Cheers
> JD
>