On Sat, Jun 12, 2004 at 08:59:05AM +0100, Peter Moreton wrote:
> Wouter,
>
> Do you have any ideas for a resistor value that is adequate to pull LVP low,
> but not overpower the RB5 weak pull up? - I'm guessing at about 100k, but it
> would be nice to use a known good value.
>
> (Much as I would like, I cannot use the WISP programmer, since my design
> needs to have an on-board programming facility)

Ah! That helpful to know. So the board will be programmed in socket. Well
now we're back to the potential programming mode glitch. Some suggestions:

1) Ensure that the board is populated with a chip that has had LVP programmed
off. No additional hardware needed then unless there is user error and the
chip is reprogrammed in socket with LVP config bit on.

2) Use a bootloader to program the part. Then you are no longer subject to the
PIC programming interface. Wouter's own ZPL would be a perfect fit for the
18F452 and would completely free the entire I/O interface of the chip from
an ICSP interface.

3) Invert the sense of RB5 and go with a strong pulldown and a switch to +5.
The problem I see with your approach is that noise is the trigger of the
problem. A weak weak pulldown may be just a subject to a noise glitch as no
pulldown at all. With the strong pulldown, no such problem can occur.

4) Have an actual switch to a pulldown for RB5.  Hey you have that switch
already right! So simply make it policy that while the part is being programmed
that whatever switch is connects RB5 to ground must be pushed. Unless the
exercise of the board is programming the board all the time, this should be
an extremely minor nuisance.

Hope these help,

BAJ
>
> Thanks, Peter Moreton
>
>
>
> > -----Original Message-----
> > From: pic microcontroller discussion list
> > [mailto:PICLIST@MITVMA.MIT.EDU] On Behalf Of Wouter van Ooijen
> > Sent: 12 June 2004 08:40
> > To: PICLIST@MITVMA.MIT.EDU
> > Subject: Re: [PIC] PIC18F452 LVP pin needs to be tied low?
> >
> >
> > > Chaps, can someone tell me if the LVP / RB5 pin on a
> > > PIC18F452 needs to be tied low?
> >
> > The problem is your word 'needs'. I don't know whether it is realy
> > needed, like I don't know whether that 100 nF cap near the PIC is
> > needed, or whether you realy need those caps at the crystal.
> > But I sure
> > know that I'd rather keep them all in place and not risk problems.
> >
> > > If I do tie RB5 low, then I lose a pushbutton, since I have a
> > > bank of 4
> > > switches on RB4,5,6,7 which use the internal weak pull-ups, with the
> > > switches pulling low through 1k when pressed.
> >
> > Use a pull-down that is weaker than the internal weak pull-ups. Or use
> > Wisp628 which has a dedicated LVP pull-down line :)
> >
> > Wouter van Ooijen
> >
> > -- -------------------------------------------
> > Van Ooijen Technische Informatica: www.voti.nl
> > consultancy, development, PICmicro products
> >
> > --
> > http://www.piclist.com hint: The PICList is archived three different
> > ways.  See http://www.piclist.com/#archives for details.
> >
>
> --
> http://www.piclist.com hint: The PICList is archived three different
> ways.  See http://www.piclist.com/#archives for details.

--
http://www.piclist.com hint: The PICList is archived three different
ways.  See http://www.piclist.com/#archives for details.