Hi
I've been following all this programming thread with some interest.
I have a lot of designs that operate at 3V but it hurts none of my
components to let the programmer take the supply to 5V.
I have to operate at 3V for of a couple of reasons.
1)In some configurations I "steal" power from the antenna voltage GPS
receivers send to their antenna so 3V consumes less power 2)some receivers
though still quite rare only supply 3V to start with.
(Actually it is always just a hair above 3V because the antenna preamplifier
operates at 3V then you need to send a little more to over come the I^2R
losses so to make a long story short an LDO reg will give you 3V no
problemo)
  
Now I have a future product where the firmware will be wildly complex but it
will not "steal" power.....if it did it would quickly drain all available
power from all but the most high end/robust GPS receivers. 
To say that I'm going to nail the code on the first release or not want to
change this or that for some of my customers in the first few revisions is
wish full thinking at best and stoopid at worst.
My quantities are fairly low and testing will be very complex.
So I have no problem with adding the header for the ICSP but I can't expect
my customers in the field to have handy dandy ICSP or to open the box to
connect to it plus when these are fielded they are some times
installed/mounted in the rafters with a man lift and I use a blue tooth
serial link to control the thing.

I wanted to leave my basic designs at 3V if I could because the RF portion
works fine there no need to reinvent the wheel etc. 
So is the lesson I need to take away from all this is that in order to
realistically be able to alter my code in the field via a serial port I will
need to operate the PIC at 5V???

(Right now I'm using a 18F6520 or it I might go larger if it turns out I
need more I/O or use a second device.


I'd love to hear any of the groups thoughts and advice....thanks in advance
for any replies.

Phillip
Things should be as simple as possible but no simpler
 
 
 
Phillip Coiner
CTO, GPS Source, Inc.
 
 
Your source for quality GNSS Networking Solutions and Design Services, Now!
-----Original Message-----
From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf Of
William Chops Westfield
Sent: Wednesday, April 05, 2006 8:15 AM
To: Microcontroller discussion list - Public.
Subject: Re: [PIC]: Microchip offer chip programming service


On Apr 5, 2006, at 6:28 AM, Xiaofan Chen wrote:

> In our experience, pre-programming is in generall cheaper than
> ICSP for almost all quantities (a few 100 to 150kpcs per year)
> in the typical manufacturing setup. But this is calculated in
> Singapore where labor cost is not low at all (about half of USA).
> If it is in China maybe Olin's calculation is correct.
>
Olin is doing ICSP as part of a "test" step that is happening
anyway; if you don't have as significant a test setup as he
does, or don't have it set up to do ICSP in the same fixture,
your numbers will come out different...

I can see Olin's point, but I think an awful lot of products
are tested without "fixtures" of any kind; someone turns it
on and confirms that the power-on diagnostics run and the
expected power-on behavior occurs, and it ships...  I can
see it depending a lot on the electrical complexity of the
product; something like the proprog has a lot of not-easily
visible things you can measure.  A more consumer-oriented
product might not...

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

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