Searching \ for '[PIC]: Jtag and more' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/microchip/devices.htm?key=pic
Search entire site for: 'Jtag and more'.

Exact match. Not showing close matches.
PICList Thread
'[PIC]: Jtag and more'
2002\06\17@144557 by Andre Abelian

picon face
Hi to all,

After carefully studding PIC micros I think this is what
Microchip should do.

1. move from OTP to Flash on all MCUs and have OTP version available
incase.
2. add Jtag hardware in pic to have inexpensive debugging capability set
brake points etc.
3. Convert boot loader software in hardware that you do not have to
program it boot loading should be part of hardware
4. add Digital to analog conversion hardware.
5. have separate interrupt vectors not one for all
6. in mplab add option to generate code like after choosing the part
  "I2C" = on or off / serial port = on  9600 /   it will generate the
code for it instead of keep looking at the data sheet looking for
registers.
7. have 2 independent PWM not 2 PWM with one timer running.
8. have 2 USART
9. have SPI and I2C separate hardware instead SPI or I2C

I am sure I will see some of this in the future

Andre Abelian

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\17@144945 by Byron A Jeff

face picon face
On Mon, Jun 17, 2002 at 11:45:05AM -0700, Andre Abelian wrote:
{Quote hidden}

I'm sure that Microchip will do all of those things...

...and the chip will cost $30 each in quantity.

As with most things in life, you get what you pay for.

BAJ

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\17@154507 by Brendan Moran

picon face
Hi,
While much of what you mention make sense, I think that you are overlooking
some things.
----- Original Message -----
From: "Andre Abelian" <spam_OUTengelecTakeThisOuTspamEARTHLINK.NET>
To: <.....PICLISTKILLspamspam@spam@mitvma.mit.edu>
Sent: Monday, June 17, 2002 11:45 AM
Subject: [PIC]: Jtag and more


> Hi to all,
>
> After carefully studding PIC micros I think this is what
> Microchip should do.
>
> 1. move from OTP to Flash on all MCUs and have OTP version available
> incase.

As near as I can tell, they are currently doing this.

> 2. add Jtag hardware in pic to have inexpensive debugging capability set
> brake points etc.

This is not necessarily a good idea for a production component.  It's nice
for the hobbyist, and the developenr, but the end user couldn't care less if
it's there.  Therefore, what is likely, to keep costs low, is if they do do
something like this, it will be available as a development model.

> 3. Convert boot loader software in hardware that you do not have to
> program it boot loading should be part of hardware

Once again, the end user doesn't care.  It is not necessary.  Only SMD MCU's
ever really needed this, and with the advent of ICSP, even the SMD MCU's
don't really need a boot loader anymore.

> 4. add Digital to analog conversion hardware.

D/A hardware gets HOT.  And it sucks up power at an incredible rate.  For a
low power MCU, not exactly the best idea, which is, I expect, why they
haven't done it.  As far as I know D/A hardware on MCU's is pretty rare.

> 5. have separate interrupt vectors not one for all

That would definitely be a nice feature, and I think that they may be moving
towards it.  It looks to me like the '87x is designed to be forwards
compatible with a future MCU that has 4 interrupt source, considering that
it has the 3 empty blocks between the reset vector and the interrupt vector.

> 6. in mplab add option to generate code like after choosing the part
>    "I2C" = on or off / serial port = on  9600 /   it will generate the
> code for it instead of keep looking at the data sheet looking for
> registers.

You could implement that yourself with macros if you really want the
feature.

> 7. have 2 independent PWM not 2 PWM with one timer running.

Some MCU's have this.  They're more expensive than PICs.  Go figure.

> 8. have 2 USART

Once agian, some MCU's have this.  Price is the difference.

> 9. have SPI and I2C separate hardware instead SPI or I2C

I know of at least one person who is claiming that it won't be long before
MCU's come with an I/O FPGA so that you can configure whatever output
devices you want.  With that sort of thing, you could even change what I/O
peripherals you had from reset to reset.

> I am sure I will see some of this in the future
>
> Andre Abelian

Some, yes, but most, probably not.  The PIC series was designed, for the
most part, to be a very inexpensive line of devices.  They wanted to build
for the common hobbyist that couldn't afford the the shiny new Motorola MCU.
Most of what you want is available on the HC12.  It covers at least
1,2,3(sort of),7,8 maybe 5 as well.

I mention the HC12, because it sounds to me like you aren't really looking
for a PIC.  The HC912 (HC12 family) costs $20USD each in VOLUME like in the
1k region.  They don't like to sell to individuals for some reason, and the
MCU only comes in SMD format.

Thus the difference in peripherals and core features.  PICs have their
place, and so do most of the other available MCU's, right down to the 10-pin
3x3mm 8051 clone.  It just looks like you're not talking about PICs anymore.

Just to summarize the HC12's features:
Check here for a complete listing:
e-www.motorola.com/brdata/PDFDB/docs/M68HC12B.pdf
CORE:
68HC912B32 Features:
16-Bit CPU12
Upward Compatible with M68HC11 Instruction Set
Interrupt Stacking and Programmer's Model Identical to M68HC11
20-Bit ALU
Hardware multiplication and division
Instruction Queue
Enhanced Indexed Addressing
Fuzzy Logic Instructions

Multiplexed Bus
Single Chip or Expanded
16/16 Wide or 16/8 Narrow Modes

MEMORY:
Flash: 32kBytes
EEPROM: 768 bytes
RAM: 1kByte

PERIPHERALS:
Timer:
-8ch, 16bit
I/O:
-up to 63 pins (thats right sixty-three, no wonder I feel starved for I/O on
the PICs)
Serial:
-SCI
-SPI
-2UART
A/D:
8-CH 10-Bit
PWM:
-4-CH 8-Bit or 2 CH 16 Bit
CAN:
-2.0a
-2.0b

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\17@161803 by M. Adam Davis

flavicon
face
Since there are already chips out there that have most or all of these
features, you should include in your list why Microchip should expand
into that market, and in what ways these chips would be better than
what's currently available.

-Adam

Andre Abelian wrote:

{Quote hidden}

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\17@164047 by Olin Lathrop

face picon face
> That would definitely be a nice feature, and I think that they may be
moving
> towards it.  It looks to me like the '87x is designed to be forwards
> compatible with a future MCU that has 4 interrupt source, considering that
> it has the 3 empty blocks between the reset vector and the interrupt
vector.

No, that's so you can jump from the reset vector before the code hits the
interrupt vector.  You need at least 3 instructions for this, two to set
PCLATH properly, and one GOTO.  I use the 4th instruction to make sure
interrupts are off so that a GOTO 0 (with PCLATH set appropriately of
course) performs a software reset.

> Some, yes, but most, probably not.  The PIC series was designed, for the
> most part, to be a very inexpensive line of devices.  They wanted to build
> for the common hobbyist that couldn't afford the the shiny new Motorola
MCU.

I rather doubt Microchip cares much about the .01% of PICs sold that are
bought by hobbiests.  Hobbiests like PICs because they are targeted for low
cost applications, and because Microchip learned early on that providing
decent software tools for free helps to sell hardware.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\17@172315 by Brendan Moran

picon face
----- Original Message -----
From: "Olin Lathrop" <olin_piclistspamKILLspamEMBEDINC.COM>
To: <.....PICLISTKILLspamspam.....mitvma.mit.edu>
Sent: Monday, June 17, 2002 1:40 PM
Subject: Re: [PIC]: Jtag and more


> No, that's so you can jump from the reset vector before the code hits the
> interrupt vector.  You need at least 3 instructions for this, two to set
> PCLATH properly, and one GOTO.  I use the 4th instruction to make sure
> interrupts are off so that a GOTO 0 (with PCLATH set appropriately of
> course) performs a software reset.

Fair enough.  Hadn't thought of that.  I'm used to a system where paging
isn't an issue, so I don't notice these things.

> I rather doubt Microchip cares much about the .01% of PICs sold that are
> bought by hobbiests.  Hobbiests like PICs because they are targeted for
low
> cost applications, and because Microchip learned early on that providing
> decent software tools for free helps to sell hardware.

Well, that's also true, that they probably don't care, but they might.
After all, look at the exposure they get simply from providing something
that can be used for hobby type solutions.  Everyone who's done anything
with an MCU knows about PICs.  And I suspect that Microchip is in favour of
lists like this one being up and running, since it sure makes things easier
for anyone developing an app for a PIC.  I don't think this list would be
here if it weren't for the hobbyists.  I suspect that the Microchip
marketing department is smart enough to see that all the PIC hobbyists are a
valuable asset to them.

So, no, you're right, I doubt they care about the *profits* from the
hobbyists, but I expect that they do care about the exposure and support
that said hobbyists provide for their big clients.  At least, if *I* were in
their position, *I* would.

As to free development tools, Motorola did some of that for one of their
processors, but what they did was a little bit chincy: they provided a buggy
IDE for free, and a much better one for a price.  The processor was really
nice, but I'd go for the Microchip IDE over that any day.

--Brendan

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics


2002\06\18@031422 by Trevor Page

flavicon
face
Hi,

I know I'm probably having a complete blonde moment here, but it is my
understanding that when at the reset vector, we know that PCLATH has to be =
0, and the initialisation routine can always be placed within the first page
in program space. Hence, we can get by with a single GOTO at the reset
vector address, which branches to somewhere within the first page.

Please tell me what I've missed. Appologies in advance if this has been
discussed to death already.

Trev




> {Original Message removed}

2002\06\18@032858 by Michael Rigby-Jones

flavicon
face
{Quote hidden}

I think Olin was refering to the case when you want to jump to code that
isn't in the first page.  All 16 series PIC's that I have used have had the
interrupt vector at 04h, so I suspect that the first 4 instructions are
designed for page setting rather than as some kind of future expansion.  If
they wanted to put multiple interrupt vectors in, they likely wouldn't place
them right next to each other anyway.

Regards

Mike

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


2002\06\18@034427 by Alan B. Pearce

face picon face
>Hence, we can get by with a single GOTO at
>the reset vector address, which branches
>to somewhere within the first page.

If you are using linkable modules you cannot guarantee that the code you
will be jumping to is in the first page.

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


2002\06\18@082813 by Olin Lathrop

face picon face
> > No, that's so you can jump from the reset vector before the code hits
the
> > interrupt vector.  You need at least 3 instructions for this, two to set
> > PCLATH properly, and one GOTO.  I use the 4th instruction to make sure
> > interrupts are off so that a GOTO 0 (with PCLATH set appropriately of
> > course) performs a software reset.
>
> I know I'm probably having a complete blonde moment here, but it is my
> understanding that when at the reset vector, we know that PCLATH has to be
=
> 0, and the initialisation routine can always be placed within the first
page
> in program space. Hence, we can get by with a single GOTO at the reset
> vector address, which branches to somewhere within the first page.

True, but I like to let the linker put my startup module wherever it feels
like.  There is little point in economizing instructions after a restart,
and it would be hard to put the remaining locations before address 4 to a
useful purpose anyway.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

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


2002\06\18@082830 by Olin Lathrop

face picon face
> >Hence, we can get by with a single GOTO at
> >the reset vector address, which branches
> >to somewhere within the first page.
>
> If you are using linkable modules you cannot guarantee that the code you
> will be jumping to is in the first page.

Actually you can, but I don't either because I see no advantage to it.


*****************************************************************
Embed Inc, embedded system specialists in Littleton Massachusetts
(978) 742-9014, http://www.embedinc.com

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


More... (looser matching)
- Last day of these posts
- In 2002 , 2003 only
- Today
- New search...