Reply by martin de lange February 8, 20042004-02-08
This might throw a spanner into your works at this last stage of design but why do you not use the 2129 part as it has ADC built in with lot's of IO and you need no multipexing here!
----- Original Message -----
From: J.C. Wren
To: l...@yahoogroups.com
Sent: Saturday, February 07, 2004 9:27 PM
Subject: [lpc2100] Evalulation Card Update

I've got an overview of what I'm thinking of going with for the
evaluation card.  < http://tinymicros.com/embedded/LPC210x >

Basic goals:  Have several of each bus type supported, i.e., several SPI
peripherials, several I2C, and several 1-wire.  Try to make devices
generally useful.

All devices are 3.3V.  All devices can be disconnected from Vcc and
their respective control signals via jumpers (can you say :lots of
jumpers"?).  In addition to the two JTAG interfaces being available, a
Wiggler type interface is built-in.  Also the "hands-off" ISP
programming mode is supported.

One DB-9 for serial port 0 (2 lines), one DB-9 for serial port 1 (with
all 8 control signals).

Where signals conflict, jumpers will allow selecting the use of one set
or the alternate set.  All I/Os are available on 2 headers.  By removing
the device jumpers, the LPC210x can be completely unperipherialed.  By
removing the device Vcc jumpers, one should be able to measure currents
for a given configuration, such as "how much current will just the
LPC2106 + SD card pull?".

Reasonable power supply for on-board devices, plus a fair chunk of
capacity for users circuity.  I'm considering either putting jumpers on
the 3 Vcc rails, or 1 ohm resistors, for current measurement.  Also
including several points to attach a 'scope ground (likely pads for a .1
or .2 inch wire loop).

There will probably be no prototype area.  However, the I/O connectors
will be arranged so that the user can plug in a card on Vector board on
.1 centers, and be able to get power, I/O, and control signals. 
Prototyping area costs real money, and I personally hate defacing
boards.  If there's enough feedback, I'll add it on.  Please specify how
much area you'd want, to .1 inch resolution.

The SD/MMC card connector will support the physical SD card.  I had
planned on only supporting the SPI mode, but there's several uncommitted
I/O lines on the processor, so I may wire those up (with jumpers, of
course) to support SD mode.

The A/D is an 8 channel 10 bit part.  One of the MCP23016 I/O expanders
will have 8 LEDS and 8 switches (drawing shows pushy buttons, they'll
actually be rockers).  The other MCP23016 pins will be available on a
2x10 header.  NVRAM shown is a Ramtrom FRAM, any 24CL64 device should
work (pin outs are the same).  RTC will have a battery (not shown),
diode OR'ed in with Vcc.  DS2405 1-wire switch has LED to indicate
state.  I2C devices with interrupt lines will be jumperable to EINT0,
EINT1 or EINT2.

I think that's everything.  I'll be doing schematics over the weekend,
most likely.  If you have any ideas, I'd love to hear them.

    --jc

An Engineer's Guide to the LPC2100 Series

Reply by David Willmore February 7, 20042004-02-07
J.D. Wren,

The design sounds good. A full kitchen sink design. :)

> The SD/MMC card connector will support the physical SD card. I had
> planned on only supporting the SPI mode, but there's several uncommitted
> I/O lines on the processor, so I may wire those up (with jumpers, of
> course) to support SD mode.

Just one little thing on this. If you put a jumpers in to support
nibble mode, support not using the SPI signals for D0/SDA and SCL--you
might want to use those signals with a real SPI peripherial. Or is
that solved with CS signals? I'm not a big SPI user, I just read the
spec that was pointed out here the other day and this was on the top
of my head.

Now I really want to stuff an SD card slot on a board and start playing
with them. Low pin count, removable, cost competative, etc. Sounds
good to me.

Cheers,
David



Reply by J.C. Wren February 7, 20042004-02-07
Arie de Muynck wrote:

> > I've got an overview of what I'm thinking of going with for the
> > evaluation card. < http://tinymicros.com/embedded/LPC210x >
>
> 1.
> In the block diagram it says "Off-board connectors are 2.5mm".
> I hope you mean 0.1", that is 2.54 mm, so one might use standard flatcable
> connectors?
>
No, the 2.5mm is correct. If you'll notice, that applies to the coaxial
power connector. 2.1mm and 2.5mm are the two standard center pins
sizes. All 2x? headers are on .1 inch centers.

> 2.
> Xtal.74 MHZ is easier obtainable and gives better baudrates?

Yes. I didn't have the exact frequency on hand last night. I plan to
socket the crystal, and the clearances are for a standard HC-49 sized
footprint. I may put a set of SMT pads down adjactent to it.

>
>
> Regards,
> Arie de Muynck
>




Reply by Arie de Muynck February 7, 20042004-02-07
> I've got an overview of what I'm thinking of going with for the
> evaluation card. < http://tinymicros.com/embedded/LPC210x >

1.
In the block diagram it says "Off-board connectors are 2.5mm".
I hope you mean 0.1", that is 2.54 mm, so one might use standard flatcable
connectors?

2.
Xtal.74 MHZ is easier obtainable and gives better baudrates? Regards,
Arie de Muynck




Reply by haack0815 February 7, 20042004-02-07
Hi Matthias,

The max. SPI rate is 7.5MHz (PCLK/VPBDIV = 60MHz/8)

Andreas

--- In , Matthias Weingart <lpc2100@p...> wrote:
> On Sat, Feb 07, 2004 at 03:27:23AM -0500, J.C. Wren wrote:
> > The SD/MMC card connector will support the physical SD card. I had
> > planned on only supporting the SPI mode, but there's several
uncommitted
> > I/O lines on the processor, so I may wire those up (with jumpers, of
> > course) to support SD mode.
>
> Some comments to SD and MMC cards. In SPI mode all commands from the
SD mode
> are also available. One-Wire SD mode has the same speed as the SPI mode.
> (3.125MBit/s burst rate at 25Mz clock). You can speed up this to
12MBit/s by
> using the 4 data lines in 4bit SD mode, but I guess this can not be
done by
> using simple IO-Pins, or timer inputs/outputs (much too slow). With no
> additional hardware the SPI-mode of the SD card is the best solution
for the
> lpc21xx. And I doubt that you can use it at the full speed. (What is the
> max. frequency of the lpc21xx SPI?).
>
> Matthias




Reply by Matthias Weingart February 7, 20042004-02-07
On Sat, Feb 07, 2004 at 03:27:23AM -0500, J.C. Wren wrote:
> The SD/MMC card connector will support the physical SD card. I had
> planned on only supporting the SPI mode, but there's several uncommitted
> I/O lines on the processor, so I may wire those up (with jumpers, of
> course) to support SD mode.

Some comments to SD and MMC cards. In SPI mode all commands from the SD mode
are also available. One-Wire SD mode has the same speed as the SPI mode.
(3.125MBit/s burst rate at 25Mz clock). You can speed up this to 12MBit/s by
using the 4 data lines in 4bit SD mode, but I guess this can not be done by
using simple IO-Pins, or timer inputs/outputs (much too slow). With no
additional hardware the SPI-mode of the SD card is the best solution for the
lpc21xx. And I doubt that you can use it at the full speed. (What is the
max. frequency of the lpc21xx SPI?).

Matthias




Reply by J.C. Wren February 7, 20042004-02-07
I've got an overview of what I'm thinking of going with for the
evaluation card. < http://tinymicros.com/embedded/LPC210x >

Basic goals: Have several of each bus type supported, i.e., several SPI
peripherials, several I2C, and several 1-wire. Try to make devices
generally useful.

All devices are 3.3V. All devices can be disconnected from Vcc and
their respective control signals via jumpers (can you say :lots of
jumpers"?). In addition to the two JTAG interfaces being available, a
Wiggler type interface is built-in. Also the "hands-off" ISP
programming mode is supported.

One DB-9 for serial port 0 (2 lines), one DB-9 for serial port 1 (with
all 8 control signals).

Where signals conflict, jumpers will allow selecting the use of one set
or the alternate set. All I/Os are available on 2 headers. By removing
the device jumpers, the LPC210x can be completely unperipherialed. By
removing the device Vcc jumpers, one should be able to measure currents
for a given configuration, such as "how much current will just the
LPC2106 + SD card pull?".

Reasonable power supply for on-board devices, plus a fair chunk of
capacity for users circuity. I'm considering either putting jumpers on
the 3 Vcc rails, or 1 ohm resistors, for current measurement. Also
including several points to attach a 'scope ground (likely pads for a .1
or .2 inch wire loop).

There will probably be no prototype area. However, the I/O connectors
will be arranged so that the user can plug in a card on Vector board on
.1 centers, and be able to get power, I/O, and control signals.
Prototyping area costs real money, and I personally hate defacing
boards. If there's enough feedback, I'll add it on. Please specify how
much area you'd want, to .1 inch resolution.

The SD/MMC card connector will support the physical SD card. I had
planned on only supporting the SPI mode, but there's several uncommitted
I/O lines on the processor, so I may wire those up (with jumpers, of
course) to support SD mode.

The A/D is an 8 channel 10 bit part. One of the MCP23016 I/O expanders
will have 8 LEDS and 8 switches (drawing shows pushy buttons, they'll
actually be rockers). The other MCP23016 pins will be available on a
2x10 header. NVRAM shown is a Ramtrom FRAM, any 24CL64 device should
work (pin outs are the same). RTC will have a battery (not shown),
diode OR'ed in with Vcc. DS2405 1-wire switch has LED to indicate
state. I2C devices with interrupt lines will be jumperable to EINT0,
EINT1 or EINT2.

I think that's everything. I'll be doing schematics over the weekend,
most likely. If you have any ideas, I'd love to hear them.

--jc