Searching \ for '[PIC]: Pullups needed on an SPI circuit?' 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/ios.htm?key=spi
Search entire site for: 'Pullups needed on an SPI circuit?'.

Exact match. Not showing close matches.
PICList Thread
'[PIC]: Pullups needed on an SPI circuit?'
2005\04\16@155238 by Richard Zinn

picon face
I know that when I've implemented I2C buses with my pics I've needed to use pullups. In all the examples I've seen with SPI except one I haven't seen pullups - so can anyone tell me for sure? Have an example?
Thanks.

2005\04\16@162719 by Stef Mientki

flavicon
face


Richard Zinn wrote:

>I know that when I've implemented I2C buses with my pics I've needed to use
>pullups. In all the examples I've seen with SPI except one I haven't seen
>pullups - so can anyone tell me for sure? Have an example?
>  
>
no pullups needed,
because you output are push-pull instead of open-collector.
The difference is that with SPI each device is addressed by a separate line.

Stef

> Thanks.
>  
>

2005\04\16@164410 by Richard Zinn

picon face
awesome, thanks. And thanks for the explanation.

On 4/16/05, Stef Mientki <spam_OUTs.mientkiTakeThisOuTspammailbox.kun.nl> wrote:
{Quote hidden}

>

2005\04\16@194648 by Jinx

face picon face
> >I know that when I've implemented I2C buses

> no pullups needed,
> because you output are push-pull instead of open-collector.

The only GP o/d (open drain) is RA4

A hardware slave I2C PIC (eg F88) has both the usual N-ch /
P-ch push-pull i/o FETs. As a firmware master you use TRIS to
make "1" bits with the pull-ups, ie with TRIS <pin> set, the pin
as i/p means the SDA/SCL line pull-ups takes that line high.
PORT <pin> is always "0", so the line is either a driven "0" or
a passive "1"

2005\04\17@091259 by Gerhard Fiedler

picon face
Jinx wrote:

>>>I know that when I've implemented I2C buses
>
>> no pullups needed, because you output are push-pull instead of
>> open-collector.

> A hardware slave I2C PIC (eg F88) has both the usual N-ch / P-ch
> push-pull i/o FETs.

You probably meant the right thing, but it sounds odd. While the ports do
have the usual push-pull output FETs, in I2C mode the push-up FET is
disabled -- as it has to be. So effectively, in I2C mode the SPI data/clock
lines are open drain outputs -- on both master and slave.

Gerhard

2005\04\25@151304 by Barry Gershenfeld

face picon face

> >>>I know that when I've implemented I2C buses
> >
> >> no pullups needed, because you output are push-pull instead of
> >> open-collector.
>
> > A hardware slave I2C PIC (eg F88) has both the usual N-ch / P-ch
> > push-pull i/o FETs.
>
>You probably meant the right thing, but it sounds odd. While the ports do
>have the usual push-pull output FETs, in I2C mode the push-up FET is
>disabled -- as it has to be. So effectively, in I2C mode the SPI data/clock
>lines are open drain outputs -- on both master and slave.

All well and good but perhaps a complex way to explain, that i2c is
bidirectional and
SPI is not.  SPI is a good old fashioned TTL-style hookup -- everything that
is an output connects only to inputs, and all inputs and outputs stay that
way.  It's
also why SPI has more wires, because there's a Data Out and an Data In.

Barry

2005\04\26@125851 by Herbert Graf

flavicon
face
On Mon, 2005-04-25 at 12:12 -0700, Barry Gershenfeld wrote:
> All well and good but perhaps a complex way to explain, that i2c is
> bidirectional and
> SPI is not.  SPI is a good old fashioned TTL-style hookup -- everything that
> is an output connects only to inputs, and all inputs and outputs stay that
> way.  It's
> also why SPI has more wires, because there's a Data Out and an Data In.

Along with a CS for each device, making SPI VERY painful if you want to
use several devices. That said it is "simpler" due to the normally
single direction signals. TTYL

-----------------------------
Herbert's PIC Stuff:
http://repatch.dyndns.org:8383/pic_stuff/

2005\04\27@095824 by Gerhard Fiedler

picon face
Herbert Graf wrote:

> Along with a CS for each device, making SPI VERY painful if you want to
> use several devices. That said it is "simpler" due to the normally
> single direction signals. TTYL

And it's especially a lot simpler to implement in hardware. That's probably
the reason why SPI chips usually are cheaper than I2C, at least where the
additional I2C logic is not negligible compared to the chip complexity.

Gerhard

2005\04\27@113958 by olin_piclist

face picon face
Gerhard Fiedler wrote:
> And it's especially a lot simpler to implement in hardware. That's
> probably the reason why SPI chips usually are cheaper than I2C, at
> least where the additional I2C logic is not negligible compared to the
> chip complexity.

Or what about license issues.  I know Phillips owns IIC, but is a license
required for SPI?  I don't remember any such restriction, but I could easily
be wrong about that.


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

2005\04\27@120607 by Spehro Pefhany

picon face
At 11:40 AM 4/27/2005 -0400, you wrote:
>Gerhard Fiedler wrote:
>>And it's especially a lot simpler to implement in hardware. That's
>>probably the reason why SPI chips usually are cheaper than I2C, at
>>least where the additional I2C logic is not negligible compared to the
>>chip complexity.
>
>Or what about license issues.  I know Phillips owns IIC, but is a license
>required for SPI?  I don't remember any such restriction, but I could easily
>be wrong about that.

No, no licence/license for SPI. It was developed by Motorola. But
there are many variations of SPI, which is a downside.

SPI EEPROM chips have historically been significantly more expensive than
I2C parts, probably because of higher usage. The SPI parts are also faster
to access and  thus can command a bit of a premium. I don't think a few
hundred gates matter that much these days, even on a 35-cent part.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
.....speffKILLspamspam@spam@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




2005\04\27@131156 by Wouter van Ooijen

face picon face
> Or what about license issues.  I know Phillips owns IIC, but
> is a license required for SPI?

As I have never seen a real definition document for SPI I doubt it. (I
don't consider a chapter in some 68HC datasheet a defintion document). I
doubt anyone realy invented (as opposed to documented) SPI: a lot of
plain 'TTL' chips are SPI compatible.

Wouter van Ooijen

-- -------------------------------------------
Van Ooijen Technische Informatica: http://www.voti.nl
consultancy, development, PICmicro products
docent Hogeschool van Utrecht: http://www.voti.nl/hvu


2005\04\27@164118 by Andrew Warren

face
flavicon
face
Olin Lathrop <piclistspamKILLspammit.edu> wrote:

> I know Phillips owns IIC, but is a license required for SPI?  I don't
> remember any such restriction, but I could easily be wrong about that.

Olin:

There's no license required for SPI.  In fact, there isn't even a
real spec for it; in the design documents I wrote for Cypress's
microcontrollers, I was forced by the lack of an official Standard to
require "SPI as described in Motorola Semiconductor's M68HC11
Reference Manual".

-Andrew

=== Andrew Warren - .....fastfwdKILLspamspam.....ix.netcom.com

2005\04\27@225556 by Douglas Wood

picon face
I went looking for the "original" SPI specification from Motorola a couple
of years ago. I ended up talking to an engineer at Motorola who was
supposedly the "expert" on such documents and she said that there was no
such document (or spec, for that matter). She said that SPI was originally
not intended for public consumption; it was for some type of internal
project and it just grow from that.

In my search I did, however, find a number of articles and designs using SPI
that put a pull-up on the MOSI (master out/slave in or SDO, as it appears on
the PIC).

Douglas Wood

{Original Message removed}

2005\04\27@231048 by Spehro Pefhany

picon face
At 09:55 PM 4/27/2005 -0500, you wrote:
>I went looking for the "original" SPI specification from Motorola a couple
>of years ago. I ended up talking to an engineer at Motorola who was
>supposedly the "expert" on such documents and she said that there was no
>such document (or spec, for that matter). She said that SPI was originally
>not intended for public consumption; it was for some type of internal
>project and it just grow from that.
>
>In my search I did, however, find a number of articles and designs using
>SPI that put a pull-up on the MOSI (master out/slave in or SDO, as it
>appears on the PIC).
>
>Douglas Wood


In 2002, I posted a link to this document:
www.bio-bot.com/privateweb/datasheets/Processors/DSP/AN991.pdf
to comp.arch.embedded. It may be of some help, however I don't think
it's in any way essential.

Pullups? Well, I've designed them into SPI circuits, to handle transient
conditions.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
EraseMEspeffspam_OUTspamTakeThisOuTinterlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com




2005\04\28@064931 by Gerhard Fiedler

picon face
Wouter van Ooijen wrote:

>> Or what about license issues.  I know Phillips owns IIC, but
>> is a license required for SPI?
>
> As I have never seen a real definition document for SPI I doubt it. (I
> don't consider a chapter in some 68HC datasheet a defintion document). I
> doubt anyone realy invented (as opposed to documented) SPI: a lot of
> plain 'TTL' chips are SPI compatible.

"SPI" is nothing more than a shift register interface. There is so much
prior art for that... probably difficult to claim anything in terms of
ownership.

It would make sense though to standardize things; OTOH, it seems that most
SPI chips are following some inofficial standard WRT clock edges etc., as
it's become less frequent that I have to change modes between chips. Or
have I just been lucky?

Gerhard

More... (looser matching)
- Last day of these posts
- In 2005 , 2006 only
- Today
- New search...