Searching \ for 'I2C over "long"distance ?' 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/i2cs.htm?key=i2c
Search entire site for: 'I2C over "long"distance ?'.

Truncated match.
PICList Thread
'I2C over "long"distance ?'
1999\04\29@055650 by Benjamin Petersen

flavicon
face
I am planning to have 8 devices on an I2C interface, but with a cable length
of atleast 50 meters (thats 150 feet). I was thinking of making use of a
max483, but there must be a better way... I was also thinking of mixing
power lines with signal lines, only having 2 wires to the I2C unit. Each I2C
unit will just be a DS/LM75. I know that I2C was ment for short distance,
but i was thinking that lower the clockrate to say 100hz would help. (or
better yet a clockrate that is out of phase with AC power lines... I was
also thinking about sending the data by using the AC power lines... Does
anyone have a clue of what to do ?

Regards
Benjamin Petersen

1999\04\29@071727 by Joe McCauley

picon face
There is a Phillips bus driver chip for I2C (P82B715). This allows driving
a bus with 100 times normal I2C spec capacitance. They cost about $5 in
single quantities. I don't know how this would operate if you try to also
power the remote devices from the data & clock lines.
Clocking at lower speeds will definitly help. Let me know how it works out
please. I have a similar application which I will be testing in the next
few weeks. The remote units are 16F84s running an I2C slave program. I have
not yet tested it to 50 meters (30 so far & it works well) so I'm
interested to see how you get on.


Joe


At 11:54 29/04/99 +0200, you wrote:
{Quote hidden}

1999\04\29@075443 by Benjamin Petersen

flavicon
face
I just found :

Re: I2c or SPI over long distances
Subject: Re: I2c or SPI over long distances
From: Ian King <spam_OUTikTakeThisOuTspamADV.SBC.SONY.CO.JP>
Date: Wed, 18 Oct 1995 09:24:57 BST
http://www.am.qub.ac.uk/world/lists/pic/msg00221.html

But I would still like to know about mixing power and signal lines.

Regards
Benjamin Petersen

1999\04\29@112009 by hgraf

picon face
> I am planning to have 8 devices on an I2C interface, but with a
> cable length
> of atleast 50 meters (thats 150 feet). I was thinking of making use of a
> max483, but there must be a better way... I was also thinking of mixing
> power lines with signal lines, only having 2 wires to the I2C
> unit. Each I2C
> unit will just be a DS/LM75. I know that I2C was ment for short distance,
> but i was thinking that lower the clockrate to say 100hz would help. (or
> better yet a clockrate that is out of phase with AC power lines... I was
> also thinking about sending the data by using the AC power lines... Does
> anyone have a clue of what to do ?

       Well I am no expert, but I have done what you are proposing. I have a ru
n
of about 40 meters, 3 LM75 devices and 1 MAX I2C device that senses on or
off conditions (don't remember the part number). Anyways, I find that it
works OK, I do use a slower clock speed and usually read from devices more
than once (I have it set to 10 times) and then filter out the highest a
lowest value. It works almost flawlessly, I have all the devices hooked up
in a daisy chain fashion using 4 conductor phone cable, power, GND, SCL and
SDA. I found that you have to bypass the supply on each device (kind of
obvious, but I didn't do it at first). In my case I have a sensor outside,
in the fridge and in my upright freezer (gets down to -25 degrees C in
there) and each device works well. If you have any questions let me know.
TTYL

1999\04\29@113524 by Bernhard Kraft

flavicon
face
On Thu, 29 Apr 1999, Benjamin Petersen wrote:

>I am planning to have 8 devices on an I2C interface, but with a cable length
>of atleast 50 meters (thats 150 feet). I was thinking of making use of a
>max483, but there must be a better way... I was also thinking of mixing
>power lines with signal lines, only having 2 wires to the I2C unit. Each I2C
>unit will just be a DS/LM75. I know that I2C was ment for short distance,

I tried to communicate to a LM75 Temperature Sensor over a 150m long cable
(rolled up) but it didn't work ... The PIC I used was able to pull the
data and clock line low and the signal was quite pretty (I observed it
with an oscilloscope) but the LM75 didn't seem to be able to pull the line
low. Don't know how big the line capacitance could be so that the LM75
would still be able to pull down the line.

===================================================================
Kraft Bernhard aka. Krufti              /"\
ICQ# 3672982                            \ /  ASCII RIPPON CAMPAIGN
                                        X   AGAINST HTML MAIL
                                       / \
===================================================================

1999\04\29@114518 by Benjamin Petersen

flavicon
face
{Quote hidden}

Just to tell you what NOT to do :
Don't use twisted wires for SCL & SDA this didn't work for me not even with
30 cm (1 foot). I replased the LM 3 times, made new pcb each time... Took me
one week until i used another wire. Then it worked right away... But then
again i am just a programmer..

Regards
 Benjamin Petersen

1999\04\29@120854 by David Wong

flavicon
face
I've had some success using I2C with ribbon cable.  Alternate the signal
wires with ground wires in between this helps to shield noise.  I was able
to drive 100 ft cable without drivers with less than ¸ percent errors.

DW
       {Original Message removed}

1999\04\29@121057 by David Wong

flavicon
face
Oh also one more thing a pull up on both sides of the cable helps. Just make
sure the parallel resistance doesn't exceed the 3mA spec.

       {Original Message removed}

More... (looser matching)
- Last day of these posts
- In 1999 , 2000 only
- Today
- New search...