Multi Master RS485 is it possible ?
Dr John Tankard
I have seen many examples of a single master multiple slaves RS485, my
question is :- is there any reason why there cannot be multiple
masters on the wire ?
Thanks in advance
John Tankard
David R Brooks
jtan...@tankard.com (Dr John Tankard) wrote:
:I have seen many examples of a single master multiple slaves RS485, my
:question is :- is there any reason why there cannot be multiple
:masters on the wire ?
:
None at all: you just need a robust protocol for passing the
mastership amongst them. Remember that RS485, unlike (say) Ethernet,
cannot tolerate 2 "masters" each trying to drive the bus in opposite
states. Hardware damage may ensue.
Or (if speed is not of the essence) consider the SAE J1708 scheme:
essentially RS485 turned into a wired-AND scheme with open-collector
drives. Two "masters" can drive, and the line will take the AND state
of them. You now need to detect (the basic CSMA problem) that you
don't have absolute mastery, and back off.
-- Dave Brooks <http://www.iinet.net.au/~daveb>
PGP public key: finger da...@opera.iinet.net.au
servers da...@iinet.net.au
fingerprint 20 8F 95 22 96 D6 1C 0B 3D 4D C3 D4 50 A1 C4 34
What's all this? see http://www.iinet.net.au/~daveb/crypto.html
Uwe Hercksen
On Wed, 05 Nov 1997 09:52:36 GMT, jtan...@tankard.com (Dr John
Tankard) wrote:
>I have seen many examples of a single master multiple slaves RS485, my
>question is :- is there any reason why there cannot be multiple
>masters on the wire ?
>
Hello,
the single master multiple slaves is easier to implement.
You will have a driver and a receiver for each master and each slave
connect to the same two wire bus.
For a correct and reliable transmission, only one driver should be
enabled at a time.
A single master owns the bus: it enables its driver, send some
characters to all the slaves, send it disables the masters driver.
Only one of the slave will answer, enable its own driver, send its
reply and disables the driver again.
Now the master may talk again, with the same slave or one of the other
slaves.
This scheme requires a protocol, the slave has to know how many
characters it will get from the master for each message type and the
master has to know how many chars the slave will respond for this
message. Between the last character of a message from the master and
the first character of the slave's response, there must be a short
delay to disable the masters driver and to enable the slaves driver.
The same is true for the last char of the response from the slave and
the first char of the next message from the slave.
If you want to use a multiple master RS485, you will need a mechanism
for the avoidance of a collison ( The Token Ring uses the token
passing scheme) or at least a collision detection ( Ethernet uses
CSMA/CD "Carrier Sense Multiple Access Collison Detect)
The multiple master RS485 requires more hardware and software to
implement and the performance may less than the performance of a
single master scheme.
By
Uwe Hercksen
herc...@nospam.uni-erlangen.de
For EMail response, please replace nospam with zew
--------------------------------------------------
Elektronikwerkstatt Uni. Erlangen
Cauerstr. 5
D91058 Erlangen
Paul E. Bennett
In article <34604...@katana.legend.co.uk>
jtan...@tankard.com "Dr John Tankard" writes:
> I have seen many examples of a single master multiple slaves RS485, my
> question is :- is there any reason why there cannot be multiple
> masters on the wire ?
No reason whatsoever. When multi-mastering you will need to develop a scheme
for detecting and dealing with collisions. If all nodes were to have a
definite listening time plus a random extra period then you should minimise
when such collisions would occur. All the other techniques and protocols
discussed here previously remain valid for consideration when determining
the protocols you will use.
--
Paul E. Bennett ................... <p...@transcontech.co.uk>
Transport Control Technology Ltd. <http://www.tcontec.demon.co.uk/>
+44 (0)117-9499861 <enq...@transcontech.co.uk>
Going Forth Safely
Manfred Bartz
jtan...@tankard.com (Dr John Tankard) writes:
> I have seen many examples of a single master multiple slaves RS485, my
> question is :- is there any reason why there cannot be multiple
> masters on the wire ?
No, but:
If your drivers are directly connected to a properly terminated RS485
bus, only one driver may transmit at any one time or you'll blow 'em
up. You can still implement a multimaster system by using a token
passing scheme, maybe with one super-master which has ultimate control
of the token.
You can use a different scheme of connecting your drivers as defined
in SAE J1708 (NS application note AN-915). That scheme allows
collision detection on an unterminated bus. You will need to allow
for random retry and progressive back-off as used in ethernet (IEEE-802.3).
Direct driver connection and token passing allows you to utilize the
full bandwith of the RS485 specs. -- SAE J1708 gives you collision
capabilities at the expense of bandwidth and cable length.
If you need guaranteed response times, a token passing scheme will be
much more predictable.
Regards
--
Manfred Bartz <mba...@werple.net.au>
-------------------------------------------------------------------------------
Unsolicited commercial E-Mail will be billed at $500 per message for proof-
reading services. This rate is not negotiable, there are no quantity discounts.
-------------------------------------------------------------------------------
Dr John Tankard
z...@ds.com (Peter) wrote:
>>No reason whatsoever. When multi-mastering you will need to develop a scheme
>>for detecting and dealing with collisions. If all nodes were to have a
>>definite listening time plus a random extra period then you should minimise
>>when such collisions would occur. All the other techniques and protocols
>>discussed here previously remain valid for consideration when determining
>>the protocols you will use.
>The problem with a "collision" between two or more RS485 drivers is
>that the drivers are NOT intended to be used that way.
>One driver, driven with (e.g.) a 1 driving another driver driven with
>a 0 is worse than driving a short circuit. Look at the safe operating
>area of the output transistors.
>No doubt the devices will withstand it - they should be designed for
>that - but I would not like to be standing up in a court, being sued
>by somebody, and have to answer a question on this subject composed by
>an expert witness. "Yes, sir, you are right in that the manufacturer's
>data does not specify this, but I found it actually works quite
>well..".
>Worse, most of the people who regularly post this (or very similar)
>question are going to try to detect collisions via their
>microcontroller's UART. A normal UART will not give you any indication
>of anything happening until it sees a start bit, followed by enough
>time to receive a nominal character at the configured baud rate. Then,
>you get something in the RX buffer, and/or a framing error.
>So one could get lots and lots of collisions, but the UART will not
>appear to see them, unless/until a start bit manages to appear out of
>the bus contention garbage. This is not good.
>IMHO the only decent way to run multi-master with normal (i.e.
>DC-coupled) RS485 is to have extra circuitry to sense the DC levels.
Peter,
You say 'A normal UART will not give any indication' but are there
UART's that will give line status info ?
Dr John Tankard
z...@ds.com (Peter) wrote:
>Peter.
You say that 'A normal UART will not give you any indication' but are
there UART's that will give line status info and that will stand
collisions in there normal operating parameters ?
John
Lee Mitchell
Dr John Tankard wrote in message <34604...@katana.legend.co.uk>...
>I have seen many examples of a single master multiple slaves RS485, my
>question is :- is there any reason why there cannot be multiple
>masters on the wire ?
>
You may find that if the cable run is very long, (100 Metres +), it can be
hard for a device on one end of the line to detect a collision with another
device at the other end.
eg.
If two devices (A & B) begin transmitting at the same time, the amplitude of
of the signal sent by "A" , when received at the far end of the cable by
device "B" , may not be great enough to overpower the signal sent by "B" and
trigger a bus contention signal, therefore both devices continue to
transmitt. However if the cable is short, then its just a case of protocols
!
--
Lee Mitchell
L...@spamtastic.demon.co.uk
"Life !... Don't talk to me about life .."
Harold Rabbie
Dr John Tankard <jtan...@tankard.com> wrote in article
<34604...@katana.legend.co.uk>...
> I have seen many examples of a single master multiple slaves RS485, my
> question is :- is there any reason why there cannot be multiple
> masters on the wire ?
>
John,
What you need to support multiple masters (a.k.a. peer-to-peer) is a
media access control (MAC) algorithm. This is part of the link layer in
the 7-layer protocol model.
One easy-to-use protocol stack that includes a high performance
peer-to-peer MAC layer is the LonTalk protocol. It supports RS-485 as well
as other media. Microcontroller chips that contain this protocol (called
Neuron Chips) are manufactured by Motorola (part numbers MC143150xx,
MC143120xx) and Toshiba (part numbers TMPN3150xx, TMPN3120xx)
See the web sites http://www.mot.com/SPS/MCTG/MDAD/lonworks/
http://www.toshiba.com/taec/ and http://www.echelon.com/Core/protocol/ for
more information.
Regards
--
Harold Rabbie
Saratoga, CA
Remove .NOSPAM from my address before replying
Werner Huerttlen
z...@ds.com (Peter) wrote:
>>If you need guaranteed response times, a token passing scheme will be
>>much more predictable.
>IMHO one should be careful with this assumption, although it is one
>often repeated.
Anyone tried to use ArcNet Chips, like the SMC20051, to drive data
over RS485 ?
I am in thinking to use this chip, let it do all the low-level-work
and hope it solves all the problems with accessing the media and error
correction.
Werner
Best regards,
Werner
* Opinions expressed herein are my own and may
* not represent those of my employer
Pascal Dornier
Werner Huerttlen wrote in message <6473o4$6...@zam201.zam.kfa-juelich.de>...
>Anyone tried to use ArcNet Chips, like the SMC20051, to drive data
>over RS485 ?
>
>I am in thinking to use this chip, let it do all the low-level-work
>and hope it solves all the problems with accessing the media and error
>correction.
In the "typical" case it would work, but you may still need high level
code to handle time-outs and lost packets.
Arcnet can have relatively long delays for reconfiguration when a
device is added or removed, or when the token is dropped due to a
bit error.
--------------------------------------------------------------------
Pascal Dornier pdor...@pcengines.com http://www.pcengines.com
Your Spec + PC Engines = Custom Embedded PC Hardware
--------------------------------------------------------------------
erikc
On 10 Nov 1997 06:48:49 GMT
"Harold Rabbie" <hzra...@worldnet.att.net>
-- origin: comp.arch.embedded:
start around US$10K.
Erikc (WA #002) | "An Fhirinne in aghaidh an tSaoil."
| "The Truth against the World."
| -- Bardic Motto
Eric Doenges
fire...@MAPSITNA.insync.net (erikc) writes:
>Aren't the Neuron chips rather expensice, though? IIRC, the development tools
>start around US$10K.
You don't have to use the Neuron chips, as Echelon has opened their
protokoll to everyone who wants to implement it. I believe an
implementation for the Motorola 68302 already exists. Echelon does want
something in the order of 40 cents or so to have a unique (world wide !)
ID assigned to your LonTalk device.
--
Eric Doenges
EMail:<doe...@lpr.e-technik.tu-muenchen.de>
"You don't have to swim faster than the shark,
just faster than the guy next to you" - anonymous
Dr John Tankard
Harold,
Thanks for the links, it looks as though i have some reading to do !
Best regards
John Tankard
jtan...@tankard.com
"Harold Rabbie" <hzra...@worldnet.att.net> wrote:
>John,
> What you need to support multiple masters (a.k.a. peer-to-peer) is a
>media access control (MAC) algorithm. This is part of the link layer in
>the 7-layer protocol model.
>
> One easy-to-use protocol stack that includes a high performance
>peer-to-peer MAC layer is the LonTalk protocol. It supports RS-485 as well
>as other media. Microcontroller chips that contain this protocol (called
>Neuron Chips) are manufactured by Motorola (part numbers MC143150xx,
>MC143120xx) and Toshiba (part numbers TMPN3150xx, TMPN3120xx)
>
>See the web sites http://www.mot.com/SPS/MCTG/MDAD/lonworks/
>http://www.toshiba.com/taec/ and http://www.echelon.com/Core/protocol/ for
>more information.
>
>Regards
Dr John Tankard
Dear Lee
It looks as though the problems with collision detection are much
greater that i thought, perhaps i should look as some of the
alternatives which have been posted on this thread.
Thank you
Jack Bonn
This is a multi-part message in MIME format.
--------------189A5AC344B344B30B8E5847
Content-Type: text/plain; charset=us-ascii
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Content-Transfer-Encoding: 7bit
Uwe Hercksen wrote:
> On Wed, 05 Nov 1997 09:52:36 GMT, jtan...@tankard.com (Dr John
> Tankard) wrote:
>
> >I have seen many examples of a single master multiple slaves RS485,
> my
> >question is :- is there any reason why there cannot be multiple
> >masters on the wire ?
>
> the single master multiple slaves is easier to implement.
No doubt! Many a bullet sweat on getting multi-master to work right.
> If you want to use a multiple master RS485, you will need a mechanism
> for the avoidance of a collison ( The Token Ring uses the token
> passing scheme) or at least a collision detection ( Ethernet uses
> CSMA/CD "Carrier Sense Multiple Access Collison Detect)
> The multiple master RS485 requires more hardware and software to
> implement and the performance may less than the performance of a
> single master scheme.
I implemented a multi-master protocol that ran on RS-485 a few years
back. It used a token bus (using nothing but the RS-485 for
interconnection) and utilized a "poll for master" such that a new node
could listen and eventually link himself into the master chain. It had
32 masters and (255-32) slaves, leaving one address for broadcast.
Target hosts included 68HC11s, 68302s, 68HC05 (slave only), and Intel
PCs.
Looks like a multi-master I^2C would blow the socks off this, however.
--
Jack Bonn <> Software Design Labs, Inc.
ja...@swlabs.com (847)526-1337
Ignorance and arrogance do a pathetic duet
--------------189A5AC344B344B30B8E5847
Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Jack Bonn
Content-Disposition: attachment; filename="vcard.vcf"
begin: vcard
fn: Jack Bonn
n: Bonn;Jack
org: Software Design Labs, Inc.
email;internet: ja...@swlabs.com
title: Consultant
x-mozilla-cpt: ;0
x-mozilla-html: TRUE
end: vcard
--------------189A5AC344B344B30B8E5847--
Werner Huerttlen
Jack Bonn <ja...@swlabs.com> wrote:
>I implemented a multi-master protocol that ran on RS-485 a few years
>back. It used a token bus (using nothing but the RS-485 for
>interconnection) and utilized a "poll for master" such that a new node
>could listen and eventually link himself into the master chain. It had
>32 masters and (255-32) slaves, leaving one address for broadcast.
would you please explain the key-features of your implementation ?
James Beck
w.hue...@kfa-juelich.de (Werner Huerttlen) wrote:
>Anyone tried to use ArcNet Chips, like the SMC20051, to drive data
>over RS485 ?
>I am in thinking to use this chip, let it do all the low-level-work
>and hope it solves all the problems with accessing the media and error
>correction.
I played with the SMC part# COM200191 ArcNet IC and RS-485 drivers. I
hate to say it, but I couldn't get the puppies to work reliably, but I
doubt it was the IC's fault. I keep saying I'll get the proto-boards
out agin and see what I did wrong, but "real" work keeps getting in
the way.
Jim
Mel (Stephen Mellor)
| James Beck said something that I couldn't ignore ...
> I played with the SMC part# COM200191 ArcNet IC and RS-485 drivers. I
> hate to say it, but I couldn't get the puppies to work reliably, but I
> doubt it was the IC's fault. I keep saying I'll get the proto-boards
> out agin and see what I did wrong, but "real" work keeps getting in
> the way.
Hum. I've been considering going this route - any out there have a success
story, or is it all bad news?
*Mel!*
Project engineer and bum.
Jack Bonn
This is a multi-part message in MIME format.
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Werner Huerttlen wrote:
> Jack Bonn <ja...@swlabs.com> wrote:
>
> >I implemented a multi-master protocol that ran on RS-485 a few years
> >back. It used a token bus (using nothing but the RS-485 for
> >interconnection) and utilized a "poll for master" such that a new node
> >could listen and eventually link himself into the master chain. It had
> >32 masters and (255-32) slaves, leaving one address for broadcast.
>
> would you please explain the key-features of your implementation ?
What more would you like to know?
It used:
o Standard (8 bit) UARTs
o Static address assignments
o Masters could originate communication to either other masters or slaves
o Token bus over RS-485
o Dynamic configuration, i.e. nodes could come and go at will
o CRC protection
Implementation was done in C and ran on various platforms.
--
Jack Bonn <> Software Design Labs, Inc.
ja...@swlabs.com (847)526-1337
Ignorance and arrogance do a pathetic duet
--------------F89ECB037374FB06BD9DBFA7
Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Jack Bonn
Content-Disposition: attachment; filename="vcard.vcf"
begin: vcard
fn: Jack Bonn
n: Bonn;Jack
org: Software Design Labs, Inc
adr: 26985 W. Fenview;;;Barrington;IL;60010;USA
email;internet: ja...@swlabs.com
title: Consultant
tel;work: +1 847 526 1337
tel;fax: +1 847 526 1311
x-mozilla-cpt: ;0
x-mozilla-html: TRUE
version: 2.1
end: vcard
--------------F89ECB037374FB06BD9DBFA7--
Werner Huerttlen
rksj...@mindspring.com (James Beck) wrote:
>I played with the SMC part# COM200191 ArcNet IC and RS-485 drivers. I
>hate to say it, but I couldn't get the puppies to work reliably, but I
Jim, thaks for this story. Exactly this I have assumed. We all would
be happy to hear the end of the story. What kind of CPU do you use ?
What is the cost of the COM200191 ?
Werner.
Werner Huerttlen
Jack Bonn <ja...@swlabs.com> wrote:
>What more would you like to know?
>It used:
>o Standard (8 bit) UARTs
ok
>o Static address assignments
ok
>o Masters could originate communication to either other masters or slaves
how is this done ?
>o Token bus over RS-485
how is the time-out set ? How do you manage if the master is lost,
e.g. powerer down.
>o Dynamic configuration, i.e. nodes could come and go at will
How does this work ?
>o CRC protection
Ok
>Implementation was done in C and ran on various platforms.
very nice
I am in designing a multi-master RS485 network. But actually found
out, collision is not allowed, because of chip-design. Possibly
token-passing is a fine way, but I assume very complicated.
Werner
Phil Dixon
David R Brooks wrote:
>
> jtan...@tankard.com (Dr John Tankard) wrote:
>
> :I have seen many examples of a single master multiple slaves RS485, my
> :question is :- is there any reason why there cannot be multiple
> :masters on the wire ?
> None at all: you just need a robust protocol for passing the
> mastership amongst them. Remember that RS485, unlike (say) Ethernet,
> cannot tolerate 2 "masters" each trying to drive the bus in opposite
> states. Hardware damage may ensue.
I had always understood this to be the difference between RS422 and
RS485.
Looking at a (very old, granted) TI Apps note, I found a note to the
effect that:
Standard RS485 differs from the RS422 primarily in the features that
allow reliable multipoint communication. For the drivers these
features are:
(Cut a few features)
o Drivers must have self protection against contention. That is, no
driver damage shall occur when its outputs are connected to a
voltage source of -7V to 12V whether its output state is a binary 1,
binary 1 or passive.
I would have thought that no matter what combination of drivers are
active, the result will not fall outside this range - i.e. an RS485
driver is protected against arbitrary numbers of collisions.
Have I missed something? I expect to need to implement a multi-master
RS485 bus soon!
Claudio Potenza
Werner Huerttlen <w.hue...@kfa-juelich.de> wrote:
> Anyone tried to use ArcNet Chips, like the SMC20051, to drive data
> over RS485 ?
I have extensively used SMC chips (20020 and 20050) with intel 186
microcontrollers at 2.5 and 5 Mbps but not with 485 on the line, using
instead the standard arcnet line interface (it sends on the line a
single period of a sinusoid for logic 1 and nothing for 0).
I had no real problem in using the SMC chips.
Arcnet is very reliable: it uses a token passing scheme
without the need for any 'master' node, you can insert and remove nodes
on a working network provoking usually a very short interruption (tens
of microseconds, but i seem to remember that in some circumstances it
can become of the order of half a second).
With a little software effort you can assign dinamically the addresses,
simplifying the configuration.
when you send a packet, the destination arcnet chip immediatly gives the
ACK or NACK for the packet so that you can (with a little software
effort) make a zero-overhead transport protocol.
Jack Bonn
This is a multi-part message in MIME format.
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Werner Huerttlen wrote:
> Jack Bonn <ja...@swlabs.com> wrote:
>
> >o Masters could originate communication to either other masters or slaves
> how is this done ?
This is from memory (standard disclaimers apply):
Masters only talk when they have the token. Slaves speak when spoken to.
Token goes around the low addresses (modulo 32). These are the addresses that
belong to the masters.
> >o Token bus over RS-485
> how is the time-out set ? How do you manage if the master is lost,
> e.g. powerer down.
Each node verifies that the token is acknowledged whenever he sends it. If it
is not acknowledged (after some retries), a poll for master is sent to see who
is next.
> >o Dynamic configuration, i.e. nodes could come and go at will
> How does this work ?
Master nodes come up and listen for a poll for master. They have time slots
they can respond in, depending on the difference between their address and the
node doing the poll for master. The time to send and acknowledge a poll for
master is 10 times the difference between my address and the sender of the
poll for master (modulo 32). The nearer you are to the address of the sender
of the poll for master, the sooner you get to reply. If a node closer to the
node doing the poll for master replies, forget it, wait for a node closer to
you to poll for master.
Each active master sends polls for master at a programmable rate, but are
guaranteed to occur often enough to eventually let every new master get linked
into the master list.
> I am in designing a multi-master RS485 network. But actually found
> out, collision is not allowed, because of chip-design. Possibly
> token-passing is a fine way, but I assume very complicated.
Link level was less than two man-months to implement. It was over 4 years
ago, so I could be a little off.
--
Jack Bonn <> Software Design Labs, Inc.
ja...@swlabs.com +1(847)526-1337
Ignorance and arrogance do a pathetic duet
--------------4927D8AD81D769DC157A480B
Content-Type: text/x-vcard; charset=us-ascii; name="vcard.vcf"
Content-Transfer-Encoding: 7bit
Content-Description: Card for Jack Bonn
Content-Disposition: attachment; filename="vcard.vcf"
begin: vcard
fn: Jack Bonn
n: Bonn;Jack
org: Software Design Labs, Inc
adr: 26985 W. Fenview;;;Barrington;IL;60010;USA
email;internet: ja...@swlabs.com
title: Consultant
tel;work: +1 847 526 1337
tel;fax: +1 847 526 1311
x-mozilla-cpt: ;0
x-mozilla-html: TRUE
version: 2.1
end: vcard
--------------4927D8AD81D769DC157A480B--
Dr John Tankard
Dear Harold
I looked up LonTalk, which looks very good, just what I need but the
price of the development systems is vey high $19,999 for LonBuilder
Developer kit and $4,495 for LonWorks NodeBuilder, do you know of any
third party development systems that are cheaper ?
Best Regards
John
Joe Snyder
jtan...@tankard.com (Dr John Tankard) wrote:
>... LonTalk,...>price of the development systems is vey high $19,999 for LonBuilder
>Developer kit and $4,495 for LonWorks NodeBuilder, do you know of any
>third party development systems that are cheaper ?
And wait until you see what they want for support fees!! In addition
to about $40K for a usable LonBuilder setup, I paid $4K for a year
of support during which I asked three simple questions, none of which
was their tech support able to answer. We ended up developing a much
cheaper (free tools), faster and non-proprietary ethernet solution.
Joe Snyder
Dr John Tankard
jjsn...@mindspring.com (Joe Snyder) wrote:
>
>And wait until you see what they want for support fees!! In addition
>to about $40K for a usable LonBuilder setup, I paid $4K for a year
>of support during which I asked three simple questions, none of which
>was their tech support able to answer. We ended up developing a much
>cheaper (free tools), faster and non-proprietary ethernet solution.
>
>Joe Snyder
Joe are you saying that I am on a looser without a LonBuilder (ie
NodeBuilder is no good) if so I think I shall have to look in another
direction $20k is just out of the question
Joe Snyder
jtan...@tankard.com (Dr John Tankard) wrote:
>jjsn...@mindspring.com (Joe Snyder) wrote:
>>
>>And wait until you see what they want for support fees!! In addition
>>to about $40K for a usable LonBuilder setup, I paid $4K for a year
>>of support during which I asked three simple questions, none of which
>>was their tech support able to answer. We ended up developing a much
>>cheaper (free tools), faster and non-proprietary ethernet solution.
>Joe are you saying that I am on a looser without a LonBuilder (ie
>NodeBuilder is no good) if so I think I shall have to look in another
>direction $20k is just out of the question
Not a loser, just poor. The LonBuilder is a complete network
development and simulation environment, while the NodeBuilder just
allows you to develop a node to be attached to a LonWorks network.
If you go the NodeBuilder route, you will also have to buy
configuration software to actually attach your nodes together in a
network. At the last time that I looked at LonWorks, the configurator
was extremely primitive, but I think that there are third party
products now that implement this fairly well. Some of them may have
the node development tools built in as well. Check the Echelon &
Motorola web sites for third party vendors.
My principal gripe with Echelon is that they were very long on
marketing hype and extremely short on delivering usable software
tools. The LonBuilder environment was (and maybe still is) a very
primitive DOS based IDE similar to Borland's, but without the features
you would expect of an integrated editor/project manger software
program. Mark, cut and paste, for instance, was out of the question.
The build process was a series of going back and forth between
unrelated screens to set a multitude of chip parameters.
Fred Graham, or Harold Rabbie, maybe you can respond to my comments
with some evidence of upgraded tools.
Joe Snyder
jtan...@tankard.com (Dr John Tankard) wrote:
>jjsn...@mindspring.com (Joe Snyder) wrote:
>>much very useful stuff.
>>
>Just so I am clear, if I got Node Builder does this mean that I could
>not put together a simple network of several simple nodes ?
I not sure what the situation is now, so it would be best to check
this out directly with Echelon.
-Joe Snyder
Dr John Tankard
jjsn...@mindspring.com (Joe Snyder) wrote:
>much very useful stuff.
>
Just so I am clear, if I got Node Builder does this mean that I could
not put together a simple network of several simple nodes ?
du...@dukelec.com
> I have seen many examples of a single master multiple slaves RS485, my
> question is :- is there any reason why there cannot be multiple
> masters on the wire ?
>
David Brown
While it is good of you to try to help the OP, I doubt if he has been
sitting for over 20 years waiting for an answer!
Tauno Voipio
to have invented years ago. There have been such protocols
in the RS-485 bus for years before his invention.
--
-TV