Long distance comms

Hi

I have a requirement to communicate between a slave device and a master where the distance between them is approximately 3km. I don't need high speed comms, 10-20kbaud is sufficient. Unfortunately I don't have the luxury of being able to use repeaters, fibre optics or wireless. If RS422 was spec'd beyond 1.2km then that would be ideal.

It seems that CAN will work up to 6km at bit rates of 10kbaud but I'd rather not have the complexity required to implement CAN especially as I only need point-to-point comms. I was thinking of using CAN transceivers & connecting the Rx/Tx pins to UARTs at either end.

Does anyone have any experience of such an application or have any suggestions ?

Thanks Dave

Reply to
Daveb
Loading thread data ...

Some of the metering busses are designed for this kind of rate/distance. One that comes to mind has Master -> slave comms voltage modulated (12-24V) and slave to master comms current modulated . Each slave needs to have a constant current drain for this to work.

-Andrew M

Reply to
Andrew M

Daveb scrobe on the papyrus:

A current loop with opto isolators should do the trick. You might be limited to a lower baud rate.

--
John B
Reply to
John B

Andrew

Thanks for the reply. I searched for metering busses & the only one I can find is M-Bus which is restricted to 350m without repeaters. Do you know of any which go up to 3km ?

Thanks Dave

Reply to
Daveb

Using current loop would also be my first suggestion with a sufficient voltage headroom (24-60 V) in the constant current generator.

However, the RS-422/485 is specified for 1200 m at 100 kbit/s for a specific wire gauge with the terminating resistors present. If you can use a thicker wire and possibly get away with the terminating resistor due to the low data rate, the link might be doable with standard RS-422/485 hardware.

Due to ground potential differences (especially during thunderstorms), it is essential that the systems have a galvanic isolation regardless which technology is used.

Paul

Reply to
Paul Keinanen

Look up DSL line drivers.

donald

Reply to
Donald

Thanks for all the suggestions.

I realise my requirement isn't a common one but I'm hoping to find a comms standard I can use that will save time in designing/testing the physical implementation from scratch. I've seen mentioned in several overviews of CAN that it will work up to 6km @ 10kbps but can't find any specs that confirm this. Could anyone point me in the direction of a spec that details this ?

Thanks all Dave

Reply to
Daveb

There may not be one. CAN is primarily a *protocol*, with more than one physical line specification choice to run it over. What limits the length of a CAN bus depends on the baudrate *and* the physical medium choosen. The 6km @ 10kbps is probably a protocol limit --- basically the speed-of-light barrier applied to bit timing requirements of CAN. That doesn't automatically guarantee that a plain vanilla two-wire CAN bus transmitter will actually be able to drive a line of that length.

IOW, you may be chasing a red herring.

--
Hans-Bernhard Broeker (broeker@physik.rwth-aachen.de)
Even if all the snow were burnt, ashes would remain.
Reply to
Hans-Bernhard Broeker

AFAIK that is the case -- when you read the CAN document itself it's pretty adamant about _not_ being a spec for the physical layer. There _is_ an SAE spec for a physical layer which uses 1/2 VCC on both wires for 1 and pulls the wires to 0V and VCC for 0 -- but I would expect that it wouldn't be any better than RS-485 for really long distances.

--
Tim Wescott
Wescott Design Services
 Click to see the full signature
Reply to
Tim Wescott

I would care less about the specified lenght of RS422, which is 1500m, than about common mode issues. With sufficiently low baud rate, the RS422 in point-to-point mode might do even a bit longer than specified. Else a closer look at the attenuation of the cable might be helpful.

Rene

--
Ing.Buero R.Tschaggelar - http://www.ibrtses.com
& commercial newsgroups - http://www.talkto.net
Reply to
Rene Tschaggelar

We used to employ both 60ma current loop and also limited distance modems for this sort of application; why not consider off-the-shelf modems and cheap FXS and FXO interfaces?

Michael Grigoni Cybertheque Museum

Reply to
msg

Buy a couple of 28 or 56KBaud POTS modems. (Do they still make those??)

Mark Borgerson

Reply to
Mark Borgerson

Hi all,

It looks like a current loop is the way to go. Opto isolating might not be so easy as I have elevated temperature to contend with (105degC). A standard modem solution is likely to be unworkable due to space constraints in the slave device.

Thanks again for all the suggestions! Dave

Reply to
Daveb

Dave,

TI has some modern "optimized" RS422/485 transceivers (SN65HVD05,06,07,20,21,22), which have controlled slewrate and are well suited for long distance applications. I am using the HVD21 in a 1MBaud,

200m multidrop application. Extrapolating from that 6km at 10kbaud shall work. IMO the TI RS485 chips are as good or better than the CAN approach (the CAN transceiver advantage is that their slew rate can be "programmed" by an external resistor), otherwise using CAN transceivers for UART protocol should readily work, but noise margin is smaller, as their is no voltage on the line in the recessive state.

Regards

Thomas

Reply to
Thomas Gallenkamp

P.S.: The DC resistance of the wire pair is critical. A 24AWG pair has about

1.2kOhm loop resistance at 6km. Differential voltage at the receiver will be some 300mV, which is still satisfactory.
Reply to
Thomas Gallenkamp

Hello Dave,

If isolation is a requirement look into toroids. You'll lose the DC component but if you use a protocol that only listens to the transitions it should work. Hysteresis receivers, for example.

Regards, Joerg

formatting link

Reply to
Joerg

Thomas

Thanks for the tip. Unfortunately these devices are 85degC max & I need 105degC+

Dave

Reply to
Daveb

Hello Dave,

You could roll your own receivers using mil/automotive grade chips. For example there are versions of the LM139 that go to 125degC.

If differential noise immunity is a concern at such low signal levels you might want to look into using coax. You could pick whatever impedance comes in a rugged and reasonably priced variety.

Regards, Joerg

formatting link

Reply to
Joerg

This is meaningless. Include context. Usenet articles need to stand by themselves, as there is no reason to expect that your reader has ever, or ever will, be able to read preceding articles. To implement proper quoting even on the broken Google interface to usenet, see my sig below. Please read the referenced URLs.

--
"If you want to post a followup via groups.google.com, don't use
 the broken "Reply" link at the bottom of the article.  Click on 
 Click to see the full signature
Reply to
CBFalconer

You're quite right. Thanks for pointing this out. Dave

Reply to
Daveb

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.