View previous topic :: View next topic |
Author |
Message |
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Wed Apr 07, 2010 7:44 pm Post subject: MAX487 Protection? |
|
|
Hi all,
for home application (no industry) of about 200m cable length (twisted and shielded, ground included) and <= 100 Members, I want to use the MAX487E as RS485 transceiver.
The MAXIM homepage states "The MAX481E, MAX483E, MAX485E, MAX487E- MAX491E, and MAX1487E are low-power transceivers for RS-485 and RS-422 communications in harsh environments. Each driver output and receiver input is protected against ±15kV electro-static discharge (ESD) shocks, without latchup."
Does anyone has experiences in terms of protection with this IC?
Do I need additional external protection devices (e.g. TVS) or not?
Many thanks and best regards,
T.D. |
|
Back to top |
|
|
ex4
Joined: 13 Jan 2006 Posts: 1062 Location: indonesia
|
|
Back to top |
|
|
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Fri Apr 09, 2010 11:49 am Post subject: |
|
|
Hi,
you're certainly right, that isolated devices are the best, but:
1) the application has no further ground available (one GND for supply AND ref)
2) I didn't find a isolated devices which can be used with more than 64 members
regards,
T.D. |
|
Back to top |
|
|
ex4
Joined: 13 Jan 2006 Posts: 1062 Location: indonesia
|
Posted: Fri Apr 09, 2010 12:30 pm Post subject: |
|
|
Quote: | 1) the application has no further ground available (one GND for supply AND ref) |
see ISO COM1 and ISO COM2, thats the common isolated point
Quote: | 2) I didn't find a isolated devices which can be used with more than 64 members
|
please read this carefully
http://www.maxim-ic.com/quick_view2.cfm/qv_pk/1020
it is clearly stated there that the max device on bus is 128, which is clearly more than you need |
|
Back to top |
|
|
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Fri Apr 09, 2010 1:06 pm Post subject: |
|
|
Hi ex4,
maybe I was not clear enough:
The members are connected to each other via a 4-wire twisted pair
cable. I use one pair for the communication and the other for supply.
Since there is only one GND for communication and data processing,
the isolation won't help.
To stick to the page you mentioned: COMx and GNDx will be the same.
You're right 128 members are possible - I overlooked it, sorry.
Another point I chose MAX487 was the slew rate limitation. The appl.
is rather slow (9.6kBd) and I did not want to run into EM trouble due
to the steep signal edges. |
|
Back to top |
|
|
ex4
Joined: 13 Jan 2006 Posts: 1062 Location: indonesia
|
Posted: Fri Apr 09, 2010 1:13 pm Post subject: |
|
|
Quote: | The members are connected to each other via a 4-wire twisted pair
cable. I use one pair for the communication and the other for supply.
Since there is only one GND for communication and data processing,
the isolation won't help.
To stick to the page you mentioned: COMx and GNDx will be the same. |
oh my god, 4wire twisted shielded pair cable and 200m length with centralize power supply if i got it right,how much voltage and current running on the cable? and is it enough until the last client(on the other end of the cable)? |
|
Back to top |
|
|
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Fri Apr 09, 2010 1:31 pm Post subject: |
|
|
Certainly good questions.
You're 100% right. The power is more or less centralized, so it's 100m to each side. Since the power is 12V/max. 1A per side and the members are regulated to 5V, I should be fine with up to 6V voltage drop, meaning 6Ohms/100m. So I'll use 2x2x0.5mm˛ or even 2x2x0.75mm˛.
(This is worst case. Since the members are spread over the distance and not only at the end,the scenario gets a lot more easy.)
I did some tests with normal 200m cable (0.6mm diameter only -> 0.3mm˛!!) for the communcation. Since I've only 9.6kBd this was no issue. But certainly, the supply voltage drop over this distance is a point to be reflected. This is why I stick to the bigger diameter now. |
|
Back to top |
|
|
ex4
Joined: 13 Jan 2006 Posts: 1062 Location: indonesia
|
Posted: Fri Apr 09, 2010 1:43 pm Post subject: |
|
|
certainly the power supply is the main part that need consideration. bigger is better
as far as i can tell, using simple rs485 driver(74176 for example), is more than enough, even for industrial standard(from my own experience). although not so safe,hehehehe
6V on 7805 will certainly not make the output near 5V, so take care on that |
|
Back to top |
|
|
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Fri Apr 09, 2010 2:00 pm Post subject: |
|
|
Thanks for sharing your experience.
Do you mean with "simple RS485" device, that you did not spend any
additional external protection? Don't be shocked, I won't terminate the bus, since I'm using low data rate and it's a LOW power application. So I should reach the rails - at least after time x
On the other hand there is no 120Ohm resistor which "clamps" any spike. |
|
Back to top |
|
|
ex4
Joined: 13 Jan 2006 Posts: 1062 Location: indonesia
|
Posted: Fri Apr 09, 2010 2:13 pm Post subject: |
|
|
yup, no additional protection
but termination over long wire is very very suggested, more sustain to any small interference, even at low speed
ok, just design and test it |
|
Back to top |
|
|
T.D.
Joined: 18 Feb 2010 Posts: 14
|
Posted: Fri Apr 09, 2010 2:58 pm Post subject: |
|
|
Many thanks, that's good news.
Certainly I will attach an oscilloscope and check, whether drops and/or
communication signals get in bad shape.
Have a nice weekend,
T.D. |
|
Back to top |
|
|
|