RS-485 Troubleshooting

Started by RahulS January 25, 2006
Hi,
I am using Rabbit development board which has separate ports for RS232
and Rs485. Right now i am using a single port both as RS232 as well as
RS485.
The port works fine for RS232 signals but for RS485 signals, the
differential output is not proper.I think there is some problem with
the differential voltages.
Please help.

"RahulS" <rahuls@kpitcummins.com> schreef in bericht 
news:1138186965.580233.127170@o13g2000cwo.googlegroups.com...
> Hi, > I am using Rabbit development board which has separate ports for RS232 > and Rs485. Right now i am using a single port both as RS232 as well as > RS485. > The port works fine for RS232 signals but for RS485 signals, the > differential output is not proper.I think there is some problem with > the differential voltages.
Well if you suspect the differential voltages, have you taken a look at hem? An oscilloscope would be the most appropriate tool. But what makes you think it is the voltage levels? If you are using a proper driver chip with proper supply voltages, and only correct load on the RS-485 bus, why then should the levels go wrong?
I am using SP483EN chip for RS485 signals. I used oscilloscope to test
the differential signals.
The signals were not differential in nature. When i short the CRO
ground and RS485+, it works.
When i monitor the waveform between RS485+ and GND, the waveform is
correct.
When i monitor the waveform between RS485- and GND, the waveform is
correct.But when  i monitor the waveform between RS485+ and RS485-, the
waveform is incorrect.
Could this be the case that the RS485 driver IC is not functioning
properly
Regards

"RahulS" <rahuls@kpitcummins.com> wrote in message
news:1138188073.191302.290970@z14g2000cwz.googlegroups.com...
> I am using SP483EN chip for RS485 signals. I used oscilloscope to test > the differential signals. > The signals were not differential in nature. When i short the CRO > ground and RS485+, it works. > When i monitor the waveform between RS485+ and GND, the waveform is > correct. > When i monitor the waveform between RS485- and GND, the waveform is > correct.But when i monitor the waveform between RS485+ and RS485-, the > waveform is incorrect.
How do you measure that? Not by clipping the probe on RS485+ and the ground clip on RS485- I hope?
> Could this be the case that the RS485 driver IC is not functioning
The only proper way of measuring this is by using a dual channel CRO, one channel on each RS485 pin and checking if the waveforms are eachothers' opposite. Meindert
I have already done that, after which i was able to conclude that
differential signals are not correct

"RahulS" <rahuls@kpitcummins.com> schreef in bericht 
news:1138189568.656826.165680@g47g2000cwa.googlegroups.com...
>I have already done that, after which i was able to conclude that > differential signals are not correct
Answers like this offer little hope of progress. Since you seem to need a bit of help in answering, please answer yes/no to each of the following: -) are you using a dual-beam (two-channel) oscilloscope? -) did you attach one channel to RS485+ ? -) did you attach the other channel to RS485-? If "yes" to all three, please describe the waveforms observed, and explain what you think incorrect about them.
"RahulS" <rahuls@kpitcummins.com> wrote in message
news:1138189568.656826.165680@g47g2000cwa.googlegroups.com...
> I have already done that, after which i was able to conclude that > differential signals are not correct
To whom are you replying now? Please leave a piece of the original text you are replying to, so everyone reading this thread, knows what you are talking about. If you are replying to me: what exactly was the difference between both signals? Can you make a simpla ASCII (character) drawing of that? Meindert
RahulS wrote:

> The port works fine for RS232 signals but for RS485 signals, the > differential output is not proper.I think there is some problem with > the differential voltages.
RS485 is almost idiot proof. If you aren't getting the right voltages at the outputs you've: - loaded the output with something that stops it working - stopped loading the output but buggered the chip - forgotten to turn the thing on (enable) - got the scope ground clipped to somewhere whappy - don't know what to look for or how to look for it. Paul Burke
"Paul Burke" <paul@scazon.com> schreef in bericht 
news:43pbdiF1ou81sU1@individual.net...
> RahulS wrote: > >> The port works fine for RS232 signals but for RS485 signals, the >> differential output is not proper.I think there is some problem with >> the differential voltages. > > RS485 is almost idiot proof. If you aren't getting the right voltages at > the outputs you've: > > - loaded the output with something that stops it working > - stopped loading the output but buggered the chip > - forgotten to turn the thing on (enable) > - got the scope ground clipped to somewhere whappy
all of these had already been suggested
> - don't know what to look for or how to look for it.
this one seems more and more obvious
On 2006-01-25, Meindert Sprang <mhsprang@NOcustomSPAMware.nl> wrote:

>> When i monitor the waveform between RS485- and GND, the waveform is >> correct.But when i monitor the waveform between RS485+ and RS485-, the >> waveform is incorrect.
RS-485 drivers are differential but not floating.
> How do you measure that? Not by clipping the probe on RS485+ and the ground > clip on RS485- I hope? > >> Could this be the case that the RS485 driver IC is not functioning > > The only proper way of measuring this is by using a dual > channel CRO, one channel on each RS485 pin and checking if the > waveforms are eachothers' opposite.
Or using a scope with a floating ground. Or using a differential scope probe. If you do a lot of RS-485 work, it can be a handy (albiet expensive) thing to have. -- Grant
On Fri, 27 Jan 2006 11:47:41 +0100, Meindert Sprang wrote:

>> we can see continuous rectangular waveforms ( on -off- on -off ) as i >> am sending data 1 and 0 alternately but continuously >> data can be seen on hyperterminal ( with 485-232 convertor in between) > > Good. If you can see data in hyperterminal, everything seems to be ok.
I agree with Meindert. If you're reading the data with HT, then you should be good to go. Having said that however, I recently had some reliability problems with a PCMCIA RS485 card, and needed to add 620 ohm biasing resistors from RS485+ to VCC and RS485- to ground in addition to a 120 ohm terminating resistor across the pair. Bob
"RahulS" <rahuls@kpitcummins.com> wrote in message
news:1138357530.405792.162570@g49g2000cwa.googlegroups.com...
> Three connections were tried as follows: > 1. CRO channel 1, > its signal pin connected to 485+ > its ground pin connected to ground > > we can see continuous rectangular waveforms ( on -off- on -off ) as i > am sending data 1 and 0 alternately but continuously > data can be seen on hyperterminal ( with 485-232 convertor in between)
Good. If you can see data in hyperterminal, everything seems to be ok.
> 2. CRO channel 1, > its signal pin connected to 485+ > its ground pin connected to ground > > similar ouput.
Typo? Do you mean 485+ or 485-?
> > 3.CRO channel 1, > its signal pin connected to 485+ > its ground pin connected to 485- > > no waveform observed ( coincides with reference line on CRO)
Logical. You short-circuit 485- to ground through the ground of the CRO.
> 4.CRO channel 1, > its signal pin connected to 485+ > its ground pin connected to 485- > its ground pin touched to 485+ ( indirectly it becomes case 1 )
I have no clou as to what you are doing now???? The only proper way to measure this is to connect CRO channel 1 to 485+, ground to ground and CRO channel 2 to 485- and ground to ground. Observe both waveforms simultaneaously on the screen, they should be in opposite phase and all the flanks matching.
> rectangular waveform observed, data can be seen on hyperterminal ( with > 485-232 convertor in between) > > please note that I am only using differential output and not ground for > communication of two devices
Which is bad, because the common mode voltage on RS485 may not exceed 7V. Always use a ground connection too. Meindert
karel wrote:
> "RahulS" <rahuls@kpitcummins.com> schreef in bericht > news:1138189568.656826.165680@g47g2000cwa.googlegroups.com... > >I have already done that, after which i was able to conclude that > > differential signals are not correct > > Answers like this offer little hope of progress. > Since you seem to need a bit of help in answering, > please answer yes/no to each of the following: > > -) are you using a dual-beam (two-channel) oscilloscope? > > -) did you attach one channel to RS485+ ? > > -) did you attach the other channel to RS485-? > > If "yes" to all three, please describe the waveforms observed, > and explain what you think incorrect about them.
Three connections were tried as follows: 1. CRO channel 1, its signal pin connected to 485+ its ground pin connected to ground we can see continuous rectangular waveforms ( on -off- on -off ) as i am sending data 1 and 0 alternately but continuously data can be seen on hyperterminal ( with 485-232 convertor in between) 2. CRO channel 1, its signal pin connected to 485+ its ground pin connected to ground similar ouput. 3.CRO channel 1, its signal pin connected to 485+ its ground pin connected to 485- no waveform observed ( coincides with reference line on CRO) 4.CRO channel 1, its signal pin connected to 485+ its ground pin connected to 485- its ground pin touched to 485+ ( indirectly it becomes case 1 ) rectangular waveform observed, data can be seen on hyperterminal ( with 485-232 convertor in between) please note that I am only using differential output and not ground for communication of two devices
Steve at fivetrees wrote:
> "Bob Stephens" <roberts@dcxchol.com> wrote in message > news:hewtwp16d2b6.1r7fp0inum465$.dlg@40tude.net... > > On 25 Jan 2006 03:02:45 -0800, RahulS wrote: > > > >> Hi, > >> I am using Rabbit development board which has separate ports for RS232 > >> and Rs485. Right now i am using a single port both as RS232 as well as > >> RS485. > >> The port works fine for RS232 signals but for RS485 signals, the > >> differential output is not proper.I think there is some problem with > >> the differential voltages. > >> Please help. > > > > Is it 2-wire or 4-wire RS485? If it's 2-wire (half duplex) make sure you > > are switching the transceiver at the proper time from transmit to receive. > > And, as I've said a few times, don't forget the ground reference (i.e. those > 2 or 4 signals are measured relative to a common 0V reference - miss this > out and fun will ensue. Not). > > Steve > (who is therefore in the habit of referring to 3- and 5-wire RS485 > architectures) > http://www.fivetrees.com
Three connections were tried as follows: 1. CRO channel 1, its signal pin connected to 485+ its ground pin connected to ground we can see continuous rectangular waveforms ( on -off- on -off ) as i am sending data 1 and 0 alternately but continuously data can be seen on hyperterminal ( with 485-232 convertor in between) 2. CRO channel 1, its signal pin connected to 485+ its ground pin connected to ground similar ouput. 3.CRO channel 1, its signal pin connected to 485+ its ground pin connected to 485- no waveform observed ( coincides with reference line on CRO) 4.CRO channel 1, its signal pin connected to 485+ its ground pin connected to 485- its ground pin touched to 485+ ( indirectly it becomes case 1 ) rectangular waveform observed, data can be seen on hyperterminal ( with 485-232 convertor in between) please note that I am only using differential output and not ground for communication of two devices
"Bob Stephens" <roberts@dcxchol.com> wrote in message 
news:hewtwp16d2b6.1r7fp0inum465$.dlg@40tude.net...
> On 25 Jan 2006 03:02:45 -0800, RahulS wrote: > >> Hi, >> I am using Rabbit development board which has separate ports for RS232 >> and Rs485. Right now i am using a single port both as RS232 as well as >> RS485. >> The port works fine for RS232 signals but for RS485 signals, the >> differential output is not proper.I think there is some problem with >> the differential voltages. >> Please help. > > Is it 2-wire or 4-wire RS485? If it's 2-wire (half duplex) make sure you > are switching the transceiver at the proper time from transmit to receive.
And, as I've said a few times, don't forget the ground reference (i.e. those 2 or 4 signals are measured relative to a common 0V reference - miss this out and fun will ensue. Not). Steve (who is therefore in the habit of referring to 3- and 5-wire RS485 architectures) http://www.fivetrees.com
On 25 Jan 2006 03:02:45 -0800, RahulS wrote:

> Hi, > I am using Rabbit development board which has separate ports for RS232 > and Rs485. Right now i am using a single port both as RS232 as well as > RS485. > The port works fine for RS232 signals but for RS485 signals, the > differential output is not proper.I think there is some problem with > the differential voltages. > Please help.
Is it 2-wire or 4-wire RS485? If it's 2-wire (half duplex) make sure you are switching the transceiver at the proper time from transmit to receive. Bob
RahulS wrote:
> > I am using SP483EN chip for RS485 signals. I used oscilloscope to > test the differential signals. The signals were not differential > in nature. When i short the CRO ground and RS485+, it works. When > i monitor the waveform between RS485+ and GND, the waveform is > correct. When i monitor the waveform between RS485- and GND, the > waveform is correct. But when i monitor the waveform between > RS485+ and RS485-, the waveform is incorrect.
This sounds as if the RS485 network is incorrectly terminated and biased. Always include adequate context in your replies. Even if using broken Google usenet interface, when you can follow the instructions below. -- "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 "show options" at the top of the article, then click on the "Reply" at the bottom of the article headers." - Keith Thompson More details at: <http://cfaj.freeshell.org/google/>
On 2006-01-25, Meindert Sprang <mhsprang@NOcustomSPAMware.nl> wrote:

>> When i monitor the waveform between RS485- and GND, the waveform is >> correct.But when i monitor the waveform between RS485+ and RS485-, the >> waveform is incorrect.
RS-485 drivers are differential but not floating.
> How do you measure that? Not by clipping the probe on RS485+ and the ground > clip on RS485- I hope? > >> Could this be the case that the RS485 driver IC is not functioning > > The only proper way of measuring this is by using a dual > channel CRO, one channel on each RS485 pin and checking if the > waveforms are eachothers' opposite.
Or using a scope with a floating ground. Or using a differential scope probe. If you do a lot of RS-485 work, it can be a handy (albiet expensive) thing to have. -- Grant
"Paul Burke" <paul@scazon.com> schreef in bericht 
news:43pbdiF1ou81sU1@individual.net...
> RahulS wrote: > >> The port works fine for RS232 signals but for RS485 signals, the >> differential output is not proper.I think there is some problem with >> the differential voltages. > > RS485 is almost idiot proof. If you aren't getting the right voltages at > the outputs you've: > > - loaded the output with something that stops it working > - stopped loading the output but buggered the chip > - forgotten to turn the thing on (enable) > - got the scope ground clipped to somewhere whappy
all of these had already been suggested
> - don't know what to look for or how to look for it.
this one seems more and more obvious
RahulS wrote:

> The port works fine for RS232 signals but for RS485 signals, the > differential output is not proper.I think there is some problem with > the differential voltages.
RS485 is almost idiot proof. If you aren't getting the right voltages at the outputs you've: - loaded the output with something that stops it working - stopped loading the output but buggered the chip - forgotten to turn the thing on (enable) - got the scope ground clipped to somewhere whappy - don't know what to look for or how to look for it. Paul Burke