Corrupt input on UART (both UEXT1 and UEXT2)

[Roux]

Hi,

I have connected an inclinometer
www.beisensors.com/pdfs/Inclinometer_analog_heavy-duty_bei.pdf
to an UART port of A20 (MOD-RS232 connected to UEXT bys) at default
9600 8N1.

Instead of normal output

.....
X=-05.73;Y=+03.97
X=-05.79;Y=+04.01
.....

i get an output where most bytes are corrupted.
Hete is hexdump of it (could not find how to attach the file to the post):
0000000 58ff 2d3d 54c1 11e4 9a67 593b b6f5 4611
0000010 1746 cd35 98d4 58c1 2d3d 54c1 11e4 8d67
0000020 593b b6f5 4611 1746 e135 98d6 58c1 2d3d
0000030 54c1 11e4 9a67 593b b6f5 4611 1746 c135
0000040 98d4 58c1 2d3d 54c1 d1e4 8db3 593b b6f5
0000050 4611 1746 e535 98d4 58c1 2d3d 54c1 11e4
0000060 4667 9667 2b3d 34c1 51e4 ce23 0a0d 58ff
0000070 2d3d 54c1 11e4 9a67 593b b6f5 4611 1746
0000080 d535 98d4 58c1 2d3d 54c1 11e4 8d67 593b

Despite it preserves some bytes, the output is comletely useless.
The problem is reproducible with either UEXT connector.
The same device works perfectly with few PCs,
and the same A20 with few few USB-serial adapters.
Same UEXT ports and MOD-232 perfectly work with other serial
devices at various baud rates.


Any clue?

Thank you!
BR,
Rostislav

[JohnS]

Voltage levels / TTL/RS232 (may even need an inverter)?  # stop bits? grounding?

John

[dave-at-axon]

This sounds like a missing GROUND connection or the 8N1 setting is wrong.

How have you wired the connections from the inclinometer to the MOD-RS232?

[Roux]

Thank you all!

Dave-at-axon was right.  It should be 8E1 (finally got a proper manual from the
manufacturer).

What confused me, was that few other ports happily ignored parity by default, so they simply worked.