Weird problem on rs232 vs USB-serial...

Steven Grunza grunza at
Mon Feb 23 19:23:25 UTC 2009

My first suggestion would be to use a Voltmeter to measure the Voltage
on the PC's COM1.  Assuming a DB9 connection you should see -5V or more
between pins 3 and 5.  If you have a DB25 then check between pins 2 and

Back in the old days (before Maxim's MAX232 and similar chips) it wasn't
unusual for a cheap RS-232 interface to "steal" negative Voltage power
from the other end of the connection.  This usually worked well unless
the data rate was too high or the other end was also trying to "steal"
negative Voltage power.  I know of some Tandy machines that used this

The USB part probably uses either a Maxim or other vendor's RS-232
interface powered from the 5V available on the USB interface.

That's it for my USD 0.02

>-----Original Message-----
>From: rtems-users-bounces at 
>[mailto:rtems-users-bounces at] On Behalf Of mike
>Sent: Monday, February 23, 2009 2:03 PM
>To: RTEMS Users
>Subject: Weird problem on rs232 vs USB-serial...
>Hi all,
>We have a Leon2-based board using the RTEMS leon2 BSP. The 
>host is x86 running CentOS 2.6.
>With a simple application that outputs to /dev/console for 
>stdio (printf):
>- with the board connected directly to COM1 of the host, no 
>output is seen
>- with the board connected to a USB<->serial adapter 
>(Keyspan), outputs and inputs were fine
>* This is with the same app initializing the UART to 38400, 
>8N1, no flow control.
>* On the host, we just use 'minicom' set for 38400, 8N1, no 
>flow control
>    (/dev/ttyS1 in the 1st case and /dev/ttyUSB0 in the 2nd case)
>What could possibly be the cause?
>Thanks for any comments/pointers.
> -mike-
>rtems-users mailing list
>rtems-users at

More information about the users mailing list