A8N-E comm port inop.

[Howard Delman]

I just built up a new system with an A8N-E, Athlon64 3800, &
WinXpProSr2. I cannot get the comm port to work. I use it for various
pieces of electronic hardware, and none can communicate with it.

As the simplest test, I've connected the machine to a single board
computer with a null modem cable. When I start Hyperterm, and try to
connect to the SBC, CPU usage goes to 100%, and my system is almost
totally unresponsive until I can get hyperterm to disconnect. If I move
the cable from this new machine to my old Win98 machine, all works fine.
The old computer and the SBC happily exchange information. I have
configured the com port (com1) on the new machine exactly the same as on
the old machine.

The only other com port on the new machine is a modem, which is assigned
to com4.

Any ideas?

Howard Delman

 

[Paul]

I just built up a new system with an A8N-E, Athlon64 3800, &
WinXpProSr2. I cannot get the comm port to work. I use it for various
pieces of electronic hardware, and none can communicate with it.

As the simplest test, I've connected the machine to a single board
computer with a null modem cable. When I start Hyperterm, and try to
connect to the SBC, CPU usage goes to 100%, and my system is almost
totally unresponsive until I can get hyperterm to disconnect. If I move
the cable from this new machine to my old Win98 machine, all works fine.
The old computer and the SBC happily exchange information. I have
configured the com port (com1) on the new machine exactly the same as on
the old machine.

The only other com port on the new machine is a modem, which is assigned
to com4.

Any ideas?

Howard Delman

On some other boards, Asus uses Texas Instruments GD75232
RS-232 interface chips. These are old style chips, that
run from +12V and -12V.

http://www-s.ti.com/sc/ds/gd75232.pdf

Is there any possibility that your -12V is out of whack ?

The fact that the new power supplies have no -5V any more,
should not affect a chip like the GD75232.

Also, I notice in the manual, that the A8N-E package
does not come with a serial port adapter plate for a
PCI slot. With third party adapter assemblies, there
are two wiring patterns for the nine pins. The result
is, not all adapter plates will work.

Here is a previous effort to debug an Asus COM port:

http://home.att.net/~Tom.Horsley/p4c800e.html

Paul

 

[Mark A]

Also, I notice in the manual, that the A8N-E package
does not come with a serial port adapter plate for a
PCI slot. With third party adapter assemblies, there
are two wiring patterns for the nine pins. The result
is, not all adapter plates will work.
Paul

My AN8-E (an early model) came with serial port adapter plate for a PCI slot
opening in the rear of the case. I have it hooked up to a 56 kbs US Robotics
modem that I occasionally use for faxing. It works perfectly.

It is described on page 1-27 of the user manual.

Maybe the OP did not hook it up properly to the motherboard header.

 

[Paul]

My AN8-E (an early model) came with serial port adapter plate for a PCI slot
opening in the rear of the case. I have it hooked up to a 56 kbs US Robotics
modem that I occasionally use for faxing. It works perfectly.

It is described on page 1-27 of the user manual.

Maybe the OP did not hook it up properly to the motherboard header.

OK. I see an adapter pictured here. So that is not the problem.

http://www.newegg.com/Product/ShowI...IA nForce4 Ultra ATX AMD Motherboard - Retail

Paul

 

[Howard Delman]

On some other boards, Asus uses Texas Instruments GD75232
RS-232 interface chips. These are old style chips, that
run from +12V and -12V.

http://www-s.ti.com/sc/ds/gd75232.pdf

Is there any possibility that your -12V is out of whack?

I've had a 'scope on the Txd and Rxd lines, and they are wiggling at the
right baud rate, and their levels are reasonable. I know the port is
connected, and I know it is being written by software. But clearly...
something is severely messed up.

Howard

 

[Paul]

I've had a 'scope on the Txd and Rxd lines, and they are wiggling at the
right baud rate, and their levels are reasonable. I know the port is
connected, and I know it is being written by software. But clearly...
something is severely messed up.

Howard

Could it be a serial port interrupt happening over and over
again ? Would a flow control RS-232 signal stuck in the
wrong state do that ?

Paul

 

[Howard Delman]

Could it be a serial port interrupt happening over and over
again ? Would a flow control RS-232 signal stuck in the
wrong state do that ?

Paul

I have flow control set to "none."

A flailing interrupt would certainly overwhelm the CPU, but that would
be a hardware fault. There is no data being sent, and the various
control lines are inactive, so there is nothing to generate an interrupt.

If you recall my test configuration, I have the comm port connected with
a null modem cable to an rs232 port on a single-board-computer. The SBC
is idling. There is nothing being sent between the machines. But
hyperterm is clearly being overwhelmed by something, and using 99% of
CPU capacity.

Since this is a brand new computer, it is possible that there really is
a hardware fault. But my experience suggests that is very very
unlikely. And I would dread having to rebuild the machine, and to
convince Asus to exchange my board.

Howard