Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

FaxModem Com Port Woes

4 views
Skip to first unread message

em...@owlcroft.com

unread,
Apr 5, 2006, 7:58:19 PM4/5/06
to
Not long ago, I moved to an entire new hardware system. I have a PCI
modem card that I am trying to make work with FaxWorks, and am unable
to do so.

The modem card is a USR 2976, which *does* have an onboard controller
(it is, I believe, essentially an OEM USR 5610). I have set both
onboard mobo COM ports to "Disabled" in the hardware BIOS.

Running pci.exe gives the following report (essential points only):

Bus 1 (PCI), Device Number 8, Device Function 0
Vendor 12B9h 3COM Corp, Modem Division (Formerly US Robotics)
Device 1008h USR 56k Internal Modem
Command 0001h (I/O Access)
Status 0210h (Has Capabilities List, Medium Timing)
Revision 01h, Header Type 00h, Bus Latency Timer 00h
Self test 00h (Self test not supported)
PCI Class Simple Communication, type Serial - 16550
Subsystem ID 00AA12B9h USR 56k Internal Voice Modem (Model 2976)
Subsystem Vendor 12B9h 3COM Corp, Modem Division (Formerly US
Robotics)
Address 0 is an I/O Port : A800h..A807h
System IRQ 11, INT# A

Running RMVIEW.exe shows:

Adapter: SERIAL_1 Serial Controller
Device Type: COMM-SERIAL Bus/Width: PCI 16 BIT
I/O = 0XA800 Len = 8 Flg = MULTIPLEXED Addr Lines = 16
IRQ Level = 11 PCI Pin = A Flg = SHARED

So we seem well agreed that the port is A800h and the IRQ is 11.

In CONFIG.SYS, I have these two lines (not actually adjacent);

DEVICE=C:\OS2\BOOT\COM.SYS /F
DEVICE=E:\PROGRAMS\WORKWITH\FAXES\Fmd.sys (2,A800,11)

(I read somewhere that the /F parameter forces COM.SYS to read all the
actual hardware settings.)

In the FaxWorks Settings Notebook, at "Ports", I find--

Port 2 (io=A800 irq=11) -- Send/Receive

--and no other ports showing. But as soon as I click on "Receive all
calls", I get a popup saying:

Fax Device FMD2 is unavailable. The COM device may be in use by
another program.

I *have* also tried explicitly giving the port to COM.SYS with--

DEVICE=C:\OS2\BOOT\COM.SYS (2,A800,11)
DEVICE=E:\PROGRAMS\WORKWITH\FAXES\Fmd.sys (2,A800,11)

--with no different results.

RMVIEW does not show any other use of IRQ 11, but pci.exe reports that
3 devices are sharing it (modem, SCSI adaptor board, video card).

Could I just have a bad modem card, or is there something I'm missing
here? I *really* need to get fax working on this machine.

Thanks for any and all help.

--
Cordially,
Eric Walker
em...@owlcroft.com

William L. Hartzell

unread,
Apr 5, 2006, 9:17:43 PM4/5/06
to
Sir:

Try removing the com.sys driver and using the fmd.sys driver alone.
--
Bill
Thanks a Million!

Paul Ratcliffe

unread,
Apr 6, 2006, 4:24:29 AM4/6/06
to
On Wed, 05 Apr 2006 20:17:43 -0500, William L. Hartzell <wlhar...@comcast.net> wrote:

>> Running RMVIEW.exe shows:
>>
>> Adapter: SERIAL_1 Serial Controller
>> Device Type: COMM-SERIAL Bus/Width: PCI 16 BIT
>> I/O = 0XA800 Len = 8 Flg = MULTIPLEXED Addr Lines = 16
>> IRQ Level = 11 PCI Pin = A Flg = SHARED
>>
>> So we seem well agreed that the port is A800h and the IRQ is 11.
>>
>> In CONFIG.SYS, I have these two lines (not actually adjacent);
>>
>> DEVICE=C:\OS2\BOOT\COM.SYS /F
>> DEVICE=E:\PROGRAMS\WORKWITH\FAXES\Fmd.sys (2,A800,11)
>>

>> In the FaxWorks Settings Notebook, at "Ports", I find--
>>
>> Port 2 (io=A800 irq=11) -- Send/Receive
>>
>> --and no other ports showing. But as soon as I click on "Receive all
>> calls", I get a popup saying:
>>
>> Fax Device FMD2 is unavailable. The COM device may be in use by
>> another program.
>>
>> I *have* also tried explicitly giving the port to COM.SYS with--
>>
>> DEVICE=C:\OS2\BOOT\COM.SYS (2,A800,11)
>> DEVICE=E:\PROGRAMS\WORKWITH\FAXES\Fmd.sys (2,A800,11)
>>
>> --with no different results.
>>
>> RMVIEW does not show any other use of IRQ 11, but pci.exe reports that
>> 3 devices are sharing it (modem, SCSI adaptor board, video card).
>

> Try removing the com.sys driver and using the fmd.sys driver alone.

Try removing the fmd.sys driver and using the com.sys driver alone.

The version of PMFax for OS/2 which I use is 3.2h.Lite and it doesn't rely
on fmd.sys any more - it was always a stupid idea for applications to rely
on their own dumb device drivers to provide what should be built in to
standard drivers.

The other thing the OP could do is just test the COM port with some other
comms. type program to make sure it is working properly.

Aidan Grey

unread,
Apr 6, 2006, 8:21:13 AM4/6/06
to

( Snip )

You should check that your COM ports are correctly installed.

Do you have any other COM ports in your system? Add-in cards
usually start at COM3. Does anything change if you try assigning
this add-in card to COM3?

What results do you get from a command such as "MODE COM2:"?
What results do you get if you try the same command for a COM
port that does not exist, such as COM5?

If you have a copy of HyperTerminal on you system, test the
COM port with that. The command "AT" should get a reply back
from the modem.


Aidan Grey


em...@owlcroft.com

unread,
Apr 25, 2006, 3:49:11 PM4/25/06
to
Sorry to be so long in answering the replies.

The modem will both dial and answer, so I presume the COM port is
communicating.

Looking at a VOUT log, it appears that the difficulty is a failure to
train (that is the final error status); but that may simply reflect a
failure or inability of the modem to talk to or understand the other
modem. Here is what nonexpert I think is the crux of the matter:

10/14 14:36:08.53 [2] put_command_19 'AT+FDT\r'
10/14 14:36:08.73 [2] wait_for 'CONNECT\r\n' 60sec
10/14 14:36:08.90 [2] <\r\n+FCS:> !
10/14 14:36:08.91 [2] <1,5,0,2,0,0,0,0\r> @
10/14 14:36:08.91 [2] dcs info 1 5 0 2 0 0 0 0
10/14 14:36:08.91 [2] fine capability
10/14 14:36:08.91 [2] pad = 0ms
10/14 14:36:08.91 [2] negbps = 14400
10/14 14:36:10.01 [2] <\n>
--------------------------------------------------------------
10/14 14:36:14.73 [2] <\r\n+FCS:1,4,0,2,0,0,0,0\r\n>
10/14 14:36:21.62 [2] <\r\n+FCS:1,4,0,2,0,0,0,0\r\n>
10/14 14:36:27.47 [2] <\r\n+FCS:1,3,0,2,0,0,0,0\r\n>
10/14 14:36:33.23 [2] <\r\n+FCS:1,3,0,2,0,0,0,0\r\n>
10/14 14:36:37.89 [2] <\r\n+FCS:1,2,0,2,0,0,0,0\r\n>
10/14 14:36:45.04 [2] <\r\n+FHS:> !
10/14 14:36:45.04 [2] <25\r> @
10/14 14:36:45.04 [2] tiff send failed
10/14 14:36:45.04 [2] hangup: Train Fail (37)
10/14 14:36:45.04 [2] ***** final status: Error_Train_Fail, 406 ints
10/14 14:36:45.04 [2] FmlDone ||n/a|39|
10/14 14:36:45.04 [2] FmlClose 'FMD2'
10/14 14:36:45.05 [2] <\n\r\nOK\r\n>

I *think* that if all were working right, that where the line appears
(my add) would be a CONNECT; instead, a steady lot of speed-downgrade
FCS's all fail.

All else up to that looks ok as best I can judge (which may not be very
well).

The modem settings in the FaxWorks Notebook are:

Class 1
Check TCF Signal
Use HW FIFO (16550A)
Special Command = ATE0Q0S0=0V1X1&C1&D2

(Yes, it has a 16550A, and is supposedly a Class 1/Class 2 modem.)

Any ideas (besides hardware defect on the modem)?

Thanks again!

Paul Ratcliffe

unread,
Apr 25, 2006, 6:11:23 PM4/25/06
to
On 25 Apr 2006 12:49:11 -0700, em...@owlcroft.com <em...@owlcroft.com> wrote:

> Any ideas (besides hardware defect on the modem)?

It's a hardware problem. The modems or the line. Nothing to do with software
or configuration.

Barry Mann

unread,
Apr 25, 2006, 10:37:43 PM4/25/06
to
In <1144281499.6...@j33g2000cwa.googlegroups.com>, on 04/05/06

at 04:58 PM, em...@owlcroft.com said:

>Not long ago, I moved to an entire new hardware system. I have a PCI
>modem card that I am trying to make work with FaxWorks, and am unable
>to do so.

Your problem could easily be the wrong modem set-up string.

In my case I could never resolve the interrupt problems with the modem
card (ActionTech) I could move the modem to another slot or logically
ask it to use another port. Every time I moved something, another card
would get in the way. Probably there is a motherboard BIOS issue.
Finally I caved and attached a ZOOM RS-232 modem that works fine. (But,
as you can see, I'm still grumpy.)

-----------------------------------------------------------
spam: sp...@uce.gov
wordgame:123(abc):<14 9 20 5 2 9 18 4 at 22 15 9 3 5 14 5 20 dot 3 15
13> (Barry Mann)
[sorry about the puzzle, spammers are ruining my mailbox]
-----------------------------------------------------------

Harald Pollack

unread,
Apr 27, 2006, 7:17:20 AM4/27/06
to
Servus em...@owlcroft.com!

e> The modem will both dial and answer, so I presume the COM
e> port is communicating.

Yes, it is.

e> Looking at a VOUT log, it appears that the difficulty is a
e> failure to train (that is the final error status);

Also, yes

e> but that may simply reflect a failure or inability of the modem to talk to
e> or understand the other modem.

The problem is (or might be) that V.17 (14400bps) is selected as starting
speed.


e> 10/14 14:36:08.53 [2] put_command_19 'AT+FDT\r'
e> 10/14 14:36:08.73 [2] wait_for 'CONNECT\r\n' 60sec
e> 10/14 14:36:08.90 [2] <\r\n+FCS:> !
e> 10/14 14:36:08.91 [2] <1,5,0,2,0,0,0,0\r> @
e> 10/14 14:36:08.91 [2] dcs info 1 5 0 2 0 0 0 0
e> 10/14 14:36:08.91 [2] fine capability
e> 10/14 14:36:08.91 [2] pad = 0ms
e> 10/14 14:36:08.91 [2] negbps = 14400

V.17, 14400bps is negotiated first...

e> 10/14 14:36:14.73 [2] <\r\n+FCS:1,4,0,2,0,0,0,0\r\n>
e> 10/14 14:36:21.62 [2] <\r\n+FCS:1,4,0,2,0,0,0,0\r\n>

Trainings fails at 12000bps

e> 10/14 14:36:27.47 [2] <\r\n+FCS:1,3,0,2,0,0,0,0\r\n>
e> 10/14 14:36:33.23 [2] <\r\n+FCS:1,3,0,2,0,0,0,0\r\n>

... fails at 9600bps

e> 10/14 14:36:37.89 [2] <\r\n+FCS:1,2,0,2,0,0,0,0\r\n>

... fails at 4800bps.

e> 10/14 14:36:45.04 [2] <\r\n+FHS:> !
e> 10/14 14:36:45.04 [2] <25\r> @
e> 10/14 14:36:45.04 [2] tiff send failed
e> 10/14 14:36:45.04 [2] hangup: Train Fail (37)

Well, that's right. But why?

e> The modem settings in the FaxWorks Notebook are:

e> Class 1

Not at all! Modem uses Class 2.0

e> Check TCF Signal
e> Use HW FIFO (16550A)
e> Special Command = ATE0Q0S0=0V1X1&C1&D2

Meaningless in fax mode.

e> (Yes, it has a 16550A, and is supposedly a Class 1/Class 2 modem.)

e> Any ideas (besides hardware defect on the modem)?

Set starting speed to 9600 (V.29) which will fall back to 2400 in worst case.

Starting with V.17 (14400pbs) has the misbenefit, that the whole fall back is
performed in V.17. Starting with V.29 (9600) has the benefit, that after 7200
(V.29) fallback is done to V.27ter (4800 and 2400).

Herzliche Gruesse, Harald

-+- Message created on Thursday April, 27 2006 14:24:40 EDT

0 new messages