Failed to print V17 to FaxLab. Not passing ATD.

Good day

I'm testing a system to FaxLab to confirm reliability.  The system uses a
Class1 Davicom fax modem.  My tests only train at V17.

Sometime I have a problem printing a fax to FaxLab.  FaxLab answers, it
picks up the calling tone (CNG) then answers with CED.  It then proceeds
with NSF/CSI/DIS (I think it is phase A?).  From the logs below it does not
seem like the ATDT finished.

Nov 23 06:37:00.20: [20093]: SESSION BEGIN 000003921 1
Nov 23 06:37:00.20: [20093]: HylaFAX (tm) Version 5.5.9
Nov 23 06:37:00.20: [20093]: SEND FAX: JOB 1959 DEST 1 COMMID 000003921
DEVICE '/dev/ttyS1' FROM 'fax' USER root
Nov 23 06:37:00.20: [20093]: <-- [14:AT+FCLASS=1.0\r]
Nov 23 06:37:00.22: [20093]: --> [2:OK]
Nov 23 06:37:00.22: [20093]: <-- [14:AT+F34=14,1,2\r]
Nov 23 06:37:00.25: [20093]: --> [2:OK]
Nov 23 06:37:00.25: [20093]: DIAL 1
Nov 23 06:37:00.25: [20093]: <-- [6:ATDTxxxxxxxx\r]

At 06:37:57 FaxLab went back on hook and at 06:38:18 we killed the job
because of timeouts in our system.  I'm assuming that hylafax log did not
finish because the process was killed.

If I understand and remember the hylafax plus 5.5.9 code correct, the
NSF/CSI/DIS is part of the hylafax code and should then be after the ATD
and ATA phase correct?  I've used data modem a bit and all of the rate and
protocol support negotiation is done by the modem.  Not sure in fax
modems.  What part is done by the modem?  Is it class 2 modems that handles
everything?

If the Davicom modem stops playing the calling tone, does that mean that
the ATD part is done as it picked up the answer tone and hands control over
to hylafax to handle the rest?  Just trying to pin point if the problem is
with the modem or hylafax?

[image: Inline image 1]

I've attached a screen print from FaxLab.  Not sure if it will pass down to
the mailing list.  I also have a audio recording if anyone is interested.

Kind regards
Christo