Re: [hylafax-users] [hylafax-devel] Re: Can not initialize modem

[Lee Howard <faxguy@deanox.com>]

At 08:49 AM 7/9/01 +0200, Bernd Proissl wrote:
>
>
>> >Jul 08 12:11:04.11: [ 4235]: <-- [4:ATZ\r]
>> >Jul 08 12:11:04.25: [ 4235]: --> [2:OK]
>> >Jul 08 12:11:04.25: [ 4235]: <-- [5:ATE0\r]
>> >Jul 08 12:11:09.25: [ 4235]: MODEM <Timeout>
>> 
>> If you issue the above AT commands in cu/minicon, do they work?
>> 
>> Can you try
>> ModemATCmdDelay: 100 (200, 300 ...
>> 
>> just to see if the modem is upset to receive ATE0 too soon after an ATZ?
>
>I had similar problems with my Elsa(s) microlink 56k (the timeout does not
>happen 100% but only maybe 1%-5% and at different AT commands).
>
>Giulio, your tip help me! I did some (extreme) testing with a value of
500. Some 
>hundred faxes have been sent until now without the timeout error :-)
>
>Lee, the default value is 0, this might bee what you (or Robert?) called 
>"hylafax is demanding". In minicom you probably can not type so fast ;-)
>Is minicom using a delay here?

Hmmm... I don't think that this is it, but Robert would need to define what
he meant by that.  I suspect, however, that there still would be problems
with things where we've experienced them before.

However, that delay is after the OK response from the previous command.
Does anyone see in a developer's manual for their modem indication that
there should be a delay after OK and before the next command is issued?

>Harald, does a value of 0 make sense to you? Is something like 10, 20, 50
better?

From empirical evidence, it seems that 0 is not a proper default.  I'd say
that 500 ms is a bit extreme, but 50 ms wouldn't be a bad situation,
either.  I'm interested to know if other software packages (i.e. Harald's)
use a default delay between the OK and the next AT command.

Lee.



____________________ HylaFAX(tm) Users Mailing List _______________________
 To unsub: mail -s unsubscribe hylafax-users-request@hylafax.org < /dev/null