Re: [hylafax-users] MultiTech ISI5634UPCI/8

["Steve Tuckner" <STUCKNER@xxxxxxxxxxxxx>]

The number after the . indicates what level of Agere code that the modem is based on. So 1.32i and 1.25p are based on different levels of Agere code. Agere code consists of controller code (which runs on a Z180 controller) and DSP code which runs on the DSP in the same chip.  We get the source code for the controller code but not for the DSP code. In the case of class 1/1.0 code, there isn't much controller code to support that functionality, so my guess is that the DSP code is the difference.  There is quite a bit of a time difference in the code bases with the .32 being probably several years newer than the .25 versions.

As far as the wedging issues goes, that is probably a controller issue that can be probably be fixed by us. The problem would be for us to duplicate the issue. 

One problem with using 1.25p is that controller changes that we have put in to support new hardware (ie. MT5634ZPX-PCI-U) are not in that though it may work anyway. Let me know you experiences on that front please.

Steve Tuckner
Multi-Tech Systems

-----Original Message-----
From: Mike McMullen [mailto:mlm@xxxxxxxxxxxxxxxxxx]
Sent: Wednesday, September 14, 2005 11:04 PM
To: Steve Tuckner; Chris Funk; Lee Howard
Cc: hylafax-users@xxxxxxxxxxx
Subject: Re: [hylafax-users] MultiTech ISI5634UPCI/8



----- Original Message ----- 
From: "Steve Tuckner" <STUCKNER@xxxxxxxxxxxxx>
Subject: Re: [hylafax-users] MultiTech ISI5634UPCI/8


> There is no equivalent downgrade from 1.32i to 1.25p for the ISI5634UPCI/8.  I am also surprised 
> that that downgrade fixed the problem below. The only reason I had seen people
> downgrading was because of lockups not connect issues. Also (this question is for Mike) I thought 
> the lockups related to 1.32i had to do with your phone lines getting wet or some such thing (see 
> email as of 4/18/2005). Is that not so?
>
> Steve Tuckner
> Multi-Tech Systems

Hi Steve,

Yep we had several communication problems involving bad/wet/partially broken
phone lines at that time. We later had some problems that effected our T1 and
phone lines about 30-45 days ago.

After all that was fixed and the phone company checked our fax lines twice in two
weeks and said they were ok I was still getting wedged faxes and the dropped
connections. We have several customers that may fax us 30-100+ pages of stuff
a day and there were a certain subset that consistantly got failures during their sends.

As their business increased, our business increased and more faxes resulted in more
failures. We average about 80-100 faxes with peaks around 120-150. Our average
fax size is 10-20 pages with 50-100 pages per some faxes pretty common every day.

Failures were usually the premature v.34 termination. With 3 fax lines and the size of the
faxes we get, V.34 means a lot to us. Let me add that our pagesizes are legal so 75 pages
of legal size paper is a big deal and faster is better if reliable.

So I declared war on failed faxes last weekend and and used a modem that reliably
failed when connecting the multitechs. Per some helpful suggestions from Lee, I sent
literally over 50 faxes total to both multitechs I have and traced what was going on
as I changed first this and then that.

I also went and bought high dollar shielded twisted pair phone line to run from the wall
plates to the machines in case intereference was a problem. Still got failed faxes sending
from the Best Data modem to the multitechs. I have faith in the Best Data modem because
it is the modem that we were directing customers to that couldn't connect to the multitechs
and all their faxes came through albeit at 14.4K max.

I finally dug up the email you sent me with the downgrade firmware and waited until
Saturday night, took the faxserver down and pulled the multitechs out and put
them in a Windows machine to run FlashWiz on the them. I had to trick FlashWiz
by renaming the hex file to get it to downgrade. The hex file apparently was for a
ZPX-PCI-V92. My modems are the universal ones and FlashWiz wouldn't use
the hex file.

Everyone says they are basically the same firmware, well I have experimentally
verified that. Although the pucker factor was way up there during the first flash.

I put the one downgraded multitech and the other at 1.32k (I upgraded one
during earlier tests with the same resulting failures) back in the faxserver and
brought it back up and sent the same fax from the best data to the multitechs.

The downgraded modem had 100% success rate over the test faxes. The
multitech with 1.32k had about a 70% failure rate. The downgraded modem
only received at 9.6k but it never failed. The other modem connected at 14.4k
but failed a lot.

So it looked like I was on to something here. I downgraded the other modem,
sent the same fax a fair number of times (12 pages of legal small print and pictures,
a residential appraisal) to that modem. It got 100% success rate.

So I have monitored pretty closely over from Monday onward. After about
275 faxes and several thousand pages I have had 5 fax failures and no
wedging. 4 of the failures were from documents jamming on the customer's
sending side. The resulting PDF showed scrunched and mangled pages. ;-)
The 5th was a sending station sent a disconnect. 1 failure out of ~275 is
A-OK by me.

I was going to wait until the end of the week to write this up and send it
to you per Lee's request until I had 5 days of experience but this topic
popped up and thought my experience might be helpful.

I don't know what the differences are in the firmware but Lee surmized and
I agree that the older firmware handles retraining in a more "tolerant" way
(for lack of a better description, mine not Lee's).

Hope this helps,

Mike






____________________ HylaFAX(tm) Users Mailing List _______________________
  To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
 On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
  *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*