-=> Quoting Hayes Support to Bill Dawson <=-
 BD>  Subj: Multiple Re-dials, Original to Craig Ford
 
 HS> What is the model number of the ACCURA 28.8 that you have?  If you
 HS> could, please issue an ATi3 to the modem and let us know what it
 HS> returns so that we will be able to tell exactly what version of the
 HS> modem you have.  Is the modem external or internal?  If internal, what
 BD>                               ~~~~~~~~
 HS> com port is the modem configured for and what com ports do you have on
 HS> the system? Please let us know what you find.
Two Com Ports, with a serial mouse on #1 and the external on #2.
Let's back out out of Blue Wave for a moment to issue the ATi3, 
shall we?  All right, that's done & it says:
04-00971-720 34AC
[ ... ]
 BD> 486 (DX-100, 32 MB, 16550 for both UARTs).  Two new 28.8 FaxModems 
 BD> & one definitely doesn't do as well.  With that software, there are 
 BD> two BBS's with whom it may require anywhere from four to six re-dials 
 BD> to connect.  When using Zmodem for downloads, CRC errors can run to 
 BD> **ASTRONOMICAL** counts.  There were 50-something of them a few days 
 BD> ago on a 1.5 MB file!  
 
 BD> The number of re-dials required has remained fairly consistent all
 BD> along since replacing the Accura 14.4.  There are quite a few BBS's
 BD> with whom first-call connects continue to be the norm, of course.
 
 BD> The CRC error problem in Zmodem is not constant; fairly often the
 BD> downloads of mail run smoothly.  And these repeats of the same
 BD> block of data weren't required any more frequently when the 28.8
 BD> first went into use than had been possible to blame on phone line
 BD> problems with the prior modem.  Since mid-September (I got these 
 BD> Accura modems [direct from Hayes] in May of 1996), the CRC's have 
 BD> become endemic!
 
 BD> I can think of nothing that has been done with or to the system to
 BD> affect the operation of a transfer protocol!
 
 BD> With one particular BBS (the local 1:387 NC), downloads would hang
 BD> on a block and Zmodem would keep requesting repeats until I gave
 BD> up on the download, almost always at about the 90% complete mark.
 BD> A change to Xmodem-1K helped some, but even that protocol lost one
 BD> download in mid-stream, had to be aborted & started over.
 
 BD> At this point, neither I nor QMPro are using anything other than
 BD> the default ("ATZ"?) to wake up the modem.. BUT it does seem that
 BD> part of the trouble is the software.  I don't like ProCommWin much
 BD> at all, yet on a couple of experimental calls, it *can* manage a
 BD> connect at either "problem" BBS with one call, and *its* version of
 BD> Zmodem seemed to be devoid of the CRC's.
 
 HS> To add on to this, generally CRC errors are not indicative of a
 HS> 'modem' problem.  When dealing with internals the problem may be traced
 HS> to a hardware conflict, with both internals and externals it could be
 HS> caused (Continued to next message)
(If the second part made it through Fido, it wasn't flagged as such 
here at The Rock.)
And although it's true that multiple CRC's aren't the "fault" of a 
modem, it is a fact that when you must re-dial frequently to connect 
you do have some modem problem.
 HS> As for our online support, we can be reached via any of the following
 HS> means:
 HS> Hayes Online BBS: 770-446-6336 or telnet://hayes.com
I uploaded a query there a few days ago; the upload, however, wasn't 
sitting in the message queue when I checked back there the day before 
yesterday; so I uploaded something rather similar to the message to 
Mr. Ford on that second call.  At present, regular Email via the Inet 
isn't among my communications options.
- - - - - - - - - - - - - -> Bill <- - - - - - - - - - - - -
... * Lizzie Borden...America's first hacker.
___ Blue Wave/386 v2.21
--- Maximus/2 3.01
---------------