"Smart" serial boards for the 80386

[Howard Leadmon]

In article <218 at turnkey.TCC.COM>, jack at turnkey.TCC.COM (Jack F. Vogel) writes:
>
> I have been watching the Bell ICC reports with much interest since we are
> using one on turnkey. I think that much of the flames are deserved and I will
> add a couple of my own, but this talk of not working with a modem is not
> true. I had 2 Multitechs running on the card and they worked fine ( that is
> as well as the card worked overall ). 

 Well I don't know how the thing works with the Multitechs, but on a Telebit
Trailblazer (which a good many netters have) the board would generally run
OK with 1200 and 2400 connections (90% of the time), but if you had a connection
that ran in PEP mode (9600 or 19200) FORGET IT!! When attempting to run in
PEP mode upon termination of the connection the port would drop DTR and it 
would NEVER return. Everytime I talked to Bell about this problem I was told
two things, first the problem may be the TB+ since it is a NON-STANDARD modem,
and second I have been told that there card was to FAST for the UNIX kernel
and it was getting ahead of everything (not quite sure I follow this one, so
I won't even try and expand on it). So anyway my conclusion on the Bell ICC
is that it is good on a modem as long as you DON'T EXCEED 2400 BAUD.. Also as
I mentioned in my previous article, the Telebit Trailblazer Plus modems are now
running without a flaw on the Smart Hostess board that I installed as a 
replacement for the faulty ICC. So if the TB+ works on a standard DUMB serial
port, and also on another vendors Intelegent Board, then I can't point a 
finger at the TB+ modems...

> This talk of the wrong sex connector
> makes no sense. You must realize that Bell sells a number of different
> RJ45-DB25 style adapters. When we took delivery on the card it had only
> one such adapter, and it was for terminals, which means it is effectively
> a NULL MODEM. Now if someone tried to use this with a modem it certainly
> would not work!!! It was not the wrong "sex" however, it was male. What we
> did was make our own modem adapters so that the signals would be "straight
> through", and it worked fine as I said. The previous posters might give
> the mistaken impression that the ICC does not support the necessary signals
> to control a modem (such as at least some cards from, say, Arnet). This is
> false, it supports the standard 8 lines and provides for real hardware
> flow control. One must just make sure that the proper adapter is installed.

 No I never implied that it dosen't support MODEM signals, in fact I also had
to make my own RJ-45 --> DB-25 connectors. Also before I ordered the card from
Bell I checked to see that it would work with a modem...

> Now on to the flames...the card has been causing us considerable grief. It
> has been run in two different systems. In the first it would cause huge
> memory fault core dumps under considerable system load. In the present
> system it doesn't core dump, instead it causes general protection panics
> and brings the system down altogether, wonderful :-} :-{. At this point
> we have it out of the system. Now a bit of good news comes from Paul
> Vixie who says there is a new release that has solved the problems. I have
> corresponded with him and he says it has shipped since 2/88. Now the
> question is will Bell make good and allow those of us who bought the
> earlier lemon exchange for a product that works?? Are you listening
> Bell?????

 That would be the correct way to handle the problem, but wonder if they
will be willing to take the loss on the card. Also I wonder which version
of UNIX/Xenix your are running, and also what version of there ICC driver.
I was using there own Bell UNIX System V release 3.0, and the ICC driver
was version 0.9 (which at the time was there latest!!

> >Personally, I'd buy something other than a Bell Technologies board.
> 
> At this point I would agree with this conclusion, unless Bell can prove
> themselves with their new release. And I might add, they should do so in
> a hurry if they are to stem the proverbial tide of this bad press.

 Well I suppose in worse case we will stop others from having to go through
the same troubles we have experienced. I also really hope that Bell gets
there act togeather and resolves the problems, but at this point (and after
eight months of waiting) the possibility of that happening is looking bad..


-------------------------------------------------------------------------------
UUCP/SMTP : howardl at wb3ffv		|	Howard D. Leadmon
PACKET    : wb3ffv at w3itm-9		|	Fast Computer Service, Inc.
IP Address: 44.60.0.1			|	P.O. Box  171 
Telephone : (301)-335-2206		|	Chase, MD  21027-0171