Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!wivax!decvax!harpo!floyd!vax135!cornell!uw-beaver!tektronix!tekecs!johnl
From: johnl@tekecs.UUCP
Newsgroups: net.micro
Subject: Re: USR Password modem
Message-ID: <1527@tekecs.UUCP>
Date: Tue, 28-Jun-83 20:52:10 EDT
Article-I.D.: tekecs.1527
Posted: Tue Jun 28 20:52:10 1983
Date-Received: Thu, 30-Jun-83 02:39:04 EDT
References: hou2b.36
Lines: 63

I have seen the Password.  We have a lot (>20) of U.S. Robotics
Auto-Dial 212A modems and are very happy with them.  USR has
the modem technology down pat, and our experience and examination
of the modems indicates that they will be much more reliable than
nearly any other modem, certainly Vadics and Hayes, since they
are nearly all digital.  Specifically, the only precision
components in the whole thing are three precision resistors in
the clock circuit.

The Password is a disappointment.  Its concept is great.  Good
technology packaged without frills for low cost.  The only
problems are with execution.  The mechanical design stinks, and
I've told the local distributors about it.  Some examples follow:
The LSI components are on stilts (tall sip-style sockets) with
components underneath.  Sockets are a continuing reliability
headache, these are prototyping sockets, and the components
underneath keep them from doing anything about it.  One side
of the 2920 IC doesn't even go all the way in (I tried); a second
password I saw didn't have that problem, but it is indicative of
the problems likely to occur.  Perhaps the worst thing is that
the voltage regulator ICs are in TO-220 cases with their leads
bent flush against their (plastic) package.  This is a no-no, since
it stresses the bonds inside.  A few shocks may cause these parts
to fail.  A smaller problem is that the cable is not strain-reliefed,
although it is unlikely to pull off because the connector is at
right angles to entry/exit.

The above problems plus the (minor) advantages of a stackable metal
case, lots of LEDs, and an analog loopback switch make it unlikely
that we will buy passwords at work for a while.  (For a personal
purchase, where the modem would receive less abuse, where I could
afford to be inconvenienced by a failure, and where money is tighter,
I might buy a Password.)  Almost all of the problems can be fixed
(albeit with a new circuit board layout).  So keep checking them out.
Meanwhile USR still makes the best 1200 baud modems.


When you order a Password, be sure to specify whether you want it
with a male or female terminal connection.  This is assinine, since the
wiring is DCE with either sex connector.  RS-232 is absolutely clear
that Data Communications Equipment (modems) is wired with a female
DB-25 connector, and just as clear that there is NO legal configuration
with DCE wiring of a male connector.  USR's action is, of course, the
result of demand by owners of certain terminals and printers that
use female connectors for God-only-knows-what reason, when the RS-232
standard clearly states that not only will terminal equipment have
a male connector, but the terminal equipment will supply the cable.
If you have ever (or often, in my case) come across two pieces of
"RS-232" equipment that didn't talk, couldn't be made to talk easily,
or simply confused you for a while, it is because manufacturers
continually violate the specification to meet "demand".  Then the
next product has even more "demand".  DEMAND that your Password come
with a female connector.  SOLVE your interfacing dilemma at the other
end of the connection where the real problem is.  REFUSE to buy
equipment that claims to meet RS-232 but doesn't, and TELL the
manufacturer why.


Thank you for your time

John Light
decvax!tektronix!tekecs!johnl
ucbvax!tektronix!tekecs!johnl