hull@hao.ucar.edu (Howard Hull) (11/26/90)
Because there have been recent questions regarding AB switch use, and because
of the content of an article recently referred to me through a response to my
query concerning VGA cards that work well with a bridgeboard, I thought it
would be useful to summarize some important information concerning the use of
AB switches as applied to the Amiga parallel and serial ports, and their use
as well with commonly available video monitors. The summary is given first,
and a section concerning philosophy follows that.
STANDARD OPERATING PROCEDURE
If you wish to attach a hardware peripheral known to work well with an Amiga
personal computer, the safest procedure (the one recommended by Commodore)
does not account for the use AB switches. It is:
1. To begin with, make sure that the AC power switch for each of the
items you wish to interconnect are turned off.
2. Make any of the pre-connect adjustments (i.e., DIP switch settings,
accessory attachments) required by the installation texts from the
applicable manuals, especially those that require removal of covers or
access plates, BEFORE making any connection to AC power as well as
before making any interconnection between devices.
3. REPLACE or close covers as the manuals recommend for normal operation.
4. Plug the AC power cords for each of the devices to be interconnected
into a common AC outlet (This may be a common quad outlet for a device
plus the Amiga, plus the video monitor, or, preferably, a common AC
outlet strip box switched OFF for several devices to be used together).
5. Now connect any not yet interconnected ends of peripheral cables to
each respective peripheral device, then connect all the Amiga ends to
the Amiga. Note that the presumption is you are attaching one NEW
device along with those that are already known to work with your Amiga.
6. It should now be safe to power up the devices in any order; however,
the most convenient method for use with a switched AC strip is to,
with the strip switch off, turn on the Amiga followed by the monitor
and peripherals in any order, followed by turning on the AC strip
switch. This way, if the new surge load is too high, the breaker will
trip right away, saving the worry of having it trip after a larger
fraction of the equipment is up and running.
7. Proceed with the checkout procedure which you have in the manual(s).
AB SWITCH SERIAL PORT USE
Now then, if you wish to use an AB switch, the safest use is with equipment
which has been manufactured specifically for use according to the EIA-RS232
serial communication standard. This includes modems and serial port use of
printers. It specifically does not include parallel port printer use or the
interconnection of computers and video monitors. It is very wise to first
check out the new equipment without the AB switch so that you will have some
recognized correct behavior to go on before testing the two devices with the
AB switch in place. Avoid at all cost, though, the error of connecting a
serial device to the parallel port! Refer to the truly electrifying article
in material appended below following the philosophical discussion. Follow
the procedure given above, except insert the following:
5A. Attach the two RS-232 peripherals to the AB switch with the switch
set to connect the peripheral formerly known to work correctly with
Amiga hardware, then attach the new peripheral to the AB switch, and
finally, attach the C port on the AB switch with a suitable cable to
the Amiga. This last cable needs to be a 1 through 8 plus 20 cable.
Use of a cable with pin 22 connected is not usually recommended with
an Amiga which also has a device connected to the parallel port, as
such interconnection may interfere with operation of the parallel
port. Do note, however, that some serial port modem software may
require signal from pin 22 (Ring Indicator) to function properly.
If you wish to operate such software, it is recommended that you
achieve a functioning configuration through acquisition of a multi-
port serial device card (e.g, the one made by ASDG) and use software
that properly handles multiple Amiga serial port devices (usually
accompanying the hardware package).
Begin testing with a checkout of the first formerly known functional device.
With satisfactory completion of the test of this device, close the program
(freeing the serial port resource), flip the AB switch to the new device and
(with your eyes open and your nostrils clear) hit the Enter key a time or two
to assure that the Amiga is still alive and kicking, and then proceed with the
test of the new device. Of course, failure of the Amiga to respond to the
Enter key would mean something was improper. Time is of essence in this case,
though you will need to rely on your best instincts in handling these awkward
situations. My reaction is usually to first kill the power switch, then to
disconnect the new device and start over with a thorough checkout of the
formerly functioning systems. I do this in preference to flipping the AB
switch back to the other device, as it is possible that the switch may have
welded, or that the port may have been damaged - a condition that has some
potential for carrying destruction to the formerly functioning device.
With RS232 devices, we are relying on the specification for the line drivers
and receivers used in that context to withstand certain kinds of anticipated
abuse. RS232 line drivers are specified to withstand connection to a low
impedance source of electrical power in the range of -12 to +12 volts DC and
unrestricted ampere capacity, while driven to either marking or spacing state.
RS232 line receivers are likewise specified to withstand input signal potential
in this range. While the 8520 CIA chip used in the Commodore Amiga 1000 and
2000 computers CANNOT withstand this abuse, it does survive capacitive current
surges associated with the interconnection of normal cable lengths (i.e., 6FT)
between the Amiga and peripherals which are already interconnected by common
green-wire (third prong on the power plug) case grounds. The use of specified
RS232 line drivers and receivers in the Commodore Amiga 3000 indicates that
usage considerations have finally outleveraged cost considerations in the
manufacture of the latest members of the Commodore Amiga computer lineage.
AB SWITCH PARALLEL PORT USE
Connection of devices and use of an AB switch with the Amiga parallel port is
not covered by any standard, and is thus not something Commodore can support
any more than it can support the repair of damage to your equipment that may
have been caused by lightning strikes. An authorized CBM repair facility will
treat any such related complaint much as though they thought you HAD been hit
by lightning. :-) They are, of course, in a position to utilize a description
of whatever happened just before you started smelling the smoke, so stay alert.
As it was with the serial port, the key to success is in the order with which
devices are connected, and, in particular, the pin numbers of the wires in the
AB switch port C cable. It is best to use cables which come labeled from the
manufacturer with the identity of the devices expected to be connected by the
cable. Redmond Cable in Redmond, Washington understands this and supplies an
astonishing variety of cables for use with the Amiga. If they make a cable
specifically for your application, by all means take advantage of the product
they offer. The same goes for CBM. If you have a cable which came with the
CBM Amiga peripheral device you wish to use with the AB switch, by all means
use that cable in the connection from the device to the AB Switch. But...
remember in any case that you are on your own, and any damage done is your
responsibility.
Let's say here that with the parallel port (as well as with video monitors)
there is one aspect of the AB switch which represents a positively identifiable
hazard. This is the matter of assuring that the device and computer grounds
are connected together before any of the signal lines are connected in order
that the electrical potential differences between the devices may be held
within specified tolerances. The return of currents associated with applied
power or ground potential differences between the computer and peripheral
devices intended to be connected to the computer via the signal drivers is
a likely source of damage to the peripheral and the computer alike. That is
the case for anybody's computer and anybody's peripheral equipment - it is not
an Amiga-specific problem in the slightest. The typical AB Switch - anybody's
AB Switch - DOES NOT by itself CONNECT THE GROUNDS before it connects the
signals! This is because such connection is application specific, and there
is no way the switch manufacturer can know a priori what will need connection
before whatever else.
So, to protect your equipment from misadventure, follow the steps in the
procedure above that begin by assuring that the equipment protective grounds
are connected via a common AC source (i.e., get an outlet strip with an ON-OFF
switch). Continue your efforts in this regard by purchasing only shielded
cables for your applications, and by making sure that the shields are then
connected appropriately (the shield is to be considered in this regard as an
extension of the device chassis). An AB Switch which interconnects the
Subminiature D connector SHELLS is one which will help enforce this
requirement, and which may be required for some Amiga parallel port devices
(i.e., Digiview and the Perfect Sound digitizer). However, I do note the
following VERY SERIOUS caveat from the original Amiga 1000 manual:
"Parallel Connector
WARNING: Pin 23 on the Amiga parallel connector supplies +5 volts
of power. Connect this pin ONLY if the power from it is required
by the external device. NEVER connect this pin to an output of an
external device or to a signal ground.
Pins 14-22 are for grounding signals. DO NOT connect these pins
directly to a shield ground."
NOTE THAT THE ABOVE APPLIES ONLY TO THE AMIGA 1000 COMPUTER AND THAT THE
AMIGA 500/2000~ USE A DIFFERENT ARRANGEMENT, PROTECTED BY A PORT GENDER
CHANGE BETWEEN THE COMPUTER TYPES!
Interestingly enough, NO PIN on the Amiga 1000 is specified as a shield pin
in the supplied manual, and no pin outside the range of pins 14-22 has a
name even closely resembling shield, ground, or common designation. :-(
Nothing whatever is stated about the SHELL of the connector, which seems by
default to be the place where power and shield termination is left to take
place.
Given that the control of Electromagnetic Interference generated by digital
computer equipment is attained in no small way via the handling of cable
shields, failure to make the method of shield termination and power return
specific is a grave oversight. I therefore am forced to come to the conclusion
that the SHELL is the expected termination point for such functions.
Normally in the interconnection of electronic peripherals, the use of shield
elements to return power (as from the +5v pin cited in the above) is resolutely
discouraged, as it tends to exacerbate system ground potential differences.
The technical people at CBM are not as careless as this commentary of mine may
have them seem. This is far more likely a problem coming from some aspect of
the way Big Blue or Centronics has done things in the past than it is just
an Amiga problem...
Even if the Amiga 1000 can meet the A or B FCC spec, without shield management,
combinations of that computer and other peripherals are rather likely to fail
to meet it. As an example tending to the case, the difference between cables
with and without shielded connector hoods, in so far as radiated energy is
concerned have been found to be typically in excess of 60 to 90 decibels for
those frequencies thus elected most likely to cause interference to broadcast
television transmissions.
Now, from the Amiga 2000 manual:
"WARNING: Pin 14 on the Amiga parallel connector supplies +5 volts of
power. Connect this pin ONLY if the power from it is required by the
external device. NEVER connect this pin to an output of an external
device or to a signal ground. Pins 17-25 are for grounding signals.
DO NOT connect these pins directly to a shield ground."
THE ABOVE INFORMATION APPLIES SPECIFICALLY TO THE AMIGA 2000 COMPUTER.
Again, no pin has a designation even remotely interpretable as either a shield
or a power return ground. What is not said, but which should be clearly
specified, is that this function is available via connection to the SHELL of
the parallel port Subminiature D connector! For the most part, devices which
use power from the parallel port are designed to operate as the sole device
connected to the port, and are by policy to be installed or removed with the
computer power OFF. The risk of using these devices - from the standpoint of
both safety and function - with an AB switch is chancy at best, and who knows
what at worst. However, it must be said that there are reports of successful
use where the shells are connected in a non-interruptible fashion across the
switch.
Last but not least, do note that the series resistance of the +5v supply pin
on the Amiga 1000 and 2000 are very different, the resistance being substantial
in the case of the Amiga 2000. With the Amiga 1000, this resistance is low
enough that serious damage can be done to both the computer and an attached
peripheral if connection to this service is inadvertently mismanaged (i.e.,
via neglecting to consider the fact that it is THERE and like the proverbial
genie, is WAITING...).
Thus a cable or switch box intended for use with the Amiga parallel port must
be checked against the computer device and peripheral pin signal lists on a
pin-by-pin basis before it is connected for use. Do note that some AB switch
boxes, to add to this misery, may not connect all 25 pins. Check with a
continuity tester to be certain that the pins you need are carried through
the switch, and that the pins which are not carried are of no consequence.
It is important that pins 17 through 22 are carried through the switch. It
would be nice if there were eight pins thus assigned so that single bit
twisted pairs could be handled together. If you have an Amiga 2000, you can
use pins 17 through 25. But such is definitely not the case for the Amiga 1000
due to the usurpation of pin 23 for +5 volts; so don't get careless in your
use of parallel port cables and gender changers with an Amiga 1000 hanging
around the house. MAKE NO ASSUMPTIONS ABOUT GROUNDS with respect to these
pins: 15, 16, 23, 24, and 25! (Pin 16 is the RESET* signal.)
Use of the AB switch with parallel port the after installation is fairly
straightforward, though the more survivable situation is one in which both
peripherals are kept powered down until after selection with the switch,
and are powered off before changing the switch. Changing the switch with
the devices powered is a trifle risky, and it may take some time before a
specifically existing hazard is encountered or determined.
AB SWITCH VIDEO MONITOR USE
The situation for AB switch use with video monitors is similar to that with
the parallel port, though there are in most cases specific pins assigned for
signal and shield grounds. The chance of these being the same on any random
monitor as it is on the Amiga is not at all good. The connection of low
impedance or a logic low level on pin 2 (XCLKEN*) will cause the Amiga to
perceive that the connected device is a genlock - The Amiga will hang waiting
for external clocking on pin 1 (XCLK*). Similarly, except for the exact
logical process, low impedance on the sync pins 10 (CSYNC*), 11 (HSYNC*) and
12 (VSYNC*) may trip up some models of the Amiga during ROM boot. There is
power on pins 21, 22, and 23. The Amiga manuals have a note warning about
this, so one is forewarned. However do note that the Amiga 1000 manual says
"Pin 21 -5V POWER (50 mA)
22 +12V POWER (175 mA)
23 +5V POWER (300 mA)"
whereas the Amiga 2000 manual (319927-02), page A-9 says
"Pin 21 -12V POWER (50 mA)" <--- PROBABLE TYPO, it's -5V, really!
22 +12V POWER (100 mA)
23 +5V POWER (100 mA)
->Note that the voltage shown is different for pin 21 and the maximum current
->is different for pins 22 and 23. I believe the reference for -12 volts is a
->class 1 TYPO, as the schematic, page F-13 show this line as connected to
->the -5V supply.
Similarly, the A3000 electrical schematic diagrams show pin 21 as -5V.
The Power Budget on page 226 of the Commodore Amiga A500/A2000 Technical
Reference Manual lists it as -5V with an expected load of 10mA. The appendix
schematic on page B2000-11 in that publication agrees with the -5V rating.
HOWEVER, the schematic in the "Rock Lobster" section, on page A500-4 shows
the voltage as sourced in -12V and emerging as -5V on pin 21 of DB235 after
passing through only a ferrite doughnut (component marked "EMI"). Yeeeeek.
It IS true that in going from the A1000 to the A500/A2000 one of the pins on
the expansion bus (internal to the A2000 only, external with the A500) is
-12V while it was -5V on the A1000. But this is not the case for the DB23
video connector. For the most part, the differences will not affect the
monitors, as they do not use power from external sources. You can make sure
they don't by seeing to it that the cables you use to attach a monitor do not
have these pins wired.
All this does not make life simple; do be aware that being strung up by your
thumbs without an available monitor port is infinitely more aggravating than
being similarly geo-potentially tethered without a parallel port.
It is IMPORTANT to try to get a monitor that allows switching between various
video input signal types by means of a front or back accessory switch. The
availability and use of such a switch instead of an AB switch will protect your
equipment by allowing the grounds to remain connected while the signals are
left to float (the switch may/should remove the termination when it disconnects
a video input cable, but this is something you'll want to test out with an
analog or digital volt-ohm-meter (VOM). Follow the usual rule with terminators,
i.e., see to it that there is only one, and that it is at the end of the line).
The Amiga 1080/4 monitor has as perhaps its only salient saving feature a front
access control row which allows excellent flexibility with respect to signal
selection choices. I hope the 1950 is as good and Commodore keeps up at least
this part of the good work.
The monitor I use has separate switchable 9-pin TTL and 25-pin ANALOG video
inputs; this saves some grief when used with the EGA modes of a separate video
adapter and a bridgeboard. Of course, it doesn't help at all with the analog
signal from a VGA card, but oh well, we tried, didn't we?
Other than that, the risks are the same as with the parallel port. An aux
video monitor that is seldom used is best left powered off until connected
via the AB switch, then powered down before moving the switch back. With
the bridgeboard VGA this is not a usual expectation, so one must accept the
risks associated with the AB switch and try to rely on safety grounded 3-wire
AC cords connected to a common receptacle row at the first level, and well
managed and connected cable shields at the second level. Using an AB switch
without some attention to these two details will probably cause system damage
problems just when you need them least.
With regard to the Amiga 3000, there is nothing I have seen in the news that
would indicate that the parallel port specification for it is different in any
important way from that of the Amiga 2000. The 1950 monitor video is changed,
in that there is a 15-pin Sub-Miniature High-Density D connector for use by
the monitor for signals of the VGA class. You have to be nuts to want to
use a bridgeboard in an Amiga 3000 (and to help you out in this regard, none
of the earlier bridgeboards will function in a 3000 anyhow). Other than that,
the same gruesome rules regarding the use of an AB switch on any of the ports
applies: identify and avoid the power pins, common those AC receptacles, and
connect those shield grounds - avoid the use of gender changers wherever
possible in order to assure signal and ground continuity and to minimize the
possible mixup with other non-similar Amigas/devices.
PHILOSOPHY
The Commodore philosophy (now this is a personal opinion of mine) appears to
be one of recommending that users operate with a hardware configuration that
can be stable for months at a time. Changing needs are to be met by the
flexibility built into the software, and through stable hardware additions.
The AmigaDOS operating system provides a greater measure of flexibility via
multitasking and varied processor use than just about any personal computer
now available on the market. If you find yourself changing the hardware
configuration on a daily basis, you are functioning outside the operating
envelope intended for the machine, and you need to look into the possibility
of acquiring more than one machine and taking advantage of the machine's
networking capabilities (i.e. Ethernet). Even public domain programs such
as ParNet will do a great deal for you, given that you survive the initial
connection of machines in this way. ParNet is a fairly hazardous adventure,
though, and is not for the uninitiated. Until a multiple parallel port card
is commonly available for the Amiga, ParNet will remain mostly a curiosity
for the intellectual Amiga user.
THE INTERCONNECT DAMAGE RISK AND DATA RELIABILITY
In the early days of IBM and DEC computers, data reliability diagnostic
tools were part of what was delivered with the computer. What these tools
tested were the computational and execution consistency of the digital
hardware; they tested it by submitting abundant varieties of data and
program execution tasks, and checking the computer for expected output.
Damage to computer hardware is oft-times much the same as what occurs in
human injury, a graceless degradation in some important operating facility
with potentially disastrous consequences - given some infrequently encountered
situation. No balm for any wounds, the wrong number on line such-and-such of
some federal form can spell disaster for a professional accountant who depends
on computers to produce results formulated through a specific program. The
small PC is not particularly accountant-reliability rated without software to
test for data reliability problems. The first IBM PC released had problems
with small negative numbers, and the first releases of AmigaBASIC (not ABasiC)
did not calculate the squares of some numbers correctly. While these rude
encounters are not data reliability problems of the sort I am trying to
address, they are examples of how such matters escape the unattentive.
If you subliminally damage your hardware through the potentially unbounded
effects of use of an AB switch, it may come through in ways not explained
to you, even by those who know specifically what can go wrong. If you're
out there kicking electronic butt through adventurous hardware changes, you
can expect occasional shorts in your electric underwear, no doubt about it.
This is for anybody's computer, and anybody's peripheral hardware. Some time
ago, someone posted an article with diagnostic information that was likely
extracted from Commodore service bulletins intended for use by authorized
repair centers. The re-publication, while not particularly official in its
content, was useful to a number of users who had encountered problems with
their computing machinery that were stranger than life. There is at least
one indication that a fault on one of the ports, in addition to blowing the
8520 CIA chips, can reach through to other hardware and can be the source of
data reliability problems. To wit, all 633 lines of it, the referenced
article is a real gem. Has this ever been updated?:
<Article 43213 of comp.sys.amiga:
<Path: ncar!ames!pacbell!sactoh0!americ!erk
<From: erk@americ.UUCP (Erick Parsons)
<Newsgroups: comp.sys.amiga
<Subject: Fix that baby!
<Message-ID: <1982.AA1982@americ>
<Date: 18 Sep 89 21:54:11 GMT
<Followup-To: comp.sys.amiga
<Distribution: na
<Lines: 633
<
<I've noticed a lot of help me messages on the board lately and can sympathize
<with the poor person that loses use of his machine due to a shortage of
<that magic smoke that holds silicon chips together soooo....
<
<For those of you that don't have access to a descent BBS here is something
<that I found on one of our local boards that looks VERY useful. I took this
<file printed it out and filed it so's that if and when my beloved machine
<should stop ticking I could fall back on this... Looks good to me :-)
<
<--------------------------------cut here------------------------------
<
<
< THE FLAKE REPORT
<
<"Flaky" (July 15, 1989) Hacker's slang as meaning:
< 1. Erratic and/or unpredictable behavior or experience
< 2. Sometimes it works; sometimes it don't
< 3. Primary cause of the following auditory spasms:
< a. "ARGH!!!!"
< b. "AH XXXX..."
< c. "#@&$!"
< d. "I need a career change; marketing!"
<
<Usage: "He's getting FLAKY..."
<Means: "RUN! HIDE!"
<
<A Bruce's Repair quote:
<"If it works; it won't."
<
<Ok folks, I'm compiling a list of what has caused flaky operation of the
<Amiga computers. This is a very unpopular subject with Amiga worshippers,
<but it should be one to pay particular attention since it can harm any
<product faster than obsolescence and DOES cause untold user-hours of
<frustrations. Flakiness affects everyone and everything from product
<development to end-users and especially service. If you can add some more
<to this list, you can give me a call at (415) 525-6973 and ask for Bruce.
<I'm busy most of the week so the best time is on Sundays in the morning. If
<you don't mind listenning to a morning grouch call any other morning.
<
<...
<A500
<...
< 5) flaky keyboard assembly with spurious characters or crash with
< CAPLOCK LED blink
< a) replace keyboard and tell customer of possibility of a serial
< device being connected to the parallel connector
<
For all my adventurousness, I have an A2000 keyboard which does this once
or twice a week, whether I need it or not. I have turned it upside down and
flushed it with low velocity compressed air. I have NO IDEA what's going on,
as it will do it on either of my two A2000 computers... It is too infrequent
to trap realistically.
Please post corrections to this article as you feel appropriate.
Flames should be directed to me by email.
BZZZT.
Best Regards, Howard Hull
hull@hao.ucar.edu
--- From the film Carnal Knowledge: ---
--- "My father was a failure. He gave me lots of advice. ---
--- The more he failed, the more advice he gave me." ---