Path: utzoo!utgpu!water!watmath!clyde!bellcore!decvax!decwrl!labrea!agate!pasteur!ames!elroy!cit-vax!ucla-cs!wales
From: wales@CS.UCLA.EDU
Newsgroups: comp.sys.ibm.pc
Subject: Speaker "clicks" that should have been "beeps"
Message-ID: <9506@shemp.CS.UCLA.EDU>
Date: 17 Feb 88 06:09:07 GMT
Sender: news@CS.UCLA.EDU
Reply-To: wales@CS.UCLA.EDU (Rich Wales)
Organization: UCLA CS Department, Los Angeles
Lines: 72

I have long noted that a couple of programs I run have a tendency to
mess up my speaker, so that one or more "beeps" from the speaker end
up being toneless "clicks" instead.

My machine is a Wugo PC-II-AD (a Taiwanese "turbo" XT clone with an
8-MHz NEC V-20 processor), marketed by Sun Computers Inc. (not to be
confused with Sun Microsystems Inc.).  As far as I can tell, it is com-
pletely IBM-compatible; I have yet to encounter any "IBM or 100% compat-
ible" software that fails to run on it.

Two programs I am aware of which tend to cause this "click" mode are
the Leading Edge Word Processing program and the Spinnaker Backgammon
game.  Among the programs which thereafter suffer from "click" syndrome
are ProComm 2.4.2, and a disk parking utility (PARK, version 2.00, by
Corona Data Systems).  Spinnaker Backgammon suffers from the "click"
problem as well as causing it -- about half of its "beeps" come out as
"clicks", with no pattern that I have been able to discern.  I am sure
there are other programs which cause and/or suffer from this problem,
but these are the ones I am personally familiar with.

If I run the DOS command "ECHO ^G" (control-G), it always "beeps", and
a subsequent command does not suffer from "click" syndrome.  (I am using
the NANSI terminal driver.)  So, whatever the NANSI driver does for a
"bell", it clears up the problem with apparent 100% reliability.

After running one of the "click-causing" programs, I used DEBUG to look
at the status of the 8253 programmable timer chip.  The count register
of timer #2 (I/O port 42H) consistently shows a value of 2 (i.e., the
first "i42" DEBUG command gives "02", while a second "i42" gives "00").

If I subsequently reset the counter manually (e.g., to 400H, via the
DEBUG sequence "o43 6B", "o42 00", "o42 04"), and then activate the
speaker by setting the two low-order bits of I/O port 61H, frequently
all I get is a "click" -- no tone.  If I repeat the "counter set"
sequence, however, I get a tone as expected.

The above manual procedure is *not* 100% reliably repeatable.  However, 
it does work often enough that there seems to be little doubt in my mind
but that I am seeing some kind of problem relating either to my 8253
timer chip or to the support circuits.

I haven't been able to get my hands on a "true blue" IBM PC or PC/XT to
see if the problem occurs there.  I did try running Spinnaker Backgammon
on an IBM PC/AT and PC/XT Model 286, without observing the "click" prob-
lem; however, these are 286-based systems and undoubtedly differ in the
clock timer components and circuitry they use, so my failure to get them
to "click" is probably meaningless.

My questions for the net:

(1) Has anyone ever observed this problem on a "true blue" IBM PC or
    PC/XT?  Is this a generally recognized problem and not simply a flaw
    in a handful of Taiwanese turbos?

    I assume, by the way, that it *is* a generic problem -- since, for
    instance, I remember several people talking about it on Datastorm's
    BBS (Datastorm being the ProComm company).  For what it's worth, the
    Datastorm people were apparently unaware of the problem and had not
    a clue as to what might be causing it.  This may mean it doesn't
    happen on "true blues" -- or it might simply mean that the Datastorm
    people aren't using any of the "click-causing" programs.

(2) Does anyone know what is causing the problem?  Is there, by any
    chance, a standard hardware fix for it on the IBM PC and PC/XT?

(3) What programming techniques are commonly accepted as workarounds for
    the problem?

-- Rich Wales // UCLA Computer Science Department // +1 (213) 825-5683
	3531 Boelter Hall // Los Angeles, California 90024-1596 // USA
	wales@CS.UCLA.EDU           ...!(ucbvax,rutgers)!ucla-cs!wales
"Sir, there is a multilegged creature crawling on your shoulder."