jallen@netxcom.UUCP (John Allen) (09/03/87)
We are having some trouble with our IMS AT compatible PC's, and I would like to find out if other compatibles have the same problem, in order to determine whether the trouble is specific to the IMS PC, the Phoenix BIOS, or some other source. The problem is this: The damned keyboard locks up during data entry! Here are the things we've been able to determine: 1. It can be caused by using software to toggle "numlock" while typing at the keyboard. The uuencoded archive file below contains the batch file "num", and some .com files which allow this to be tested. Running "num", we can produce a lock-up within 1 to 3 minutes. 2. It would appear that only the keyboard sub-system locks up; in a mouse driven application, everything else continues to work. 3. It is not an isolated case; we have four IMS PC's, all of them exhibit the problem. 4. An IBM AT does not exhibit the problem. 5. It happens much more often in 10 MHz turbo mode than in 6 MHz mode. 6. It happens with any of three different clone keyboards or a "true blue" AT keyboard connected to the machine. 7. It happens under both DOS 3.2 and DOS 3.3. 8. It happens under at least two versions of the Phoenix BIOS. 9. IMS has a *hardware* modification which they claim solves the problem at 6 MHz, but not at 10 MHz. They want to buy back our machines instead of taking the time to solve the problem. Anybody know what the problem is? If you can, and would care to repeat the test using our test software, the uuencoded archive file below contains the following test files. Name Length Date ============ ======== ========= num.bat 64 2 Sep 87 numlock.com 15 1 Sep 87 numlock.scr 93 1 Sep 87 reply.com 14 1 Sep 87 reply.scr 102 1 Sep 87 ==== ======== Total 5 288 run "num", and type away, trying not to fill the type-ahead buffer completely, as the "buffer full" bell slows down the whole process. As num runs, you will see the numlock light toggle on and off. If this toggling stops, you'll find that the keyboard is locked up, and you'll need to reboot. If you do run this test on any turbo AT clone, please email your results to me, whether positive or negative. If we get any responses I'll publish a summary of the results. Thanks, John Allen ========================================================================= NetExpress Communications, Inc. seismo!{sundc|hadron}!netxcom!jallen 1953 Gallows Road, Suite 300 (703) 749-2238 Vienna, Va., 22180 ========================================================================= ----- Don't cut here, just uudecode! ------- begin 644 IMSbug.arc M&@AN=6TN8F%T````````/0```"(/7(%NWT`````,9<:@>0/BC1DS#10$'`AB M"ITP<NB`H)('3AHW9Q+J8//F#9R$;NJTX3AF34(Y9>"PR9/PS!LZ!#EZ3*@! M&@)N=6UL;V-K+F-O;0``#P```"$/L)`W6`\```"X0`".V(`V%P`@PP/X+HL: M"&YU;6QO8VLN<V-R``!6````(0^ID)0$70````QA0,2``:.!@C9O[(`(@X<% M01H%#R8$068."X8&\;R1`T),'CIE0,"APW$+P1@WNK"0$5%.&3H(<@\>@ M&8-N0+BITX;-FS%K7(QYT\;@'8-Q#&H`&@)R97!L>2YC;VT`<@``#@```"$/ M-I$0_`X```"T",TA/`!U`LTAM$S-(1H(<F5P;'DN<V-R`'(``%<````A#RZ1 M:1AF````#&%`Q(`!HX&"-F_L@`B#A@4.@VG<T`$A(X;!,6W@+&3#HJ`"-6[T M"(01!J)$BA8/)ES8D,88DQ,K&C0H!\08/`;+&'0#0DX9.&SRN!CSIHW!.P;C %&-0`&@!$ UUCUUC