Troubles with Transwarp

[news@serval.net.wsu.edu (USENET News System)]

I recently purchased a Transwarp and a SCSI hard drive with an AMR controller.  I
installed both at the same time (TW in slot 3 and HD in slot 7).  Here is my
system before hand:

Aux. Slot:  100% RamWorks III compatible Q-RAM //e with 1 megabyte
Slot 1:  Apple Super Serial Card (set for interrupts and printer mode)
Slot 2:  ThunderWare ThunderClock Plus
Slot 3:  Empty
Slot 4:  Apple II Mouse Interface
Slot 5:  Empty
Slot 6:  Apple Disk II Interface
Slot 7:  Empty

I put the TW in slot 3, the HD controller in slot 7 and the hard drive was
unable to load ProDOS.  Frustrated, I put the TClock in slot 5 and left the TW.
The hard drive started, but, then halted and would often break (randomly) while
loading the Apple II Desktop v1.1 (aka MouseDesk 2.0).  I finally figured out
that the TClock was conflicting with the TW somehow.  Anyone know why?  BTW I set
the TransWarp to acclerate slot 4 (Mouse) and slot 7 (hard drive).  I then pulled
the TClock and the system responded the same.  Next I migrated the TClock to slot
2 and moved the TW to slot 5.  The system, once again, printed 'Unable to load
ProDOS'.  I pulled the TClock and the TW cooperated just fine (except for what
follows).  I dropped the TClock back in slot 2 and booted the Disk II controller.
Karateka loaded, but, when it animated I noticed 7bit glitches on the screen. 
This had _never_ happened before the new hardware. 

I booted the hard drive and suprisingly enough ProDOS finally loaded.  I went to
run Publish-It! 4.0 and while operating, the program experienced the same
side-effect as seen in Karateka.  Bamfoozled, I ignored the problem hoping it
would go away.  Did it? NO!  I loaded Karateka and the randomly placed bytes
still existed.  I pulled the clock and bytes from Karateka disappeared.  Next, I
ran Winter Games and these bytes existed, though, the bytes were not present in
Karateka.  I guessed that the problem might be related to interrupts and the TW. 
Is this true?

Also, the TW manual explicitly warns against accelerating Clock cards and ACIA
Serial cards since they may depend upon the hardware at 1mhz.  I frequently use
Agate 0.69 for ANSI emulation, and, while watching the screen load, I noticed
that auxiliary portion of memory loaded at the same performance as at 1mhz.  I
ignored this as part of the program.  Later, I noted that Publish-It 4.0 seemed
much slower than a Laser 128EX which also runs around 3.7Mhz.  I finally figured
out that with the clock in the machine, I HAD to accelerate the clock.  The
TClock _was not_ affected by the acceleration process and the aux. memory also
sped up drastically.  Why is it doing this??

As a method of determination, I felt that in order to prove the interrupt conflict
theory I should try this out with the mouse interface.  I then deccelerated the
Mouse card without the Clock installed, and sure enough, the Auxiliary RAM slowed
down to 1Mhz.  Are the interrupts to blame?  and, how can I possibly get the TW
to function in slot 3, with the TClock in slot 5 or slot 2 so as to possibly free
up a spot for a Mockingboard or a UniDisk controller?

Please E-mail...

Michael Kelsey
mkelsey@eecs.wsu.edu