Transwarp //e woes

[DAKIDD <Dakidd@cris.com>]

Hello...
OK, first my machine setup:
Slot 1: GPI parallel printer interface card with Star SG-10 printer
Slot 2: Apple SSC with a Zoom 14.4V modem attached
Slot 3: Transwarp //e
Slot 4: Empty
Slot 5: Apple Unidisk Controller card with one Apple Unidisk 3.5" drive
Slot 6: Apple Disk controller card with 2 Apple Unidrive 5.25" drives
Slot 7: Apple SCSI controller, with Apple 40SC hard drive into two 20Meg 
Prodos Partitions using Apples SCSI card utilities partitioning program.
 
Aux   : QRAM //e 1 Meg memory expansion card

No-Slot-Clock installed under the C/D ROM chip

Base machine is an enhanced //e.
 
OS is ProDOS 8 V2.01 with the NAMEFIX patch from the 2QWK! package installed.
(And before you blame the patch, the problem I'm about to describe 
started BEFORE I applied the patch) and a custom-written ATINIT file (but 
no Appletalk) that loads the NSC driver code over the standard ProDOS 
ThunderClock driver at boot time.


The problem: From time to time (with no real predictable pattern) I find
that the boot partition of the hard drive has had the second 256 bytes of
block #2 overwritten with $40 ('@') characters. It is entirely filled with
them. File data is left intact for any file entries that occupied that
block, but of course, there's no way to read it because the pointers,
name, etc. have gone away. I'm adept enough to retrieve things lost this
way, and I keep backups current, but this happening usually means almost a
day worth of fishing around on the drive trying to retrieve the last week
or so worth of changes that have been lost, moving them to the second
partition, restoring the backup, putting the changed stuff back over onto
the first partition, etc. 
 
Translation, it's inconvenient as all get-out, but I almost never lose 
anything that I really want to get back.
 
The question:
WHAT THE HECK IS CAUSING IT?!?
 
I seem to recall that when I plugged in the TW//e, that was the beginning 
of my troubles. I've run it for extended periods without problem, only to 
have the disk-wipe kick in on me. I've run the machine with the TW//e 
disabled. I STILL sometimes get it. I've swaped out the memory expansion 
and replaced it with the standard Apple Extended 80-col card. I get 
errors (and lose my nice big ProTerm scrollback and editor, and most of 
Appleworks, and...)
I've tried everything I can think of, including strapping heatsinks to 
every chip that isn't nailed down, freeze spray to check for thermal 
intermitents, the whole shebang. Nothing has isolated the problem, and I 
still get disk-munches happening. (Last night was the latest one)
 
Two programs, (dare I say it) ProTerm 3.1 and the Manx Aztec C 
development system that I use, seem to be the programs that most often 
tell me that something has gone wrong. ProTerm by crashing 
unceremoniously into the monitor as it goes after something on the 
affected disk, and the C system by simply hanging and dying until I try a 
control-reset, at which point, it too crashes to the monitor.
 
A three-finger salute at this point invariably returns the dreaded "Can't 
load ProDOS" message.

Does ANYBODY have anything even remotely like this that happens on their 
system? I'm virus free, as far as I can tell, And I've heard nothing of 
any bugs that cause this. Am I having a hardware problem, or it it software?
 
I've run memory tests on all of it that I can test (everything except the 
TW//e's memory) and I get back perfect results just fine -- UNLESS THE 
TRANSWARP IS LIVE. Then, ALL of memory tests fail, but don't give me any 
kind of indicationg what the problem is. The Self-test, whether with or 
without Transwarp active, invariably tells me that all is well.
 
I can find no nasties like shorts on the PC boards, and I've got a System 
saver sucking the air through the rig for cooling. I'm at wit's end. I 
DON'T want to have to try to find another //e (and I'm NEVER going to go 
blue, and want nothing to do with M*C unless I'm forced to it), and GS 
prices are ridiculous in my area from what I've seen. I WANT to keep this 
rig going for a while longer, but I'm beginning to worry that it's on 
it's last legs. If anybody can help me pinpoint the problem, I'd like to 
hear about it. 
 
Drop me a followup here, or Email me at DaKidd@cris.com or Dakidd@grfn.ORG
The cris.com address is preferred, but I still check mail daily at grfn.
 
Many thanks in advance for any help anybody can give me.


Apple // FOREVER! IBM? TARGET!