System tests utility in ECDC...what's the usefulness of it if the results are never the same?

Hi! Thanks for your attention.I'm using the cd copier portion of ECDC (began with version 4.02d and
now since summer 2002 using 5.1) since almost three years to make backup copies of all my disks
successfully, and i keep them somewhere else than at home just in the case of fire,theft...etc...
Since two weeks, i have a new machine i've built (ASUS p4p800-vm, Celeron 1700Mhz, 256Mb DDR 400Mhz,
Win98SE) because the other machine (IBM Aptiva, AMD K6-2 400MHz, 256MB pc100, Win98 1st edition) had
a big stability problem due to a slowly dying motherboard. I then built this new machine in a new
standard ATX case with a decent Antec 350Watts power supply, bought the p4p800-vm mobo with a
Celeron CPU, and reused the same drives that were in the Aptiva: Maxtor Diamond Max plus9 80GB
7200rpm HDD(primary master), LG gcr-8523b 52x cd-rom drive(primary slave), Sony crx-140E 8x-4x-32x
cd-rw drive(secondary master). Now to go finally to the point, i can't copy anymore a disk from the
cd-rom drive toward the cd-rw drive(on the fly copying),a thing i could do without any problems with
the old IBM Aptiva. The program simply freezes at a random time and the drives stops spinning. On
the interface, i can see 1% left from buffer so i suspect that the cd-rom dive can't send data fast
enough towards the cd-rw drive, a thing i find surprising (copying from the HDD is not a problem,
the buffer always stay in the 100% range,then). Since the first coaster a few days ago, i'm now
only doing write tests (without actually burning anything on the CD) that fails most of the time.

During the unsuccessful test, i can see that the buffer percentage never sticks to 100% but tends to
drop most of the time in the 80~85% range, 'till the data stream is probably slowed enough to cause
a buffer underrun situation. If i lower the write speed from 8x to 4x, then 3 out of 5 write tests
are successful, but this situation is still unreliable. Any idea where to troubleshoot? By looking
in ECDC's System Tests results i can see that the cd-rom drive is "supposed" to read data faster
than the selected 8x write speed of the cd-rw drive: small files= 9x, large files= 15x, read
speed=17x. I did another system test for the data read speed of the cd-rom drive and guess what? The
results displayed aren't the same at all than those displayed before the test! This pushed me to do
further testing with four different data CDs. It took some time, but it was interesting to see how
much the results may vary with each data CDs.
For example with the CD#1 i had: small files= 9x, large files= 15x, read speed=17x
with CD#2= small files= 8x, large files= 16x, read speed=18x
with CD#3= small files= 6x, large files= 33x, read speed=20x
withCD#4= small files= 3x, large files= 21x, read speed=17x
If the results vary so much depending on wich CD used for that test, what's exactly the usefulness
of the System Tests utility,then? Another thing that i can't understand is why a 52x max speed
cd-rom drive can have a read speed reported so low as 17 or 20x? And to complete this post, besides
when copying on the fly from a cd-rom drive to a cd writer (a thing that missed me suddenly!), do
you know under wich circumstances those tests results are used by ECDC?
TIA for any reply and your time to read this.


--
Alain(alias:Kilowatt)
Montréal Québec
PS: 1000 excuses for errors or omissions,
i'm a "pure" french canadian! :-)
Come to visit me at: http://kilowatt.camarades.com
(If replying also by e-mail, remove
"no spam" from the adress.)