Diagnosis Req: SafeMode=DVD ok, Normal=not

I would especially appreciate ideas from the experienced minds here…(not new at this myself, but here’s a stumper).

Symptom: Writing to DVD using any of Nero (6 something), NTI (6.5 something), Sonic (it came with the drive) - all yield non-verifiable results. In particular, larger files have corruption, in that they don’t match the original files on the hard drive.

Nero’s “quality” scans of the resulting disc are actually good. No PO errors, PIE and PIF within norms. Most small files write ok, anything over 300 Mbytes is hit or miss.

If I boot in safe mode (win2k), I can write without problems (other than having to boot in safe mode).

Boot normally (win2k, sp4, etc) - verification failures happen every time.

Been testing with Ridata 4x +RW’s for a few hours so I don’t waste a complete stack of +R’s. Quality scans are consistently reasonable, considering the +RW has been through 50 write sessions. Same disc records fine in safe mode, verification fails under normal boot.

Other background info: This OS and the OEM Benq 1620 have been working without flaw for at least a year. This problem appeared about 2 weeks ago.

My thinking is that it’s software / os configuration. Not sure if WMWare could have been at fault, but I removed it without behavior change.

Reset the ASPI (Adaptec 4.60).

Turned off the anti-virus (you never know…)

So far, I haven’t found “it”.

Put the BenQ into another machine - it writes without problems.

Computer is an oldie goldie M7D K7D with twin MP2600’s.

I’m about to try an alternate installation of the OS, on a fresh hard disc, to see if the problem vanishes (my thinking here is that a really need to prove a normal boot mode write works on this machine, otherwise I could be chasing RAM or some other hardware problem).

Thanks in advance for any ideas

…minor update on my problem:

I found that by switching the IDE controller for the DVD burner from DMA to PIO mode, it will burn 4x media (the +RW’s) just fine…

What’s up with that?

Data corruption is normally a Ram problem or a bad IDE cable (but that is not written in stone). I would try swapping the IDE cable first and if that doesn’t help download memtest86 to check and see if you have any memory problems.

Yes, thanks, back to checking the basics…

And for anyone else reading through, here’s what happened:

It was RAM, but from an odd reason for the failure.

It appears that a slightly gummy buildup of dust, blown onto the motherboard right where the RAM socket sits by the CPU’s fan, was conducting just enough to cause occasional RAM error. It failed under Memtest86, but the OS otherwise seemed to run just fine.

I didn’t notice it at first…swapped in another RAM chip, same problem.

Dusted off the motherboard with a dry, small paintbrush…works like it was brand new.

Glad to hear you got the problem licked…

That is one problem I haven’t come across so far. but one to remember for the future in case i do. :smiley: