PX-755A writes corrupted bytes to DVD+-R

[qanda]This thread is about the Plextor PX-755A. Click here to see full specs[/qanda]Hello,
first of all - I am new to this forum, but not to this world.

I have one very strange (and much more annoying) problem with my Plextor PX-755A (TLA 0001, current firmware 1.08), which I bought brand new about 14 months ago.

I RECENTLY noticed that my drive sometimes writes some corrupted bytes to DVD±R discs. I do not know if this had happened anytime before, since I did not notice until recently. Imagine my FRUSTRATION when I realised that all discs which I have written since the purchase may be faulty!

DETAILS:

By performing a byte-to-byte comparisions of the content of just-written DVDs against the source content, I realised that in some cases there are few (sometimes a few dozens) bytes which are not the same as their source copies, although all those DVD discs have EXCELLENT after-write performance measuraments, without any glitches (PI/PO, etc.)!!!

Those bytes ALWAYS differ in a way that the source byte has its highest bit =1, while the written byte has its highest bit = 0 (highest bit inverted from 1 to 0). This is the only difference which may appear. So, these are the possible combinations of “defects” (hexadecimal notation, and X means “any hex number”).

SOURCE WRITTEN
BYTE BYTE

8Xh 0Xh
9Xh 1Xh
AXh 2Xh
BXh 3Xh
CXh 4Xh
DXh 5Xh
EXh 6Xh
FXh 7Xh

PARAMETERS:

  • I mostly use Verbatim MCC 004 (DVD+R) discs, but I tried a few MCC 03RG discs and a few cheapo discs, and got the same behavior regarding this problem;

  • I use my drive in a VIA KT133-based PC (1.1 GHz), but I also tried it in a 440BX-based PC (1 GHz). Both of them are NOT overclocked, have 100% error-free RAM (tested with MemTest), top-quality 350W Enermax PSU’s and different PATA cables. The drive had the SAME issue on both PC’s.

  • Both of these PCs use Windows 2000 Pro SP4 with latest updates (trough AutoPatcher 2000). ASPI is v4.60 (ForceASPI), UDMA is used.

  • I used ImgBurn 2.4.1.0 and PlexTools Professional v2.35 for writing. The issue appeared with both of them!

MEASUREMENTS:

  • It SEEMS that the issue DOES NOT appear when writing CD-R, CD-RW and DVD+/-RW discs at ANY speed!

  • After I “detected” this awful problem, I deliberately spent around 20 DVDs (10 MCC 004s, 5 03RGs and 5 some cheapos) just to test this issue. It SEEMS from those test that the drive WORKS FINE when writing at 6x, but most of DVDs written at 8x had 5-15 errors (distributed everywhere on the media), every DVD written at 12X had around 30-60 errors (distributed all over the place), and a few which I wrote at 16x had several hundred corrupted f*****n bytes all over the media.

MY WEAK CONCLUSIONS:

  • The problem seems to lay in the drive, since it exhibits on two PCs with completely different hardware, and with two writing applications.
  • The problem SEEMS to be related only to DVD discs, and only when writing at 8x or faster (I wrote a few DVD+RW discs and one DVD-RW disc about 20 times at 4x and no corrupted bytes occured).

THE SOLUTION?

The simple solution would be to return the drive to the shop (since I still have the 2-year warranty), but they would give me PX-800A, which I can not accept. That’s like replacing an older Audi A8 for a newer WV Golf.
I insisted on receiving a PX-7xx drive, but they got a reply straight from Plextor saying that they have not PX-7xx drives in stock at all.

So, my plan is to first find out if I can fix this drive somehow (that’s why I am here). If not, I will replace it for PX-800, sell PX-800 immediately and hunt for a good PX-716A on Ebay (but that will be a hassle since it’s a lottery to get a drive which is not worn off and with a newer TLA).

So, PLEASE help me to solve this ugly issue!! I can’t believe that I am experiencing this, it must be a curse! I still have nightmares when I think about dozens and dozens of my written DVDs on the shelves which may or may not have corrupted bytes …

P.S.: I will provide any additional needed info.

Could you describe the method you use for your “byte-to-byte” comparisons?

ImgBurn does come with a verify-option. Have you used this and did it report any errors? If it did not report errors did you find errors with your own method?

Plextools Pro does not have a verify option, but PTP LE has. You can download PTP LE from www.plextools.com and give it a try.

Thank you for your response. I was afraid that nobody would read my post because of its size (but the forum rules and the common sense tell me to write as much useful details as possible …).

First of all, I read the written discs both on the PX-755A and on another DVD-ROM drive. They always read the SAME (I extract TAO/ISO images from both drives and then compare them using the FC command-line application).
So, I am confident that the PX-755A always reads CORRECTLY (it does not “produce” corrupted bytes when reading, only when writing).

I used several methods for comparision of written discs against their source content.

If the disc was written from an image, then I would verify it against that image in ImgBurn, or create a new image from the written disc and compare those two images using “FC /B image1.iso image2.iso”. Both ImgBurn and FC ALWAYS (of course) reported the SAME results (only in different “notations” - ImgBurn in sector+offset number, FC in absolute byte offset).

If the disc was written from a bunch of files, then I would compare two folder structures (the one on the written disc and the one on the source HDD) using Total Commander (Menu “Commands -> Synchronize dirs …”) just to see which files are affected.

I hope that this answers your question.

  • Bosanek

Several days have passed and yet nobody answered. This is really demoralizing. :frowning:

Am I really having some utterly unique, “never heard of before” problem? Please try to help me by giving any hints or ideas or suggestions or …

I have contacted Plextor’s technical support TWICE about this issue, and got NO reply both times. :frowning: :sad:

If nothing else, please direct me at some other forum where I might have better chances of finding an answer (although I don’t think that there’s a more suitable forum for that than this one).

Have you seen bytes being changed within the files written to the disc, or just bytes being changed in the part of the image that doesn’t contain file data?

Welcome to CDFreaks! :slight_smile:

[QUOTE=Bosanek;2156076]

  • It SEEMS that the issue DOES NOT appear when writing CD-R, CD-RW and DVD+/-RW discs at ANY speed!

  • After I “detected” this awful problem, I deliberately spent around 20 DVDs (10 MCC 004s, 5 03RGs and 5 some cheapos) just to test this issue. It SEEMS from those test that the drive WORKS FINE when writing at 6x, but most of DVDs written at 8x had 5-15 errors (distributed everywhere on the media), every DVD written at 12X had around 30-60 errors (distributed all over the place), and a few which I wrote at 16x had several hundred corrupted f*****n bytes all over the media.
    [/QUOTE]

I’ve given this some thought and I think the above passage holds the key to the solution. Look at these data-rates:

CD-R (max)
48x: 7.200KB/s (CAV)
CD-RW (max)
24x: 3.600KB/s (PCAV)
DVD-R (at 16x and 6x)
16x: 22.160KB/s (CAV)
6x: 8.310KB/s (CLV)
DVD+RW
6x-8x: 8.310-11.080KB/s(PCAV)
DVD-RW
6x: 8.310KB/s (CLV)

Notice that according to your own observations your problem is at datarates > 8MB/s only. So something in your setup doesn’t like speed. I think it’s unlikely the drive, since reading (also at higher speeds) seems to work OK. This, IMHO, points to data-corruption on computer side. Something can’t handle high data rates.

Do you have any special VIA or Intel IDE drivers installed? If yes uninstall them and let windows install the Microsoft IDE drivers. Also make sure you’re using the latest chipset software provided by VIA and MS.

Years ago there was a VIA driver (4 in1 or something like that) that caused data corruption. You might want to use Google to search for reports about it.

Thank you both very much for your replies.

@DrageMester:

All corrupted bytes belong to files themselves, NOT to the ISO/Joliet/UDF filesystem’s metadata. I double-checked this by directly comparing files from written discs against their “originals” which are on HDD (by using FC or Total Commander).
The conclusion - this is a real data corruption, not a filesystem difference.

@hwp:

I forgot to mention that I have only high-speed CD-RW (8-12x) and 4x DVD+/-RW media.

So, I wrote CD-RWs at 10x max (around 1500 KB/s) and DVD+/-RWs at 4x max. (around 5500 KB/s IIRC). All other max. speeds which you pointed are correct.

As for the idea of chipset or driver being a cause, I doubt it. This is why:

  • As I said in my first post, I tested the drive on a VIA KT133-based PC (my main PC), and also tested it on a Intel 440BX-based PC (my secondary PC). Both PCs are NOT overclocked, and I tested their RAMs using MemTest86. It can’t be that both of these chipsets are bad, especially not 440BX.

I am using VIA Hyperion chipset drivers on my primary PC (the “old” Hyperion version which is recommended by VIA to be used on KT133 and similar chipsets), and AFAIR I don’t use any Intel’s chipset driver on 440BX PC because IIRC I tried to install it when I installed Win2K in last spring, and it refused to install because it had nothing to install (O.S. already had everything it needed). ForceASPI 4.60 is installed on both PCs, but I tried ImgBurn using SPTI too (besides using PlexTools also).

I believe that this set of different PC environments is enough to prove that the drive is the faulty element, but if anybody still thinks that something else could be, please say so. Also, if the high transfer speeds (8x DVD or more) are a “burden” for a chipset or driver, why doesn’t it then corrupt data when reading at high speeds, or when working with hard drives, which operate at 20-60 MB/s?

Is there any other information which I can provide?

I once had a problem with my PC writing/reading wrong data to/from my harddiscs. It didn’t happen very often but it was a wrong single bit every few gigabytes.

i cannot say for sure what the exact reason of this problem was, but I could fix it by reducing memory timings from 400 MHz (rated speed) to 333 MHz (underclocked). As far as I remember, memtest ran fine on the PC despite the errors.

If you haven’t already tried switching the data cable then I suggest you do so. You’ll need an 80-pin IDE cable.

@Liggy:
If RAM or some other motherboard-related problem is the cause, I would have to be VERY unlucky to have two completely different PCs with such problems.
But I will try to test the drive on a third PC, if I manage to find a friend who’s willing to let me dissect his PC.

@DrageMester:
PCs have different PATA cables. My primary PC has 80-pin PATA cable (drive was/is working in UDMA mode 4), while the secondary PC has 40-pin PATA cable (drive was working in UDMA mode 2). While testing, the drive was the ONLY device on the cable (no slave present), to eliminate possibilities of any interferences from additional device.