UDMA2 and UDMA5 ODD on same controller

Has anyone had experience connecting two ATAPI optical drives operating at significantly different UDMA modes to the same IDE controller?

I currently have a Plextor PX-W4824 CD-R/RW and a Toshiba SD-M1712 DVD-ROM operating on my ASUS P4C800-E Deluxe secondary IDE controller. Both drives are UDMA mode 2, set up as master and slave, respectively, and have worked together on one controller with absolutely no problems for years. An Optiarc AD-7170 DVD-R/RW is UDMA mode 4, set up as master alone on the primary IDE controller and I’d like to keep it that way. I want to replace the Toshiba with an ASUS DVD-E616A2 DVD-ROM that’s supposed to be capable of UDMA mode 5 (ATA100) for a little faster ripping speed.

Please don’t suggest that I flash my Toshiba with firmware that removes the so-called Riplock. Tried that once. Worked great until I ran into some badly manufactured discs (one of which was an ATI video card OEM driver CD). I thought it would destroy my drive before I could stop the process! Now I know the purpose of “Riplock” - and it’s not to discourage people from ripping discs. It’s to protect the hardware!

The question is - Will I have any problems with two ATAPI optical drives operating at UDMA2 and UDMA5 on the same IDE controller? Will the slower drive drag down the performance of the faster drive? I know this used to be an issue, at least with ATA HDD and ATAPI ODD on the same controller, but I’m not sure that modern IDE controllers don’t let each drive operate at it’s maximum speed like SCSI controllers. Any thoughts on this?

I personally don’t see a problem, with modern motherboards and OSes.

On my main PC, I have a LiteOn 165P6s (UDMA4), coupled with a Samsung SH-S182D (UDMA2) on my Primary IDE channel. No problems ripping or burning.

You shouldn’t experience any issues with this. I certainly had no problems with my Asus E616 and a Liteon burner on the same IDE channel.

I don’t understand what udma4(66.7MB/s) is really gonna do for an optical drive,
especially with rotational speed limits and small drive caches. 16X burning or ripping maxs at about 22MB/s, way short of udma2 speed.

That are theoretical rates, I wouldn’t always count on them.
(Many) manufacturers have switched to UDMA4 Host Interfaces, so there must be some sense behind it, I think.


I’m not sure either, but for less than $30. USD, shipping included for a new retail ASUS DVD-E616A2, I’m going to try it. I’ve had good luck before with ASUS ODD’s and the old Toshiba SD-M1712 is getting a little tired, anyway.

With the new Optiarc AD-7170A DVD burner operating at UDMA4, I guess I’m thinking my old Toshiba DVD reader (UDMA2, 192KB buffer) may be falling a little behind.

BTW, the E616A2 has a 2MB buffer, not the “small cache” you referred to and it’s supposed to operate up to ATA100 (UDMA5).

Thanks, everyone, for all the help. Keep 'em coming.

2 meg cache to me is small when you do the math, at 16x(22MB/s) 2 megs holds
2/22 (0.09 seconds worth of data)

at 66MB/s second burst ~3/100’s of a second


FWIW, when I replaced the old NEC ND-3500A with the Optiarc AD-7170A and with the Toshiba SD-M1712 DVD-ROM reader in the system, my total processing time (transcode and write using AnyDVD/CloneDVD2) dropped from an average of about 48 minutes to around 38 minutes. Due, I suppose, to being able to write MCC004 reliably at 16X with the AD-7170A and the UDMA4 mode it was capable of operating in as compared to UDMA2 for the ND-3500A.

Swapping in the new ASUS DVD-E616A2 (UDMA5) for the old Toshiba (UDMA2) has resulted in average total processing times of around 19 minutes! Regardless of the reason for the increase in speed, I’d say it’s well worth the $30. USD investment in the DVD-E616A2.

BTW, those of you who said having UDMA2 and UDMA5 ODD on same controller won’t hurt the performance of either drive appear to be right!

Thanks for all the help!

There is no UDMA5 capable ODD. It is just reported wrong.

Modern IDE should have no trouble with multiple UDMA’s on the same channel, I believe. If I remember right, some older chipsets - way older - used to limit the speed of both devices to the slowest of the two, but I do not think that has been the case for a long while now.

It might be reported wrong, but it’s in line with ASUS’s published specs on the E616A2 from their site -

[left]Burst Transfer Rate:
Ultra DMA/100: 100M Bytes/sec max.
Ultra DMA/66: 66.6M Bytes/sec max.
Ultra DMA33: 33.3M Bytes/sec max.
Rate PIO Mode 4: 16.6M B/s max.
Multi-word DMA mode 2: 16.6M B/s max

Per this link:

Last time I looked, Ultra DMA 100 was UDMA mode 5. Seems to make sense here, as my Optiarc AD-7170A which NEC specs as “Ultra DMA 66 mode 4” does, in fact, show up as UDMA4 in control panel/system/hardware.

Yup, I’ve seen the specs.

Have you made an burst rate test with nero cd dvd speed yet?


I finally got a chance to run the test you suggested.

Nero CD/DVD Speed shows a Burst Transfer Rate of 55MB/sec (55962KB/sec to be exact - personally, I’d call it 56MB/sec) for the ASUS DVD-E616A2. I used an older commercially pressed original movie disc for the test and ran it several times with the same 55MB/sec results. The best the old Toshiba SD-M1712 could manage was around 28MB/sec.

Without making a career out of testing this thing, I’d have to say that I’m more than pleased with the performance of the new ASUS as compared to the old Toshiba.

I think I’ve finally managed to figure out how to attach an image of Nero’s CD/DVD Speed Benchmark test for the ASUS DVD-E616A2. Comments welcome!

Not sure how you created the image but the simple, recommended method is to save the image as a png file directly from CD-DVD Speed (use the floppy icon, top right) & then just upload it. The image is then smaller than the one you’ve posted.

On the burst rate test I’d say that’s better than UDMA 4 as I get about 47mbs on my Pio & Liteon UDMA 4 burners.

IIRC, I’ve got 48 or 49mb/s with the Samsung 163 SATA burner.

Hmmm, I don’t know how to judge about that burst rate (I can confirm your values), as this drive reports itself as UDMA2 drive. Maybe this is just the transfer rate from the controller to the SATA/PATA bridge…


does burst have any real world relevance?

I don’t know! Chef seemed to be requesting a test, so I gave it to him. Real world relevance to me is my previous statement: “Without making a career out of testing this thing, I’d have to say that I’m more than pleased with the performance of the new ASUS as compared to the old Toshiba.”

Real world tests will be ripping dual layer dvd’s and comparing with that nec for speed and accuracy. If the asus can slow down and read damaged disks that’s better than
reading too fast.