40 vs 80 wire IDE cables for CD drives

I understand there are issues using 80 wire IDE cables for CD burners. It is suppose to be well known, but I haven't run into this yet or have seen any threads on the subject.

What is the deal, is there issues and what?
 
40 wire cables for UDMA 33 and UDMA 66 devices
80 wire cables for UDMA 100 and UDMA 133 devices

CD-ROM, DVD-ROM, CD-RW and DVD-RW drives aren't UDMA 100/133 devices;


Greetings from
Duracell
 
If u have an ata100/133 compatable motherboard u can use 80 wire cable for any IDE device.
There is no benefit in doing it for cd/dvd drives as far as I'm aware, but if thats the only spare cable u have then it will work perfectly fine.
All my drives are on 80 wire with no problems what so ever.
 
Not Quite....
40 is good for modes up to ATA/33
80 is required for ATA/66 or above

Most controllers force a maximum of ATA/33 on a 40 wire cable - and this is the highest mode I've ever seen an optical device support.

If you are forced to share an cable between a HD of ATA/66 or better, and an optical device, an 80 wire cable is required - though I seem to recall that Ultra devices may be restricted to no more than ATA/33 if there is a non-Ultra device on the same cable - though this may be a myth derived from the 40/80 wire handling!
 
daveml said:
If u have an ata100/133 compatable motherboard u can use 80 wire cable for any IDE device.
There is no benefit in doing it for cd/dvd drives as far as I'm aware, but if thats the only spare cable u have then it will work perfectly fine.
All my drives are on 80 wire with no problems what so ever.

you are correct daveml!!! ;)
 
Last edited:
though I seem to recall that Ultra devices may be restricted to no more than ATA/33 if there is a non-Ultra device on the same cable - though this may be a myth derived from the 40/80 wire handling!
That is another question, if you put a CD drive on the same bus as a HDD will that affect the xfer rate of the HDD (slow it down)?
 
Two issues on speed, with OLD controllers, two drives ran at the speed of the slowest, but modern DMA mode controllers tend to support split speed working - though Ultra modes may be restricted for reliability/stability reasons if a non-ultra device shares .... I had an ATA/33 and an older (16) device sharing, and it was not very stable until I lowered the Ultra/33 device to 16 as well.

The second issue, I forget where I saw it tested, found that although dissimilar speed devices tend to produce their own fastest speeds used individually - if simultaneous access isw attempted, the slower device appears to drag the faster one down to the same speed (bus occupancy/arbitration problem?)
 
I too have 80 wire cable on udma 33 devices with no problems, there are however CD-ROM drives emerging which are UDMA 66 so I would only us 80 wire on those personally.
 
That seems to make sense. I still see people saying you can't mix devices at all.
Why is there a limit to the number of IDE channels a MB will support without going into another type of Raid type controller (Highpoint or Promise)? Why can't there be IDE 3? Is it a O/S issue or bios issue?
 
I guess the chipset makers don't deem it neccessary to support more than two IDE channels natively within their chipset.

Tertiary IDE is definitely possible - in the days of CD-ROM interfaces on soundcards, there were soundcard which could provide a 3rd IDE - though the port and IRQ expected were not really standardized.

Way back, you had to have a card for just ONE IDE - and that was ISA bus. With the inegration into chipsets, dual IDE became the norm, and the port and IRQ mapping was pretty well set in stone at that time - It would have been much better if the original PC design had been buried, and a new start made - and if Apple had licenced clones at reasonable prices, we'd all be using Apple compatibles, they would be the leading software house, and Gates would be just another spotty geek.
 
Top