ATA 100 Cable Length

[Guest]

I have a ATA 100 Maxtor disk drive that I would like to mount in the upper slot of a full tower case but the 18in cable is too short. Belkin sells a 24in cable; will it work OK? Maxtor says that the longest length to use is 18in. What kind of problems will I be looking at?

 

[kindlr]

They seriously suggest a longest cable length? It sends data, there shouldn't be a longest length, if it were a power cable it'd be different due to resistances, but data has negligible resistance issues... You shouldn't have any problems with a 24" that I can see.

Me "I bought a TNT2 M64, the BIOS says its a Vanta"
IOMagic "Theyre the same card"
Me "Um, no"

 

[AMD_Man]

Possible slow downs in data transfer. The longer the cable length, the more likely the chance of interference and the more likely data would be corrupted. However, because IDE specs support data integry checking, the only disadvantage will be a slight slow down in data transfers.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[Arrow]

It is more reliable to use a shorter cable length - but as the above posts say, you shouldn't worry too much.

Rob
Please visit <b><A HREF="http://www.ncix.com/canada/index.cfm?affiliateid=319048" target="_new">http://www.ncix.com/canada/index.cfm?affiliateid=319048</A></b>

 

[lhgpoobaa]

well 18" (46cm) is the max reccomended length before the noise gets significant.
however, with modern drive and controllers are pretty good getting over the noise, so im using 24" (60cm) cables and are not experiencing any slowdowns or data corruption...

i wouldnt want to go longer though... there are 36" cables around and people have lots of trouble with them. just too long

and when i put in my 24" cables, i created some monster 1.5Gb files then copied them from drive to drive, and moved them around alot within drives for half a day.
then i did a file compare to check for data corruption.
to be safe i suggest u do the same.

OEMs selling "High End"PCs with integrated video will be forced into Q3tournaments using a TNT2M64!<P ID="edit"><FONT SIZE=-1><EM>Edited by LHGPooBaa on 11/10/01 09:51 PM.</EM></FONT></P>

 

[AMD_Man]

LHPooBaa, the IDE specs check data integrity so I don't think you risk data corruption with longer cables. However, you might experience slowdowns if the IDE controller detects corrupted data and asks the hard drive to resend the data. Also note, overclocking your FSB can also decrease hard drive performance because it increases the chance of corrupted data being send to and from the controller and resent.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[lhgpoobaa]

i beg to differ...

the controller and drive can correct small amounts of errors, the odd bit being sent wrong, but there comes a point where the noise and data corruption is too great. ive seen reviews where people DO get corrupted data on 36" cables.


OEMs selling "High End"PCs with integrated video will be forced into Q3tournaments using a TNT2M64!

 

[AMD_Man]

lol, 36", who needs a cable that long? Even full towers don't need cables that long, they impede air flow if they're too long.

AMD technology + Intel technology = Intel/AMD Pentathlon IV; the <b>ULTIMATE</b> PC processor

 

[lhgpoobaa]

true.
ive no idea why they used them either.
24" is pleanty long enough for a tower case, even with the connector at the bottom on the mobo connecting to a drive at the top.
and no data errors

Why do i feel like the lone sane voice in the mental assylum?

 

[CALV]

LHGPooBaa is right, there IS a max reccomended length, and data corruption DOES occur on long cables, I once made a long one out of ribbon cable- to plug in other peeples drives whilst still in their pc (to make a ghost image for a recovery cd)- had probs till i found out about this max length, seems a short maximum length, but its true.



Next time you wave - use all your fingers<P ID="edit"><FONT SIZE=-1><EM>Edited by CALV on 11/16/01 07:59 AM.</EM></FONT></P>

 

[lhgpoobaa]

holy crap...
just found a place that sells 100cm IDE cables!
thats 40 inches!!!
thats gotta be criminal!

and the reason why ide cables have this problem is that they are unshielded, unlike scsi cables.
as the length increases, the cable acts more and more like a big aerial, with electromagnetic waves going back and forward, disrupting the signal.
ata 66/100 80pin cables help the situation a little, but not much.

apparently other things make it worse...
1. cable length
2. faster interface like ata66/100 is more prone than ata33, despite the 80pin cable
3. overclocking

i believe serial ata will do away with the inherent problems of ata cables.

Why do i feel like the lone sane voice in the mental assylum?

 

[mbetea]

mmm, i wonder why noone's tried to put a few intergrated IEEE ports on mobos. no more bulky POS cables to have to worry about splicing or arranging for air flow. no clunky connectors. and i remember when the transformer watches were all the rage

well if luck is a lady, it explains why i have no luck :frown: