Jeff Vian: >>> If you are jumpering the drives any way except as cable select (cs) >>> then you may encounter problems when using them on a CS cable. You >>> *will* encounter problems if a drive jumpered for master is connected >>> to the slave connector, and vice versa. You *may* encounter problems >>> with a drive jumpered as master and connected to the master connector >>> or when jumpered as slave and connected to the slave connector (I have >>> seen this myself). Tim: >> You will ONLY have problems if you try mixing a drive set for cable >> select with one that's not. You will NOT have any problems putting two >> drives each jumpered as slave and master on a cable select cable in Mike McCarty: > This is not quite my experience. As I mentioned, Connor drives used > to be notorious as not running as slave to other brands of drives. > > Also, I don't like running a hard disc as a slave to a slower device, > like a CDROM. In me experience, it can lead to timing problems on > the cable. I won't argue with that. My beef was about the contention that jumpering or cable-select configuring of master and slave will be a problem. Whichever method was used, there are some drives that don't work well together. Cable-select only allows master/slave configuration of drives, if you look at some drives they have jumpers for variations of master and slave (single drives, compatibility modes for working with other brands of drives, etc.). I think anything with those extra options is a bit of a warning that there may be more to simple configuration of which is master or slave. >> either of the master or slave positions. All a cable select cable does >> is jumper or open pins on the IDE connector, just the same as you'd do >> with the jumpers next to it. It makes no difference to the data in the >> cable, and drives pay no attention to the cable select pins unless >> they're jumpered to do so (or are badly designed). >> >> Just about all high-speed ATA cables are cable select cables, it's >> almost impossible to find ones that aren't. Using one doesn't force you >> to jumper your drives to be cable selected. > That's what the spec says. I dunno why someone else commented that > using a CS cable with M/S jumpered drives could cause a problem. That's what I thought, too. It smacked of someone not understanding cause and effect after doing something, presuming that they knew what happened, but really misunderstood. e.g. I had some computer twerp trying to tell me that on IDE cables which drive was master or slave was determined by the drives and timing. That the furthest cable from the host was automatically the master because it made some timing detection. Of course, that's a load of rubbish. I can see one advantage of cable-select - for those who don't understand master and slave configuration of drives, they can just plug in the pre-configured as cable-select new hard drive they just bought into the second connector, and hope that it'll work. Of course, if they try plugging things in the wrong way, swap drives around, or their existing drive sits on the slave plug but jumpered as master, etc., it'll fail and they won't know why. Cable-select would only work in a foolproof manner if it was the only way that drives were connected. Personally, I prefer SCSI (able to have 7 devices, or more, on a common bus), but they're rarer and more expensive. Though the need for multiple drives has diminished somewhat with the huge drives we have available these days. -- Don't send private replies to my address, the mailbox is ignored. I read messages from the public lists.
