Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!leah!rpi!batcomputer!cornell!rochester!rit!tropix!moscom!ur-valhalla!uhura.cc.rochester.edu!sunybcs!boulder!ncar!unmvax!tut.cis.ohio-state.edu!att!mcdchg!ddsw1!karl From: karl@ddsw1.MCS.COM (Karl Denninger) Newsgroups: comp.unix.xenix Subject: Re: source for WD1007WA2 Summary: WD1007A is a very flexible board; let's not knock it folks! Message-ID: <3611@ddsw1.MCS.COM> Date: 22 Jul 89 15:54:12 GMT References: <3072@gaboon.UUCP> <5343@tekgvs.LABS.TEK.COM> <3607@ddsw1.MCS.COM> Reply-To: karl@ddsw1.MCS.COM (Karl Denninger) Organization: Macro Computer Solutions, Inc., Mundelein, IL Lines: 62 In article <3607@ddsw1.MCS.COM> nvk@ddsw1.MCS.COM (Norman Kohn) writes: >In article <5343@tekgvs.LABS.TEK.COM> keithe@tekgvs.LABS.TEK.COM (Keith Ericson) writes: >>WD has (at least) two versions of their 1007 ESDI controller; one of >>them is only capable of a 2:1 interleave and should be avoided if at >>all possible. The later version is alleged to be capable of 1:1 >>interleave but I haven't evaluated one to verify this. >I've heard that the 1:1 is not yet available, but should be soon. >I've also been told, however, that WD's ESDI card gives 1/35 fewer >sectors than Adaptek's, presumably because it allocates one alternate >sector per track (for bad sectors found at format time) and that this >feature cannot be disabled. Adaptek's is allegedly better for unix. Not true. The WD1007 has several set-ups available. It supports 1:1 interleave, first off. Secondly, you have a choice of translation mode enabled or disabled, and a choice of three (usually) translation settings. There are some catches though, which you need to be aware of: If you turn OFF translation and the BIOS, you need an external format program. This is not normally a problem, and it will allow you to use the full 35 sectors per track. HOWEVER -- your operating system must be able to deal with this, and map out the defects. Many cannot do this right (386/ix comes immediately to mind :-() If the BIOS is enabled, you are normally running at 34 sectors per track, with the 35th being set up as "sector 0". The controller is smart enough to reformat a track so that a defect (if it occurs) is in sector zero, thus making it invisible. This only works if there is ONE defect in a given track. The BIOS also has a mode to spoof the drive into looking like it has less than 1024 cylinders, even when it does not. This can be VERY useful if your operating system will not install without an exact match in the ROM drive tables, as these tables will not go out beyond 1024 cyls. Lastly, be aware that the BIOS sometimes decides to write to the CMOS memory on the system. This can cause problems with machines that have an "extended CMOS" setup, allowing changes to things like shadow RAM and I/O speeds. You'll know if this is happening because you'll get checksum errors in your CMOS (grrrrr). IF this happens, you will need to disable the onboard BIOS completely, although you can do it AFTER formatting (providing your OS can handle > 1024 cyl drives; Xenix can do this) and still get what appears to be an error-free device. >On the other hand, if unix 386 only allows a fixed number of bad sectors >(64 in uport) and an entire track is bad somewhere, with large >hard drives the table could fill up pretty quickly. This is true; you can get around this by having what appears to be a "flaw free" disk easily..... I LIKE this controller; it works great, and is very fast. Adaptec has been problematical for me in the past. DTC, which also makes ESDI boards, hasn't managed to get one here for evaluation yet, so I have no idea what their boards are like. -- Karl Denninger (karl@ddsw1.MCS.COM, !ddsw1!karl) Public Access Data Line: [+1 312 566-8911], Voice: [+1 312 566-8910] Macro Computer Solutions, Inc. "Quality Solutions at a Fair Price"