bstempleton@watmath.UUCP (Brad Templeton) (09/01/83)
How many of you have seen the ads in recent magazines for a 10 meg shugart SA-612 for $995 US? Sounds pretty good, so while I was in Chicago last week, I went to check it out. First things first. The disk is advertised for $995 mail order from "Great Lakes Peripherals" near Chicago. Similar ads offer the disk for $1595 through dealers from "Phoenix". They are the same company. The address I went to to pick up the disk was the same one in the $1595 ads. They are essentially screwing their dealers by undercutting them. Business practices aside, I saw the disk work and the company was not a garage, so I gave them my $995 US funds and took mine with me. The people with me were in a hurry so I didn't check it then and there. The technical manual didn't come with it. They are sending it. They are using both DTC controllers and ANADEK (or whatever) and I got a DTC since it is very similar to the one in the XT and I want to write a Coherent device driver for this. I put in in a PC when I got home, and it would not format. It just hung. A friend and I made a careful examination of the interface board that plugs into the PC (This is the only thing these guys make, mostly they are OEMs) and I noticed that there were 4 sets of holes to wire jumpers onto. Examination of te pinout for the PC's bus revealed the first two were send and ack for DMA ch. 1 and the second two for DMA ch 3. The manual said it used #3, but my board had the middle two jumpers wired! That meant it sent out request on ch #3 but was waiting for acks on ch. #1. Pretty dumb, so we fixed it. I then formatted the disk and got 39 bad tracks the first time, 37 the next time, with only 3 in common. When I got their return call, they said, "we know what is wrong with your disk". I told them I found that hard to believe, and explained the jumpers. They admitted they did not know how that got by them, but that the reason for the bad tracks was that DTC had started shipping them new controller cards that had different timing specs without telling them. This meant a few problems on older PC units. They said they are shipping me a new controller board right away. That's the bad news. Now the good news. The thing does work, and it is about 20% slower than an XT. Since DOS seems to make things CPU bound, though, I suspect the fault is their driver. I hope my own driver will make it just as fast. After all, the thing is supposed to DMA 5 megabits per second, but data from even the fastest hard disks seems to come in at only 20 k bytes/second. Otherwise things seem good. Slightly slower seek times for short seeks than a davong, same average time (90 ms). Quite quiet, and the chassis looks very solid and well put together. There is room and power for another disk, tape backup or floppy. It is reaonably quiet, and the ribbon cable you get is quite long so you can shove the thing out of the way where you never have to touch it again if you use a power strip. I guess the best news of all is the $995 price tag. Nothing around seems to beat it. Even if you think it is slow, with the money you save you could buy another spindle (20 megs for $1495 probably less if second disk unit, this one made by Ampex), or buy a whole bunch of ram and make heavy use of ramdisk for all your compiles. In all, I am fairly pleased assuming the new controller is good and my driver is faster. More news when I get the new board. The thing is warranted for a year. It works with dos 1.1 and 2.0, slightly faster under 2.0 -- Brad Templeton - Waterloo, Ont. (519) 886-7304