chip@ateng.ateng.com (Chip Salzenberg) (03/08/89)
A long time ago in this newsgroup, I posted a query about power-on problems with Everex Step 20 motherboards. In particular, I noted that the Rev E motherboards would sometimes fail to power-on properly when an attached printer or tape drive was turned on first. I received several replies, and I'm thankful for all of them. Here is a summary. -------------------------------------------------------------------- From: blew@tc.fluke.COM (Bob Lewandowski) I posted a request for info a while ago on the Step machine and I got several responses, some indicated problems with the bus compatability because of speed, overheating of the Weitek chips, a problem with SCSI host boards from WD and Adaptec, and problems running some versions of Unix. Your problem is news to me, but all the other problems were associated with older versions of the mother-board (prior to Rev G). What you describe sounds like a problem I ran into a while ago with a similar situation where when the printer was on you couldn't turn on a device that the printer was connected to. It was a problem with power coming back into the off device via its printer port driver logic. The printer would feed power back to pull the off device's +5V bus up enough so that it's processor wouldn't reset properly when it's power was turned on, causing it to go into outer space. We ended up putting in some diodes to isolate the supply of the off device from power coming back in it's printer output pins. It was a kind of kludgey solution but it did solve the problem. The diodes went between the +5V supply and the Vcc pins on the packs that were driving the printer port. Resistors could be put between the Vcc pins and ground to keep a minimum current through each of the diodes. You can just pull pin 16 on the driver packs out of the board and insert the diode with the anode in the hole in the board and the cathode to the lead. [A simple solution to a weird problem.] -------------------------------------------------------------------- From: Clarence Dold <cdold@starfish.Convergent.COM> I have an Everex Step386/20, running SCO Xenix 2.3, with an Epson FX286 parallel, an Archive 150MB SCSI-QIC, a Micropolis 145MB SCSI, a Future Domain SCSI controller, with Xenix kernel mods, and an Arnet smart 4- port board driving Link MC-10 terminals and slave printers. -------------------------------------------------------------------- From: uunet!ames!mips!sgi.sgi.com!calcite!vjs (Vernon Schryver) I have an Everex 386/20, with an Archive tape (Everex "Excel Stream 60"), as well as two parallel printers. The dealer who sold it to me now says "concerning adding Weitek...you're right to worry about heat...upgrade from Rev E to Rev G for $", so I presume I have a Rev E board. (That is not a complaint--I heard the rumors before I bought the machine). The number on tag on the back of the box has an 'E', as further evidence. (Sorry, but I don't feel like tearing it apart to check the foil in order to say for certain.) [So the Rev E boards have heat problems... interesting!] I have never seen the sympthoms you describe. -------------------------------------------------------------------- From: uunet!dalsqnt!vector!chip (Chip Rosenthal) [This story is a bit strong for re-publication. Suffice it to say that he had a bad experience with an Everex motherboard that wouldn't work. He's switched to Micronics and he's happy now. This is the only such story I received, compared to several that reported (almost) no problems.] -------------------------------------------------------------------- From: uunet!uunet!aablue!jb You're not the only one with problems on Everex hardware, although mine seem to be more of the ghost variety. All of my machines, so far, have been Rev E mom-boards. We have only been using Tecmar QT series tape drives, including externals, and I have not experienced any problems if they have been on when the rest of the system is powered up. I would lean towards the controller on your tape drive confusing your bus with stray signals or possibly even a shorting condition. ie. what happens if you leave the tape off, boot, turn the tape on, and press the beloved red reset switch behind the door? [It works: Reset brings a hung Rev E motherboard back to life.] -------------------------------------------------------------------- From: dave@micropen (David F. Carlson) We have used Steps here since last spring. We have had *1* customer report of this sort (rev E) and were unable to duplicate the problem here (Everex QIC-02 tape card.) By my experience, not all revs of E boards will fail with certain devices attached, but those that do [have lots of problems]. I'm glad I know what this little deal is now. [So some problems are not common to *all* Rev E motherboards.] -------------------------------------------------------------------- From: jbayer@ispi.UUCP (Jonathan Bayer) [Jonathan suggests that the "tape reset" command will notify the driver that the tape drive has just been turned on. We tried it, and the tape drive did in fact reset, but then we couldn't do I/O. Sorry, we lose.] We're running a Step 25 here, with no problems such as you described. ======================================================================== So, folks, that's the story. It looks like: 1. Rev E motherboards sometimes fail. Some failures are heat-related. Some failures just disappear. 2. Rev G motherboard seem to work much better than Rev E. 3. If you've got the hang-on-power-on problems, get an upgrade to rev G. Thank you, and goodnight. -- Chip Salzenberg <chip@ateng.com> or <uunet!ateng!chip> A T Engineering Me? Speak for my company? Surely you jest! "It's no good. They're tapping the lines."
chip@vector.UUCP (Chip Rosenthal) (03/12/89)
In article <1989Mar7.141704.14389@ateng.ateng.com> chip@ateng.ateng.com (Chip Salzenberg) writes: > From: uunet!dalsqnt!vector!chip (Chip Rosenthal) >[This story is a bit strong for re-publication. Suffice it to say that he > had a bad experience with an Everex motherboard that wouldn't work... Ahhh...good taste strikes again. Let me hasten to add that not only did I change manufactuers, but I changed suppliers. The people I was dealing with had no concept of how to deal with a non-DOS machine. Their techs were so useless that *I* spent the time in their back room trying to debug the machine. Although I have had enough bad experiences with Everex to avoid them, in this case it might have been the change in stores which did the trick. -- Chip Rosenthal chip@vector.UUCP | Choke me in the shallow water Dallas Semiconductor 214-450-5337 | before I get too deep.