Conference hydra::amiga_v1

    John,
    	Please, please, please upload to TAPE or wherever !!!
    
    				Alan T. (RZ23 w/o spinup !)

    
    Picked up a copy of RZutil from TAPE:: yesterday and tried it out on an
    RZ56 (665mb) that isn't connected to anything at the moment.
    
    I am using a GVP II with a homemade 25 > 50 pin cable.  The RZ56 is
    inside a storage expansion box at SCSI ID 1 along with a TK30 at SCSI
    ID 5.
    
    The RZutil program said it completed the Spin-up setting, I powered
    cycled everything (could not tell if the drive started or not).  I used
    the GVP Fastprep program to format drive, however when I selected ID 1
    the GVP controller hung (the led attached went solid and I could not
    access the other drive).
    Naturally I am somewhat curious as to why the controller would hang
    when I attempt to access this drive.  I did the same thing prior to
    using RZutil, so I presumed that the disk spin-up was a problem. 
    However after using RZutil to (I think) set the spin-up bit the same
    problem happens.  Perhaps this is a problem with my controller any
    suggestions on troubleshooting this problem.
    BTW: the Seagate that I also have was connected at the end of the
    homemade cable earlier without any problems so I think the cable is
    working.
    
    wayne
    

    I'm beginning to believe there's something odd about these RZ5x drives!
    I've had almost identical results trying to run these drives on my
    HardFrame.  I tried an RZ55 last night and an RZ57 tonight with pretty
    much the same non-success as reported in .2, grrrrr.

    By the way, the RZ57 has a jumper (W5) to enable/disable spinup, so you
    don't need to use RZUtil to fix that.  Neither setting appears to make
    any difference in my (in)ability to make the drive do anything beyond
    spin-up or spin-down.

>    <<< Note 4405.3 by BOMBE::MOORE "Amiga: Real computing on a PC budget" >>>
>                 -< What *IS* it with these %^@$#% things????? >-
> 
>     I'm beginning to believe there's something odd about these RZ5x drives!
>     I've had almost identical results trying to run these drives on my
>     HardFrame.  I tried an RZ55 last night and an RZ57 tonight with pretty
>     much the same non-success as reported in .2, grrrrr.
> 

Gee Bruce, now that your swearing at them, I guess I better say *something* :-)

There is nothing odd about RZ5x drives that would cause the HF or GVP
to hang. About the only thing I can think of is termination and/or terminator
power. Bad term/power can cause strange behaviour. Did you make sure that
only the drive at the end of the cable has terminators? Make sure there is
not a long stub of cable flying in the breeze (put a drive at the physical
end of the cable).

Are you powering these drives from the Amiga? The RZ5x drives take much
more power than the 3.5" ones. If you have a dvm, measure the 5 and 12 volts
to be sure. Also measure pin 26 (term power) on the SCSI cable to see that it's
at least 4.5 - 4.75 volts.

Also, the RZ5x drives support synchronous data transfer. Maybe HF or (and?!)
GVP have trouble with this? There have been known to be oddities with when 
a device negotiates for sync protocol. Some SCSI initiators will get confused
if the device asks when not "expected" to.



john

>    <<< Note 4405.4 by BOMBE::MOORE "Amiga: Real computing on a PC budget" >>>
>                        -< Startup not an issue on RZ57 >-
> 
>     By the way, the RZ57 has a jumper (W5) to enable/disable spinup, so you
>     don't need to use RZUtil to fix that.  Neither setting appears to make
>     any difference in my (in)ability to make the drive do anything beyond
>     spin-up or spin-down.
> 


Yup, as of RZ57 (I think maybe RZ56?), we tossed the modeselect stuff for
enabling/disabling power on spin. Now there is a jumper and the mode select
page is a read-only page that reflects the state of this jumper.

You are setting the SCSI ID properly, right? I mean you don't have 2 drives
with the same ID on the bus, right?  That will definetly cause a hang!

I believe both HF and GVP use ID 7 so the drives can be ID 0 - 6.

john

    Power is fine (I just measured), there doesn't even seem to be a strain
    during spin-up.  Terminator power on the cable reads 4.8 volts, I've
    got termination at both ends (and only at the ends).  I'm pretty sure
    the SCSI IDs are properly set, I can park/unpark each drive individually
    as units 0 and 1, no response at any other addresses.  I've tried a
    couple of (similar) cables.  I'm using an external (Centronics style)
    terminator on the far end of the cable, and internal termination on the
    controller end.
    
    I have discovered that once things lock up, I can recover by briefly
    unplugging the SCSI cable from the RZ5x drive...
    
    All of the Microbotics utilities that "scan for drives" hang at that
    point, but their simple PARK program (in SampleCode drawer) works fine.
    The SCSItest program hangs, reports that "INQUIRY" command failed after
    I pull the cable to unfreeze it.

>    The SCSItest program hangs, reports that "INQUIRY" command failed after
>    I pull the cable to unfreeze it.

Sounds like a protocol battle. Maybe the synchronous negotiation is messing
up someone. 

Without a scsi bus analyzer, I would call Microbotics and ask.

john

I've (very temporarily) got a RZ55 in the office which I was thinking of
connecting to my HardFrame, but it sound's a little too aggravating to me.

People have RZ2%'s working don't they?  I was seriously considering 
trying one of those.

Steve

    The RZ55 works fine.  I've used them on Supra and GVP controllers with
    no problem.  Its a really nice drive, hated to give it back.  In fact,
    I bought the 200meg Conner.  The RZ55 is the 300meg version.
    
    md

    
    A lot of further testing on my system found that the GVP FastPrep
    program will hang the controller when it queries the RZ56 drive when
    I have my Seagate 157 on the bus.  If I pull the Seagate the FastPrep
    works fine.  Even after prepping the RZ56, hooking up the Seagate and
    trying the FastPrep the controller still hangs.
    
    Thinking that there might be timing problems I copied about 25mb of
    files from the Seagate to the RZ56.  No problems occured so I don't
    know what might be going on, I plan to report this to GVP.
    
    BTW: does anybody happen to have a program which will allow you to
    build SCSI command packets (command and data) and send them to the
    specified driver and subsequently display the results?
    
    wayne
    

I am slowly working on a scsi.library that will make this easy.

don't hold your breath though...

john 

    After yet another night of experimentation, here's what I've found...

    I now have an RZ23 installed and running perfectly.  This was
    relatively straight-forward except that I couldn't get the drive to
    show up on Amiga-DOS because I'd forgotten to clear the "last drive"
    flag on the other drive.  Other than that, no real difficulty getting
    the RZ23 up and running.

    That completed, proving the cables, termination, etc. are in working
    order, I tried a number of alternate configurations with the RZ55 and
    RZ57 drives.  Based on comments a couple of replies back, I tried
    connecting each drive by itself, still no good.  I even went so far as
    to get RDPrep running with the RZ23 and then swapping the SCSI cable to
    one of the others (set to the same SCSI ID).  But further attempts to
    access the drive still just hung.

    I can't figure out what it is, but there's almost certainly some kind
    of incompatibility between the HardFrame and RZ5x drives.


    						8^(

Well Bruce,  if you want to pack up your amiga and bring it in to SHR
I can meet you here and we can figure out exactly what's going on using
the scsi bus analysers we have here...

john