Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

CyberStorm MK3 SCSI problem...?

70 views
Skip to first unread message

Jay

unread,
Sep 29, 1998, 3:00:00 AM9/29/98
to
I'm happy with my new Cyberstorm mk3/060... nice and speedy.

I just borrowed a 68 to 50 pin adapter to test/hook up my SCSI-2
drives (5400rpm Seagate Hawks), so I can try out the CyberSCSI and
eventually ditch my slow Oktagon controller.

Well, it looked fine, except now the drives are *slower* than when
they were on the Oktagon. I'm getting speeds like 200k/sec copying
large (5+meg) files to RAM: or to the other drive. I've tested these
drives at 5+ megs/sec on my PC (which also has a SymBios UltraSCSI
card).

Could this be a sign of bad termination? Or that the adapter itself
is not terminated? (I don't know if it is or not) The SCSI chain
looks like this:

mk3 --> 68 to 50 pin --> Hawk #1 --> Hawk #2 (Terminated)
(Terminated??)

--jay
ga...@ottawa.com

Doug Smith

unread,
Sep 29, 1998, 3:00:00 AM9/29/98
to
Jay wrote in comp.sys.amiga.hardware about "CyberStorm MK3 SCSI problem...?
":


> Well, it looked fine, except now the drives are *slower* than when
> they were on the Oktagon. I'm getting speeds like 200k/sec copying
> large (5+meg) files to RAM: or to the other drive.

IME, bad scsi setups tend to either not work at all, or lock after a
short time.

Check the mask value in HDToolbox.


>
> Could this be a sign of bad termination? Or that the adapter itself
> is not terminated? (I don't know if it is or not) The SCSI chain
> looks like this:
>
> mk3 --> 68 to 50 pin --> Hawk #1 --> Hawk #2 (Terminated)
> (Terminated??)
>

As I said above, I don't think this is the root cause of the problem,
however, this is not correct; you have only terminated one end of
your scsi chain.

The Mk3 has no termination on it, so you need to terminate it somehow.

What you really need is,

SCSI3-terminator--->mk3-->68 to 50 Upper8 term--->Hawk1--->Hawk2[term]

Though this will mean you buying a fast and wide cable, just to terminate
it.

You might want to have a try at plugging the 50 pin adaptor onto the
Mk3, then connecting that to one of the middle connecters on a 50 way
cable and putting one of the hawks at each end, it's not really correct,
It's going to leave the upper 8 lines incorrectly terminated, but it should
work with the bits you have, ie,


Mk3
[term]Hawk1 ---> 68->50 Adaptor---> Hawk2[term]

Cheers,

doug.
--
Email: Do...@defocus.demon.co.uk http://www.defocus.demon.co.uk
IRC : arper (IRCnet - irc.demon.co.uk:6667)
IRL : Hazel Grove, Stockport, NW England.
"reliable PC" is an oxymoron. Jolyon Ralph, 13-07-98


joe

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
** To reply in e-mail, remove "bejvuk." from address **

On Tue, 29 Sep 1998 20:20:20 GMT, Jay wrote about CyberStorm MK3 SCSI problem...? :


> I'm happy with my new Cyberstorm mk3/060... nice and speedy.
>
> I just borrowed a 68 to 50 pin adapter to test/hook up my SCSI-2
> drives (5400rpm Seagate Hawks), so I can try out the CyberSCSI and
> eventually ditch my slow Oktagon controller.
>

> Well, it looked fine, except now the drives are *slower* than when
> they were on the Oktagon. I'm getting speeds like 200k/sec copying

> large (5+meg) files to RAM: or to the other drive. I've tested these
> drives at 5+ megs/sec on my PC (which also has a SymBios UltraSCSI
> card).
>

> Could this be a sign of bad termination? Or that the adapter itself
> is not terminated? (I don't know if it is or not) The SCSI chain
> looks like this:
>
> mk3 --> 68 to 50 pin --> Hawk #1 --> Hawk #2 (Terminated)
> (Terminated??)


On my Mk3 setup I had to terminate both ends. However, I'm using SCSI3
ultra wide in addition to 50 pin devices. If it helps, my chain looks
like this with a 68 pin ribbon cable:

Viking2 HD -------Mk3-------Pioneer CDROM----------68 pin terminator
(terminated) via 68/50 adapter
only terminates
upper 18 pins

With my setup and the HD set to SYNC I can get 14MB/s continuous transfer
rates. It may be that you also need to terminate both ends of your SCSI
cable. May I suggest the following on your 50 pin cable:

Hawk #1---------------Mk3 via-----------Hawk #2
(terminated) 68/50 adapter (terminated)
only unused upper
lines terminated

Also, if your Mk3 has NEVER been updated, immediately go to the Phase5
ftp site and download the latest 060 lib and FlashROM updates in the
PPC directory (yes PPC). I believe the files are named and dated 11
July, 1998 for both the libs and FlashROMs. This will save you a lot
of headaches since the new software fixes a lot of bugs. Also, if you
ever plan to set up a new HD, look for new versions of UnitControl and
SCSIConfig. The stuff on the original System and SCSI diskettes has
bugs. Good luck.....................joe (I've been there)

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On Tue, 29 Sep 1998 23:34:47 -0000, "Doug Smith"
<dou...@defocus.demon.co.uk> wrote:
>> Well, it looked fine, except now the drives are *slower* than when
>> they were on the Oktagon. I'm getting speeds like 200k/sec copying
>> large (5+meg) files to RAM: or to the other drive.
>
>Check the mask value in HDToolbox.

I'm using the defaults... 0x7ff fffc mask, and 0x00ff ffff
maxtransfer

> Mk3
>[term]Hawk1 ---> 68->50 Adaptor---> Hawk2[term]

I just set it up exactly this way, and it makes no difference
speed-wise. Combined throughput maxes at 3 mb/sec for both drives
combined, according to SCSIBench... and copying files is 200-300k/sec
max.

odd... :/


--jay
ga...@ottawa.com

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On 30 Sep 1998 00:39:16 GMT, e...@bejvuk.epix.net (joe) wrote:
>On Tue, 29 Sep 1998 20:20:20 GMT, Jay wrote about CyberStorm MK3 SCSI problem...? :
>With my setup and the HD set to SYNC I can get 14MB/s continuous transfer
>rates.

<envious stare>

>It may be that you also need to terminate both ends of your SCSI
>cable. May I suggest the following on your 50 pin cable:
>
>Hawk #1---------------Mk3 via-----------Hawk #2
>(terminated) 68/50 adapter (terminated)
> only unused upper
> lines terminated

Doug mentioned this, I have it configured this way now, but it
doesn't seem to have any performance difference. Transfer is still
slow. 3MB/s combined throughput in SCSIBench. And much less than
that simply copying large files to RAM: and back...

However, the adapter I'm testing likely doesn't terminate the upper
lines... could that make such a huge world of difference?!



>Also, if your Mk3 has NEVER been updated, immediately go to

This is the first thing I did *before* I even installed it.. :-)

thanks

--jay
ga...@ottawa.com

Antony Alonso

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
Hello Jay, on 29-Sep-98 21:44:59, you said,

> I just set it up exactly this way, and it makes no difference
>speed-wise. Combined throughput maxes at 3 mb/sec for both drives
>combined, according to SCSIBench... and copying files is 200-300k/sec
>max.

What system is this installed in? If it's *NOT* an A4k, you'll need to
do the INT2 hack (required for A3000/A3000Ts).

>--jay
>ga...@ottawa.com


--

*ATTENTION*
==========

Do NOT auto-reply to this message. The return address has been purposefully
altered so as to prevent unsolicited spam mail. To properly respond via
email, remove the "1" from "NightShade1" in the reply-to address.

System Specs (for comp.sys.amiga.* newsgroup posts):
---------------------------------------------------
A3kT, CyberVision64, MultiFaceIII, Emplant Deluxe, BJ-610
Warp Engine 4040 w/128mb Fast, 2.1GB HD, 1GB Jaz, 4X CDROM
Paragon 600SPII Flatbed scanner, USR V.90 Courier, 17" MAG


joe

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
** To reply in e-mail, remove "bewjed." from address **

On Wed, 30 Sep 1998 02:49:40 GMT, Jay wrote about Re: CyberStorm MK3 SCSI problem...?:


> On 30 Sep 1998 00:39:16 GMT, e...@bejvuk.epix.net (joe) wrote:
> >On Tue, 29 Sep 1998 20:20:20 GMT, Jay wrote about CyberStorm MK3 SCSI problem...? :
> >With my setup and the HD set to SYNC I can get 14MB/s continuous transfer
> >rates.
>
> <envious stare>

You wouldn't be so envious if you knew the hassle I went through to get
the damn thing working :( However, by immediately downloading the latest
FlashROM and 060 libs you saved yourself a lot of trouble.

> >It may be that you also need to terminate both ends of your SCSI
> >cable. May I suggest the following on your 50 pin cable:
> >
> >Hawk #1---------------Mk3 via-----------Hawk #2
> >(terminated) 68/50 adapter (terminated)
> > only unused upper
> > lines terminated
>
> Doug mentioned this, I have it configured this way now, but it
> doesn't seem to have any performance difference. Transfer is still
> slow. 3MB/s combined throughput in SCSIBench. And much less than
> that simply copying large files to RAM: and back...

I'm not familiar with those drives but they definitly kick the crap out
of my old Samsung 2.5GB EIDE drive. That did a max of 2.5MB/s on my
A4000 IDE port. Also, make sure you have Cybermap enabled from the Mk3
early startup menu. And if you have 60ns RAM on your Mk3, that can also
be enabled from the Mk3 early startup. You probably already know this but
the menu is accessed by holding the Escape key during a reboot or cold
boot. Those 2 options will make your Mk3 even FASTER :) Ummm, the mouse
will disappear afterwards requiring another reboot. No big deal.

> However, the adapter I'm testing likely doesn't terminate the upper
> lines... could that make such a huge world of difference?!

At this point I'd like to suggest contacting Greg Scott from National
Amiga. He helped solve my SCSI3 problem and sold me that 68 pin internal
terminator for my Mk3. It solved my problems and allowed me to run at
full speed (14MB/s) :) The only downside is that when I use external
cabling/devices I have to drop to ASYNC mode which limits my HD transfers
to 7.5MB/s. Otherwise I start getting intermittent HD errors every now
and then. Scary stuff!

But I've read on these ngs many times that it is suggested to use the Mk3
and CSPPC SCSI3 interface for internal devices only to allow maximum performance.
From my experience that appears to be correct. And once you get everything
squared away the performance is top notch. There are some SCSI3 owners
that hang out in these ngs with the 10000rpm Cheetah. That drive easily
handles over 16MB/s continuous (or better). All I'm suggesting is that
when your cabling/termination problems are worked out, your Hawk SCSI
drives will perform at their peak performance with your Mk3 SCSI interface.
Best of luck.................joe

mjfu...@flasuncoast.net

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
In article <361290b5.171082183@news>,

ga...@killspam.ottawa.com (Jay) wrote:
> I'm happy with my new Cyberstorm mk3/060... nice and speedy.
>
> I just borrowed a 68 to 50 pin adapter to test/hook up my SCSI-2
> drives (5400rpm Seagate Hawks), so I can try out the CyberSCSI and
> eventually ditch my slow Oktagon controller.
>
> Well, it looked fine, except now the drives are *slower* than when
> they were on the Oktagon. I'm getting speeds like 200k/sec copying
> large (5+meg) files to RAM: or to the other drive. I've tested these
> drives at 5+ megs/sec on my PC (which also has a SymBios UltraSCSI
> card).
>
> Could this be a sign of bad termination? Or that the adapter itself
> is not terminated? (I don't know if it is or not) The SCSI chain
> looks like this:
>
> mk3 --> 68 to 50 pin --> Hawk #1 --> Hawk #2 (Terminated)
> (Terminated??)

I have a MK3 and was tempted to connect my harddrive to it but read
that only Ultra Wide SCSI devices benefit. Other devices may actually
run SLOWER after you went to the expense and trouble to get the 80 pin
cable, adapters, and terminators. IIRC, it was my dealer who pointed
this out when I e-mailed him about connecting my new Quantum to the
MK3.

My dealer could be wrong about this, not that he's stupid, but because
he gets little help from phase5.

--
I'm Mike, elsewhere. Team AMIGA, of course.

-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On 30 Sep 98 00:20:29 -0500, "Antony Alonso"

<Night...@Clarksburg.com> wrote:
>> I just set it up exactly this way, and it makes no difference
>>speed-wise. Combined throughput maxes at 3 mb/sec for both drives
>
>What system is this installed in? If it's *NOT* an A4k, you'll need to
>do the INT2 hack (required for A3000/A3000Ts).

It's an a4000. Buster 9. Though, I'm told the Buster revision
makes no difference.

The only two cards in the machine are a Zorro2 IO board, and a
Zorro3 GFX card, which is working at full speed according to
benchmarks.


--jay
ga...@ottawa.com

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On 30 Sep 1998 05:25:37 GMT, e...@bewjed.epix.net (joe) wrote:
>On Wed, 30 Sep 1998 02:49:40 GMT, Jay wrote about Re: CyberStorm MK3 SCSI problem...?:
>> <envious stare>
>You wouldn't be so envious if you knew the hassle I went through to get
>the damn thing working :( However, by immediately downloading the latest
>FlashROM and 060 libs you saved yourself a lot of trouble.

It works amazing now, is ALL that matters. :>

> And if you have 60ns RAM on your Mk3, that can also
>be enabled from the Mk3 early startup.

> Those 2 options will make your Mk3 even FASTER :)

I'm amazed how much a difference it made, too. Lots of stuff is
easily 2x faster

>At this point I'd like to suggest contacting Greg Scott from National
>Amiga. He helped solve my SCSI3 problem and sold me that 68 pin internal
>terminator for my Mk3.

Yeah... it *must* be the connector. <shrug> That's all I can think
of it being. I'll be sure to order one.

>when your cabling/termination problems are worked out, your Hawk SCSI
>drives will perform at their peak performance with your Mk3 SCSI interface.

Oh, i hope so...

>Best of luck.................joe

Thanks!

--jay
ga...@ottawa.com

Niels Knoop

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
Jay (ga...@killspam.ottawa.com) wrote:

> I'm using the defaults... 0x7ff fffc mask, and 0x00ff ffff maxtransfer

^^^^^^^^
If this is not the typo, it is the reason. It needs to be 0x7ffffffc
(eight digits). Otherwise you disallow DMA transfer to your Cyberstorm's
Fast-RAM.

--
| Bye! /// Niels Knoop /// e-mail:
ni...@rbg.informatik.tu-darmstadt.de |
--

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On 30 Sep 1998 13:06:21 GMT, ni...@rbg.informatik.tu-darmstadt.de

(Niels Knoop) wrote:
>> I'm using the defaults... 0x7ff fffc mask, and 0x00ff ffff maxtransfer
> ^^^^^^^^
>If this is not the typo, it is the reason. It needs to be 0x7ffffffc
>(eight digits). Otherwise you disallow DMA transfer to your Cyberstorm's
>Fast-RAM.

Oops, it was a typo.. yes, it's 8 digits. So I guess that's not the
problem.


--jay
ga...@ottawa.com

Jay

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On Wed, 30 Sep 1998 11:36:18 GMT, mjfu...@flasuncoast.net wrote:
>that only Ultra Wide SCSI devices benefit. Other devices may actually
>run SLOWER after you went to the expense and trouble to get the 80 pin
>cable, adapters, and terminators. IIRC, it was my dealer who pointed
>this out when I e-mailed him about connecting my new Quantum to the
>MK3.

Oh no... I hope that's not the case... <sigh> I'll shell out the $25
to see if it makes a difference, anyway. If not, my dealer said I can
return it, luckily.


--jay
ga...@ottawa.com

Ralph Schmidt

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to

Using 0x7ffffffc or 0xffffffff can´t hurt either.
--
Ralph Schmidt,la...@popmail.owl.de(private),NextMail welcome

Flemming Steffensen

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
Ralph Schmidt wrote in comp.sys.amiga.hardware about "Re: CyberStorm MK3 SCSI problem...?":

[mask, and maxtransfer]

RS> Using 0x7ffffffc or 0xffffffff can´t hurt either.

Ralph, you know much about the mask and maxtransfer settings... Could
you give some (easy to understand for everybody) information on how they
work, so that all the confusion that always seem to exist about these
settings can be cleared?

I hope you can.

Thanks in advance!

--
Flemming Steffensen
.: | FIRST Production -ETD 020-
.::: | Homepage: http://www.bigfoot.com/~fsteff/ CATS'
.;' :: | INTERNET: fst...@bigfoot.com |\ _,,,---,,_sleeping!
.;' :: | AMIGANET: 39:140/146.17 ZZZzz /,`.-'`' -. ;-;;,_
.::::::::: | FIDO-NET: 2:238/188.17 |,4- ) )-,_; ,\ ( `'-'
.::. .::. | '---''(_/--' `-'\_)
Using IS08859-1? Do you see © as copyright, ÷ as division and ½ as 1/2?

Ralph Schmidt

unread,
Sep 30, 1998, 3:00:00 AM9/30/98
to
On 09/30/98, "Flemming Steffensen" wrote:
>Ralph Schmidt wrote in comp.sys.amiga.hardware about =22Re:
CyberStorm MK3 =
>SCSI problem...?=22:
>
>=5Bmask, and maxtransfer=5D
>
>RS=3E Using 0x7ffffffc or 0xffffffff can=B4t hurt either.

>
>Ralph, you know much about the mask and maxtransfer settings...
Could
>you give some (easy to understand for everybody) information on how
they
>work, so that=A0all the confusion that always seem to exist about
these
>settings can be cleared?
>

The Mask algorythm basicly looks like this...

if ((Address & Mask)==Address)
direkt transfer
else
buffer transfer

This means...

if your mask is 0xfffffffc you only allow longword aligned addresses
for direkt transfers.
if your mask is 0x00ffffff you only allow addresses in the <24bit area
for direkt transfers(Zorro2 addressspace)
if your mask is 0xffffffff or 0x7fffffff(doesn´t matter..as there
can´t be legal *system* ram at 0x80000000) the filesystem doesn´t
buffer.

The max transfer means what it says...0x00ffffff=24bit is the highest
transfer length one device job is used for...the FFS would split the
transfers then.

--
Ralph Schmidt,la...@popmail.owl.de(private),NextMail welcome

Flemming Steffensen

unread,
Oct 1, 1998, 3:00:00 AM10/1/98
to
Ralph Schmidt wrote in comp.sys.amiga.hardware about "Re: CyberStorm MK3 SCSI problem...?":

RS> The Mask algorythm basicly looks like this...
RS>
RS> if ((Address & Mask)==Address)
RS> direkt transfer
RS> else
RS> buffer transfer

Ok.. thanks, it clears up a few things.

RS> if your mask is 0xfffffffc you only allow longword aligned addresses
RS> for direkt transfers.

Hmmm... So 0xff..ffc forces longword aligned addresses... Does that
mean that 0xff..ffe forces word alligned addresses, and 0xff..fff no
allignment at all?


If yes, then which is to prefer, and for what reasons?

mjfu...@flasuncoast.net

unread,
Oct 1, 1998, 3:00:00 AM10/1/98
to
In article <6urujk$d06$1...@news1.epix.net>,
e...@bejvuk.epix.net (joe) wrote:
> ** To reply in e-mail, remove "bejvuk." from address **

>
> On Tue, 29 Sep 1998 20:20:20 GMT, Jay wrote about CyberStorm MK3 SCSI problem...? :
> > I'm happy with my new Cyberstorm mk3/060... nice and speedy.
> >
> > I just borrowed a 68 to 50 pin adapter to test/hook up my SCSI-2
> > drives (5400rpm Seagate Hawks), so I can try out the CyberSCSI and
> > eventually ditch my slow Oktagon controller.
> >
> > Well, it looked fine, except now the drives are *slower* than when
> > they were on the Oktagon. I'm getting speeds like 200k/sec copying
> > large (5+meg) files to RAM: or to the other drive. I've tested these
> > drives at 5+ megs/sec on my PC (which also has a SymBios UltraSCSI
> > card).
> >
> > Could this be a sign of bad termination? Or that the adapter itself
> > is not terminated? (I don't know if it is or not) The SCSI chain
> > looks like this:
> >
> > mk3 --> 68 to 50 pin --> Hawk #1 --> Hawk #2 (Terminated)
> > (Terminated??)
>
> On my Mk3 setup I had to terminate both ends. However, I'm using SCSI3
> ultra wide in addition to 50 pin devices. If it helps, my chain looks
> like this with a 68 pin ribbon cable:
>
> Viking2 HD -------Mk3-------Pioneer CDROM----------68 pin terminator
> (terminated) via 68/50 adapter
> only terminates
> upper 18 pins
>
> With my setup and the HD set to SYNC I can get 14MB/s continuous transfer
> rates. It may be that you also need to terminate both ends of your SCSI

> cable. May I suggest the following on your 50 pin cable:
>
> Hawk #1---------------Mk3 via-----------Hawk #2
> (terminated) 68/50 adapter (terminated)
> only unused upper
> lines terminated
>
> Also, if your Mk3 has NEVER been updated, immediately go to the Phase5
> ftp site and download the latest 060 lib and FlashROM updates in the
> PPC directory (yes PPC). I believe the files are named and dated 11
> July, 1998 for both the libs and FlashROMs. This will save you a lot
> of headaches since the new software fixes a lot of bugs. Also, if you
> ever plan to set up a new HD, look for new versions of UnitControl and
> SCSIConfig. The stuff on the original System and SCSI diskettes has
> bugs. Good luck.....................joe (I've been there)
>

Just thought I'd say thanks for the tip! This is the first post I've seen
which mentions the FlashROM updates in the PPC directory. My Mk3 does not
have the PPC option so I would not have even snooped around in that PPC
directory without your recommendation.

The most apparent improvement came when I loaded FinalCalc and it simply
dazzled me with the speed of loading. It was like having a new accelerator!

Thanks again.


--
I'm Mike, elsewhere. Team AMIGA, of course.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own

Ralph Schmidt

unread,
Oct 1, 1998, 3:00:00 AM10/1/98
to
On 10/01/98, "Flemming Steffensen" wrote:
>Ralph Schmidt wrote in comp.sys.amiga.hardware about =22Re:
CyberStorm MK3=
>=A0SCSI problem...?=22:
>

Sigh..you should check your newsclient config...some ISO char
thing is wrong.

>RS=3E The Mask algorythm basicly looks like this...
>RS=3E
>RS=3E if ((Address =26 Mask)=3D=3DAddress)
>RS=3E direkt transfer
>RS=3E else
>RS=3E buffer transfer


>
>Ok.. thanks, it clears up a few things.
>

>RS=3E if your mask is 0xfffffffc you only allow longword aligned
addresses
>RS=3E for direkt transfers.


>
>Hmmm... So 0xff..ffc forces longword aligned addresses... Does that
>mean that 0xff..ffe forces word alligned addresses, and 0xff..fff no
>allignment at all?
>

yes

>
>If yes, then which is to prefer, and for what reasons?
>
>
>

Depends on the device..most devices handle 0xffffffff.
Maxtransfer should always be 0x00ffffff for SCSI...a lot
scsi dma counters have a 24bit limit and designing around
may need some not trivial gimmicks in the device.
For IDE the max transfer rule is a bit different..there
it may also depend on the HD you use.

--
Ralph Schmidt,la...@popmail.owl.de(private),NextMail welcome

Flemming Steffensen

unread,
Oct 1, 1998, 3:00:00 AM10/1/98
to
Ralph Schmidt wrote in comp.sys.amiga.hardware about "Re: Mask and Maxtransfer Was:Re: CyberStorm MK3 SCSI problem...?":

RS> > CyberStorm MK3= =A0SCSI problem...?=22:
RS> Sigh..you should check your newsclient config...some ISO char thing
RS> is wrong.

Strange... It seems like your reader don't support Quoted-Printable
encoded messages. You are the first to suffer from this.

RS> > Hmmm... So 0xff..ffc forces longword aligned addresses... Does
RS> > that mean that 0xff..ffe forces word alligned addresses, and
RS> > 0xff..fff no allignment at all?
RS> yes

Ok.


RS> > If yes, then which is to prefer, and for what reasons?
RS>
RS> Depends on the device..most devices handle 0xffffffff. Maxtransfer
RS> should always be 0x00ffffff for SCSI...a lot scsi dma counters have a
RS> 24bit limit and designing around may need some not trivial gimmicks
RS> in the device.

Ok, I see.

RS> For IDE the max transfer rule is a bit different..there it may also
RS> depend on the HD you use.

Thats the 64k (MS-DOS enforced) limit going into action, are there other
limits to think of?

JAN-HENDRIK MANGOLD

unread,
Oct 2, 1998, 3:00:00 AM10/2/98
to
joe wrote:

> be enabled from the Mk3 early startup. You probably already know this but
> the menu is accessed by holding the Escape key during a reboot or cold
> boot. Those 2 options will make your Mk3 even FASTER :) Ummm, the mouse
> will disappear afterwards requiring another reboot. No big deal.

OK, what if this menu does not appear when holding the ESC key? That is my situation.
A3000 (int2 upgrade) MKIII. I also wandered how to do the update of the Flash when the
menu does not show... I posted this question earlier but haven't been online lately so I
missed any possible replies.
I too want to access the MKIII SCSI ;>/

Thanks in advance...

--
__ _______ ____ ____ ____ __
/ // / __/ \/ / \/ __ \/ / // /\
/ _ / __/ / / / / / _ / /
/_//_/___/_/\_/___/__/\_/_/_/\\_\/
\_\\_\___\_\/\\___\__\/\\_\_\/ \_\


joe

unread,
Oct 2, 1998, 3:00:00 AM10/2/98
to
** To reply in e-mail, remove "sasqix." from address **

On Fri, 02 Oct 1998 11:15:47 -0700, JAN-HENDRIK MANGOLD wrote about Re: CyberStorm MK3 SCSI problem...?:


> joe wrote:
>
> > be enabled from the Mk3 early startup. You probably already know this but
> > the menu is accessed by holding the Escape key during a reboot or cold
> > boot. Those 2 options will make your Mk3 even FASTER :) Ummm, the mouse
> > will disappear afterwards requiring another reboot. No big deal.
>
> OK, what if this menu does not appear when holding the ESC key? That is my situation.
> A3000 (int2 upgrade) MKIII. I also wandered how to do the update of the Flash when the
> menu does not show... I posted this question earlier but haven't been online lately so I
> missed any possible replies.
> I too want to access the MKIII SCSI ;>/


Since you already have the int2 upgrade it will be easy.

The menu doesn't appear on the earlier FlashROMs. I had the same problem.
Goto the Phase5 ftp site at:
ftp://ftp.phase5.de/

Find the PPC directory and look for FlashUpdates-110798.lha and a very
similarly named 060libs update. It also has -110798 in the file name.
Download both archives. Install the 060 libs 1st since that is easy.

The FlashUpdates archive contains updates for ALL the latest boards.
Mk3Update is the one you need. Be sure to read all the notes and follow
the instructions exactly. Do NOT interupt the update process. It takes
less than a minute. The update does not go exactly as described in the
docs. However, it works!

Once you have completed all the updates, reboot your computer and hold
down the Escape key. A menu will appear with options for SCSI, RAM,
SYSTEM, and CDFS. Best of luck.............joe

JAN-HENDRIK MANGOLD

unread,
Oct 2, 1998, 3:00:00 AM10/2/98
to
joe wrote:

> Once you have completed all the updates, reboot your computer and hold
> down the Escape key. A menu will appear with options for SCSI, RAM,
> SYSTEM, and CDFS. Best of luck.............joe

Doah!! Did it, works. Thanks :>)

0 new messages