Amiga.org
Amiga computer related discussion => Amiga Hardware Issues and discussion => Topic started by: Lockon_15 on April 30, 2006, 06:50:07 PM
-
I have A2000 with GVP ImpactII SCSI controller. IBM harddrive is set on ID1, while Plextor CDRW occupies ID2.
When I turn on machine, harddrive spins up, CDRW lights blinks, but screen display stays white. In order to proceed booting, I have to reset Amiga. On the other hand, system boot regulary if CDRW is disconnected.
I've read some posts on this forum regarding re-arranging ID on harddrive to ID6 (factory default), but does anyone has confirmed solution ?
Thnx
-
Make sure termination is correct. Check this (http://wonkity.com/~wblock/SCSI/SCSIExamples.html) page for more information.
/Patrik
-
Hi Patrik,
I'm pretty sure termination is correct. Harddrive is connected by middle SCSI 50-pin connector and CDRW is terminated by jumper since it is connected by the end of SCSI ribbon cable.
Anyway, thnx for idea :angry:
-
Ok, so the harddrive definately has no termination, not even any old resistor-packs?
Btw, the first harddrive usually is jumpered as id 0 - ie all id-jumpers removed. This shouldnt matter as it works when only the harddrive is attached, but for wierd problems, wierd things can be worth trying :D.
/Patrik
-
Yes, harddrive has no cap on termination jumper. I'm not sure if SCSI chain is properly terminated on SCSI board since I have no user manual for GVP ImpactII controller. However, SCSI PCB has 3 jumpers for SCSI device ID (default is ID0 or even ID7).
-
The ImpactII has onboard termination and should have id 7 per default. It was the harddrive and id 0 I was talking about.
/Patrik
-
OK. Can you recommend exact jumper settings for each SCSI chain part (controller, hdd, cd)?
Device position is fixed, since SCSI cable is too short to permit hdd/cd permutation.
I've read somewhere that there is exact SCSI detect order which can have impact on boot select. So far, it seems that HDD is too slow on spin-up and so misses pre-defined time frame for boot select.
-
Sounds very reasonable.
Unfortunately, there is no option to lengthen the wait of the ImpactII controller at poweron. Is there any jumper on the harddrive to enable/disable auto-spinup at poweron which you could try changing?
/Patrik
-
Yes, I have such jumper. It is currently disabled, so drive automatically spin-up as you power up the system. Enabling this jumper will mean that spin-up will commence only after specific directSCSI command is issued.
This would render HDD even more inert so I can't see it as a practical move towards succesfull boot-up. A friend of mine told me few hours ago that he experienced same problems running IDE HDD on A1200. He had 2 choices: to perform a hardware hack to delay boot interval, or to acquire SCSI disk for his Blizzard 1240. He went for disk and got rid of troubles.
Anyway thnx for trying.
-
Aye, I know, but maybe the ImpactII acts different when it has to tell the drive to spin up.
There is one more option, a long shot actually which most likely wont work. If you have access to an EPROM-burner, you could update the onboard ROM-driver (gvpscsi.device) to the latest version and hope that version waits longer. Check this (http://babel.org/amiga/) page.
/Patrik
-
Hi,
only with 'SCSI 1' hdd you can add an another peripheral on this card due to a bios card limitation !
It happened the same to me.
The CDRW could only be seen with the old Quantum 80 Mb SCSI1 hdd. When I replaced it by any SCSI2 hdd the CDRW could not be seen any more.
We need a bios upgrade, and there is some on Ebay that are selling these :-D .
-
Lockon_15 wrote:
A friend of mine told me few hours ago that he experienced same problems running IDE HDD on A1200. He had 2 choices: to perform a hardware hack to delay boot interval, or to acquire SCSI disk for his Blizzard 1240. He went for disk and got rid of troubles.
this happened to me back in the old days with a wdc 420mb drive. when i jump to a newer 850mb problems stoped. all drives i have test after 1996 spining fast enough and there is no problem of this kind in a1200 with them
-
This becomes more and more interesting. I have just checked ROM version by both visualID and issuing "version gvpscsi.device". It seems that I have v4.5 which is not last revision. Ralph Babel offers image for v4.15
-
CLS2086 wrote:
Hi,
only with 'SCSI 1' hdd you can add an another peripheral on this card due to a bios card limitation !
It happened the same to me.
The CDRW could only be seen with the old Quantum 80 Mb SCSI1 hdd. When I replaced it by any SCSI2 hdd the CDRW could not be seen any more.
We need a bios upgrade, and there is some on Ebay that are selling these :-D .
I will go for ROM burn. Now it seems that I had flaky ROM since SysInfo reported only 1MByte per second on my 1GB IBM DPES-31080.
Thnx
-
Hang on a second, doesn't v4.5 mean it's newer than v4.15? I've seen some coders using for example v4.15 as sub-revision fifteen as opposed to one-five...
I would have thought v4.5 is better?
I know Phase 5 updated the v8.1 to v8.5 ROM on the SCSI-IV kit because of issues with CD burning etc. BlizKick can do this in software if you have the v8.5 image I believe.
Maybe the same could apply for your controller?
-
Hyperspeed wrote:
Hang on a second, doesn't v4.5 mean it's newer than v4.15? I've seen some coders using for example v4.15 as sub-revision fifteen as opposed to one-five...
I would have thought v4.5 is better?
I know Phase 5 updated the v8.1 to v8.5 ROM on the SCSI-IV kit because of issues with CD burning etc. BlizKick can do this in software if you have the v8.5 image I believe.
Maybe the same could apply for your controller?
It got me for a second, too.
But then it would be logical to expect 4.50 notation...no, I don't think so. This is quote just right off Ralph Babel's Amiga pages...
"Note: Yes, 4.15 really is the latest one. Hint: Which section would you expect to come first in a book: 4.5 or 4.15?"
4.15 is gonna burn in hell, but meanwhile I am thinking if it is possible to remap newer ROM into FASTRAM and updating system with Binddrivers...just to see if there is transfer speed increase (it should be above 2.5Mbytes per second AFAIK)
-
What the hell!?
So sub-revisions of programs don't follow decimal rules?
So for example 30.34 would be newer than 30.33 but 1.5 is older/crapper than 1.49?
That can't be right...
-
It all depends on who put the software out. GVP followed one set of upgrade numbers, but no one else is forced to follow suite. Actually due to copyright laws, they may not be able to. Then you also have the phase of the moon and sun spots to consider.........
-
Amiga version numbers are composed of two integers which are written out in the form number1.number2, so 4.15 comes after 4.5, just like chapters in a book.
But seriously, there is no need to argue about this as Ralph Babel who wrote the driver writes on his page (http://babel.org/amiga/) that 4.15 is more recent than 4.5.
/Patrik
-
Revisions like chapters of a book? What is this, the Bible?
If they wanted to do that shouldn't it be v4:5 Mark Chapter 13...
:-D
When I use the Commodore Installer and it says "Installed Version" and "Version to be installed" I'd pick v4.5 over v4.15.
Even the Europeans would get confused because they just swap the stop with a comma (./,).
And when I check the datestamp the v4.5 usually is more recent than the v4.15 if you get my meaning.
-
by patrik:
But seriously, there is no need to argue about this as Ralph Babel who wrote the driver writes on his page that 4.15 is more recent than 4.5.
Yeah, but look at what motrucker said:
by motrucker:
It all depends on who put the software out. GVP followed one set of upgrade numbers, but no one else is forced to follow suite.
And, patrik... that link talks about nothing but GVP drivers (which for some reason give me that 'low quality' vibe).
:-D
-
Uhh, based on the behaviour of the 'version' command I would assume that all coders would follow the same standard Ralph is following.
If you write (and run) a script file thus:
version >NIL: gvpscsi.device 4.10
if NOT warn
print "Ralph is Right, v4.15 is later than v4.5"
elseif warn
print "Ralph is wrong, v4.5 is later than v4.15"
endif
You will find that Ralph is right (assuming I have remembered my syntax correctly).
I use this routine, with different parameters, in my startup-sequence to selectively call bppcfix only when I have the BPPC in the machine (the 68040 library in flashrom is a later version than the one in my LIBS: drawer). Haven't yet thought of a way to selectively install/remove the warpup libraries though, having them in a seperate drawer with a 'LIBS: ADD' assign doesn't seem to work.
-
So sub-revisions of programs don't follow decimal rules?
No.
It's quite simple really: version revision [ subrevision]
It just happens that the separator is a '.'
-
Piru: If you are one of the coders that subscribes to this madness consider me on your 'Persons likely to strangle me' list!
EDIT: And what happens when things get like V4.5.15?
Is that the fifteenth sub-sub-revision of the fifth sub-revision of version 4?
You've got to be kidding me! That's more arse backwards than manufacturers calling a Megabyte "1,000,000 bytes' (when it should be 1,048,576 bytes).
-
Looks perfectly logical to me. And anyway, software cos seem to like naming their programs after the years they were released in lately :-P.
-
@Hyperspeed
Well, this is what C= coding guidelines say. It's all documented.
EDIT: And what happens when things get like V4.5.15?
Is that the fifteenth sub-sub-revision of the fifth sub-revision of version 4?
Sure. Perfectly logical. Except that's it: the fifteenth sub-revision of the fifth revision of version 4.
There can be V4.5.1243 and V4.5.2 and there never is any question about which is more recent, as there is with some weird decimal number notation.
You don't need to like it personally, but this is what most of the civiliced world uses for version numbers.