Welcome, Guest. Please login or register.

Author Topic: Strange SCSI behaviour - maybe missing terminator?!?  (Read 10566 times)

Description:

0 Members and 1 Guest are viewing this topic.

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« on: September 16, 2012, 06:36:50 PM »
Quote from: danbeaver;708238
Put an active, wide Terminator on both ends of the cable. The one I picked out with the glowing LED has "Termination Power" and thus is "Active" or there would be no power to make the LED glow. No reason to invoke, "Night and Fog" in this (Nacht und Nebel).

Active termination has nothing to do with term power. Active termination means it is using a chip internally to keep the bus signal levels closer to exact. unlike passive termination which is just resistors that pull high,low etc.The led versions are handy to be sure you have power.
I have found some LVD/SE adapters will not work on the csppc, and some seem ok. Ultra wide SE adapters always work in my experience. Plain scsi 3 terminators may or may not work. i recommend sticking with ones labeled for SE ultrawide use.

Term power is usually supplied by the controller or a device on the chain. a terminator cannot supply term power. term power can be supplied anywhere on the chain.

I have to agree with other people's post. using narrow devices on the scsi chain will slow it. Its best to use all wide devices.
When using narrow devices on the end of the UW scsi chain,a "high byte" terminator(or adapter) should be used as to keep all lines properly terminated.

Plextor makes a line of scsi cdrom drives that are 68Pin ultra interfaced,these can be found used still. many were used in IBM servers.

I have used Acard 7720UW adapters with quality cf adapters and sandisk fast cf cards and managed to attain 32MB/s transfer rates on the csppc scsi. it really does scream as far as things in the amiga world go ;)

mech
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #1 on: October 06, 2012, 07:05:29 PM »
Quote from: JohnFante;710499
WOOOHOOOO!

Things are running again. :-D

When I removed the CD-ROM (Plextor PX-32TSi) so I only had the two HD's (Seagate and IBM) and a terminator at the end the HD's spinned up and I was able to boot my OS4.1 install.

And as far as I can feel the HD's works a little bit faster with the Quantum drive of the chain.

Now I only need to get the correct jumper settings for my CD-ROM drive to solve the conflict I have.

Anybody have any experience with how to correctly configure a Plextor PX-32TSi CD-ROM drive or a Plextor-W124TSi CD-RW drive?

Thank you in advance.


Just set the jumpers so it is a device between 0-6 and does not conflict with another drives on the chain(remember each scsi drive has to be a diff. number). then set termination on or off depending where it is on the chain. generally i set the hard drive numbers lower and the cd after those. the jumper settings are usually on the back of the cover or label.
putting 50 pin narrow scsi drives on the ultrawide 68 pin chain will generally slow everything down.  Plextor did make 68Pin cdroms also.

mech
« Last Edit: October 06, 2012, 07:23:10 PM by mechy »
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #2 on: October 07, 2012, 01:08:19 AM »
Quote from: danbeaver;710587
Again, nothing in this post will solve your issue, but since the last "unit" on the bus/chain is designated by its ID number and the controller wants to know the last device and last unit number, I have always felt that it should be a drive where this is recorded in its RDB; so I use a HDD there and a CD earlier.

Not sure what you mean here,but setting last unit i.e. in hdtoolbox is just to keep it from wasting time scanning past that ID - you can also set last LUN iirc. that value is written to the RDB.
Since scanning starts at ID 0 it makes sense to have the boot drive low as possible for quicker booting.
A narrow scsi controller actually does not care what the last number is and doesn't need to know..  if no last id is set,on narrow scsi it will scan 0-6 (controller usually 7) and on UWscsi supports 15 devices.

Now in scsi,termination wise the last drive doesn't care what number it is,its just the last drive physically on the last connector.(i know this has nothing to do with what you said above,just a FYI here).

Putting a cd early also slows things since cd's come on the bus slower than the average HD(not always-but in many cases..) In the reality of things this may not make enough difference to change it.

Some like to put the hd at ID 0 and the cdrom at ID5,i never really understood this,makes more sense to me to do ID0 for the hd and say ID1 for the cdrom if its the only other device.

Setting LASTID sometimes can be a pain,if you add a device past the last ID,and forget to edit the rbd to scan to the lastID,then you set there wondering why it doesnt see the drive you added.. lol.

Mech
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #3 on: October 07, 2012, 02:51:24 AM »
Quote from: danbeaver;710610
It appears to me before booting the SCSI controller scans the entire bus querying each ID (and LUN) before it does any booting. Also whenever I change a device (CF Card added or removed at end of chain -- something I no longer do) when I go to HDToolBox or Media Toolbox I get a "need to update drive" request dealing with Last ID / LUN. From this I inferred that the controller wants to be able to write this info into the RDB or someplace on the drive; hence a CDROM would have no place to write this


Dan,
It will write the last id to the boot drives RDB to what the last drive now is in this case.It would not try to write it to a cdrom of course since its a read only removable as far as i know.

mech
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #4 on: October 07, 2012, 03:41:57 AM »
Quote from: danbeaver;710622
Yes, that was my point; it can't write last anything to a CDROM so it just keeps checking all the IDs and LUNs.

Any help with the rest of the post on how active termination circuits work without power?

i guess i misunderstood what you were saying about the cdrom.

on the terminators, Someone else answered that for you earlier(Zac iirc), but the answer is they dont,you always need term power or terminators do not work :)

most active terminators use a custom chip to keep bus levels right,it of course also needs power.


Mike
« Last Edit: October 07, 2012, 03:46:50 AM by mechy »
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #5 on: October 07, 2012, 04:22:03 PM »
Quote from: danbeaver;710634
Herein lies my problem, if the active termination circuits don't have anything to do with the "term[ination] power" then what or where comes the power used to keep them, well, "active?"

i think you misunderstood what i meant :)  i was saying they are 2 seperate things. some people comfuse term power as termination etc.
Some people don't realize what term power was for.

mech
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #6 on: October 07, 2012, 04:27:57 PM »
Quote from: JohnFante;710656
Well I moved the HD's so that the chain is as follows:

Active terminator
CSPPC
IBM ID0 - no term power
Seagate ID1 - term power
CD-ROM ID2 - no term power
Active terminator

When I have that chain I still go to the kickstart 3.1 boot menu.

When I remove the CD-ROM it boots into OS4.1 fine (placed on the IBM).

I most admit I am loosing hope here. Once during my fideling around I got the CD to work. Then I put all hardware in its places and changed som ID jumpers and then it didn't work again. Do not know why? :-(

I have an old IDE DVD-drive laying around. Can I use that on the IDE channel? I know it is not that fast but since I am will not be using it for anything but installation that would be just fine.

SCSI kind a suck right now ....

Is the cdrom a 68pin cdrom or are you trying to adopt a 50pin cdrom to a 68pin UWscsi chain?

Have you went into the csppc early menu and checked the scsi setting? perhaps something is off? You might set everything to auto until you get up and running.

Maybe the cdrom drive is bad? 15-20 year old scsi drives do fail sometimes.

You could use a cd/dvd drive on the internal ide with no problems. it will be slow,but as you say, something is better than nothing. Be sure to set it as master if its the only ide device.

Mech
 

Offline mechy

Re: Strange SCSI behaviour - maybe missing terminator?!?
« Reply #7 on: October 07, 2012, 09:44:43 PM »
Quote from: JohnFante;710678
It is a 50pin CD-ROM with a 68pin to 50pin converter. It is a Plextor PX.32TS. Se photo. Just to be clear. It has worked fine but has stopped working because I removed a HD, attached a terminator etc.



I have set everything to auto. No result



I have thought about that but since it has worked flawlessly until now combined with the fact that It worked very shortly during my "fideling" and the fact that I get the same error when I use a Plextor PX-W124TS CD-ROW dirve I would say that it is unlikely.



Great :-) I will dig one out and see what happends.


I think i see the trouble, you should be using a 68 to 50 adapter with high byte termination. The picture you have seems to show a regular 68/50 adapter.

Mech