External SCSI enclosure issue

[meall]

Hi,

I have an external enclosure that is having strange behavior... All tests have been done with a SCSI terminator on the enclosure.

If I connect the enclosure to a Mac without any HD installed internally, it boots up correctly to it, no problem. SCSI probe reports SCSI ID 3 as expected.

If I connect the internal HD, then start the enclosure then the Mac, it searches for HD all the time. Then, just shut off the enclosure the MAc boots from Internal HD no problem. SCSI ID of internal is 0.

Then I switch on the enclosure, the HD mount as second HD, and SCSI probe reports SCSI ID 0 and 3 as expected.

Any idea why this enclosure have such behavior?

Thanks

 

[Christopher]

Is it from Fantom?

 

[meall]

Is it from Fantom?

The enclosure is a "PowerUser Pro" sticker on it, other than that I cannot say!

 

[Christopher]

Oh. I was wondering because I had a Fantom Drives(not sure if thats right) that currently is sold in SCSI form but I had a IDE Firewire version that had problems like what you had.

If you can find a manual I would think there might be something in there. And possibly it might have jumper settings on the main board inside the case.

 

[trag]

Try your experiments with the SCSI terminator not on the enclosure. It is possible that the drive in the enclosure has a jumper set to provide termination, so with the added external terminator, you could be double-terminating the external end of your SCSI chain.

Alternatively, check the jumper settings of the drive in the enclosure.

Also, if these are Seagate drives, make sure that "Termination Power to Bus" is disabled. The host (the Macintosh) provides SCSI Termination Power, and providing it from additional devices can cause arguments amongst the voltage regulators involved leading to problems.

Even if they aren't Seagate drives, you should do the above, but it's only with Seagate drives that I've ever found it to be a real problem.

 

[register]

If the machine works well in combination with other external drives have a look at the jumper settings for the drive inside the external enclosure. Else you might check for proper termination of the internal HD. In rare cases the SCSI chain does not like the cable you use, also.

 

[Aoresteen]

Ditto on the termination issue. Check the external drive to make sure it is not internally terminated.

Also, pull off the terminator and add a second external SCSII device (make it SCSI ID 1 or 2). See what happens.

 

[meall]

Ok, did a few more tests!

I removed the external termination and it booted up to almost the end on the internal HD, but the computer freezes just at the boot end.

The HD is a Quantum 730 Mb (LT73S0011 or something like this). I did not had more time to test anything else...

So far, I did not see any jumpers on the hardrive itself... Maybe I did not looked correctly, bit normally they are obvious...

 

[equill]

... It is possible that the drive in the [external] enclosure has a jumper set to provide termination, so with the added external terminator, you could be double-terminating the external end of your SCSI chain ...

Double termination is also a potential pain if you at any time wish to add another SCSI device, such as a CD-ROM drive, to the chain on the remote side of the external HDD. An external terminator is easily removed for that purpose, but not so an internal terminator. The internal terminator could take the form of a jumper marked TE (termination enable) on a header block, especially on the Seagate drives that trag mentions. Alternatively, and depending on the age of the drive, termination could be passive, via two (or three) removeable in-line resistor packs on the drive's logic board.

SCSI is as much art as science, in which there are certain truths (don't put LVD devices on an HVD chain), frequent truths (all attached devices should be powered-on at startup), sometime truths (all SCSI cables are equal, but ultra-skinny cables are very unlikely to be so), and barely truths (Apple always implemented SCSI in a consistent manner).

You didn't say which model of Mac was concerned in this case, but the odds are that it uses the same bus (ie, with a single controller) for internal and external daisy-chains. If then the internal ribbon cable is much shorter than the external 'system' cable, the internal cable may act as a 'stub' rather than as a branch, thereby having the disconcering effect of a terminator within that daisy-chain.

de