SCSI supports numerous hardware devices chained together
One problem with this design is the echo effect. When an electrical signal reaches the end of the cable, the signal tends to produce an echo of itself that travels back up the cable in the wrong direction. This echo interferes with other signals that are traveling along the cable and causes chaos.
To avoid this problem, you must terminate SCSI devices. Termination refers to the practice of placing a terminator at the end of the chain. A terminator is nothing more than a resistor designed to absorb the signal, thus preventing an echo.
Every SCSI device is designed so that the device can be terminated. But you must only terminate the device at the end of the SCSI chain. If you terminate a device earlier in the chain, you'll isolate the devices beyond the point of termination, and they'll no longer be able to communicate with the SCSI controller. Terminating the chain prematurely can also in some cases damage SCSI devices.
One of my previous tips discussed that there was no standard method for setting a SCSI ID. A SCSI ID can be set through jumpers, dip switches, dials and other methods. It's the same for SCSI termination – there's no standard way of terminating a SCSI chain.
Termination is always performed on the last device in the chain, so the method used for termination is specific to that device. In older SCSI implementations, terminating a device simply meant physically installing a special resistor onto the port that would normally be used to attach the next device in the chain. But most of today's SCSI devices are self-terminating; you don't have to physically install a resistor onto a port.
But just because most modern SCSI devices are self-terminating, it doesn't mean that termination occurs automatically. It's still up to you to enable termination on the last device in the chain. Typically this involves setting a jumper or a dip switch, but some devices allow you to enable termination through a software interface.
Many varieties of SCSI have been introduced over the years. That being the case, it may come as no surprise that there are several different types of termination. The type of termination that is necessary depends on the particular flavor of SCSI being used.
Since most modern SCSI devices are self-terminating, you won't have to worry about using the correct type of termination unless you are working with an older device. Even so, it's good to be aware that various types of termination do exist. The primary types of SCSI termination are: Passive, Active and Forced Perfect.
Passive termination is the type described at the beginning of this article. It simply involves placing a resistor on the last port in the SCSI chain.
Higher speed implementations of SCSI require tighter tolerances for termination. Therefore, Active termination works not by absorbing the signal through a resistor, but by regulating voltage on the line. LVD is another type of active termination.
Forced Perfect termination also uses voltage regulation, but does so by using a diode (a special type of resistor that is polarized). This means that current can only flow through a diode in one direction. A diode can prevent the echo effect by blocking current from flowing backwards on the wire.
About the author: Brien M. Posey, MCSE, is a Microsoft Most Valuable Professional for his work with Windows 2000 Server, Exchange Server and IIS. He has served as CIO for a nationwide chain of hospitals and was once in charge of IT security for Fort Knox. He writes regularly for SearchWinComputing.com and other TechTarget sites.
More information on this topic:
- Tip: Guide to troubleshooting SCSI
- Topics: Computer buses
- RSS: Sign up for our RSS feed to receive expert advice every day.
This was first published in June 2007