Terminator Testing Primer

Terminator Testing Primer

Q. What function does an SCSI terminator serve and how does it do its job?

A. SCSI Terminators are usually put at the end of one or more SCSI cables to allow fast signals to be passed through the cable without significant distortion. Signal distortion happens when a data pulse (traveling at nearly the speed of light) hits the end of a cable and reflects back. The longer the cable, the more of a problem this reflection causes because it bounces back in the middle of another data pulse, thus mixing the signals. This could be likened to having a conversation in an echo chamber. The terminator is designed to make the end of the cable look like it goes on forever, thus no reflection. It does this by terminating a conductor with a resistor (or other components) to imitate the inductance and capacitance of a cable of infinite length.

Q.Do all types of SCSI terminators serve the same purpose? Why the different types?

A. All types of SCSI terminators serve the same basic function. Different terminators types were developed because there are different applications which need terminations to look and act in certain ways. For example, to imitate the twisted pair in a SCSI data cable, a 105 ohm resistor might be used as a termination.   LVDSE (Differential) terminators require two wires per signal at the connector, while SE (single ended) terminators assume one wire of the pair is connected to a common ground (Single Ended–SE–terminators using the “common ground” approach do not work as well as other “newer” approaches, but SE terminators were the first terminators developed).

The nominal voltage of a single ended SCSI cable is 2.85 volts. This is created with a voltage regulator that the terminator resistors connect to. The voltage of the regulator should not change when termination resistors are driven high or low. Diodes have also been used in forced perfect terminators to capture the energy that would otherwise be reflected back.

As was mentioned earlier, a terminator needs to be at the end, not in the middle, of a cable. A terminator in the middle of several cables connected in series from device to device would degrade the signal. What do you do when you need a terminator in the middle but do not want to degrade the signal? The answer is to turn off the terminator when necessary. You can do this by using active ICs rather than using passive resistors.

Q.How is the performance of a terminator characterized? How does our test equipment test these characteristics?

A. SCSI terminators have some high speed characteristics but are usually characterized by tests with low speed electronics to more accurately measure the resistances and voltage levels involved; high speed tests require very expensive equipment and are easily distorted by the measurement interfacing.

Unitrode and Dallas Semiconductor make chips used in SCSI terminators that support the LVD/SE/HVD modes used on the newer SCSI lines. We test terminators that use either manufacturer’s chips. The terminators’ resistors are built into the three chips typically used on these designs. The resistors can’t be tested unless the terminator is turned on.

SE terminators (the older-style, Single Ended active terminators) consist of a 2.85 volt regulator tied to a bank of resistors that terminate the data lines. The resistors on these terminators can be tested on any Cirris tester that supports Complex Test, but the regulator should be powered up to make sure it still works. Testing of the regulator would not be done on the Cirris tester in these instances.

Q.How does the Touch 1 Terminator Test, differ from the dedicated Cirris SCSI Terminator Tester?

A. Manufacturers of cable assemblies are generally interested in two types of tests: (1) Verifying that the assembly was put together with the correct components (without electrical damage), and (2) Verifying that the component values in the chips are correct.   The original Cirris SCSI Terminator Tester was designed for customers in category (2). It is capable of making measurements on terminators at better than 1% accuracy because of the HP Meter which is part of the test device. This is particularly useful for manufacturers who laser-trim resistors and then want to make sure that they did it correctly. Very few customers do this anymore, (probably none now that SE and LVD/SE chips are available).

The Touch 1 Terminator test was designed for customers in category (1). The high-speed test was developed to completely test the terminators for correct assembly and possible damages due to static electricity.

Q.What is in the future for such devices and how much variation can we expect?

A. Special scripts for the Touch 1 along with other adapter sizes can be manufactured to support future terminator types. This approach does lack some of the precision of the dedicated SCSI Terminator Tester, however, the new versions of the Touch 1 tester (both the 1000V & 1500V versions) now support accuracy in the 1% tolerance range.

Q.What hardware do I need to test terminators and how does the hardware work? What is the test really doing?

A. As mentioned above, the Touch 1 tester has scripting capability which will allow it to be configured to test a wide varity of configurations. Testing terminators is just one of those options.   The table below lists the part numbers and descriptions of the hardware items needed to test terminators.

The table below list the part numbers, descriptions and prices of the hardware items needed to test terminators.

QUANTITY ORDER NUMBER DESCRIPTION
ONE (1) C101-T1 Touch 1 (1000V) – Base Unit
(Base unit includes 128 test points.)
ONE (1) C101-TA Touch 1 (1000V) – Expansion Box
You will need at least one expansion box possible two boxes, if you wish to test the cable while you test the terminator
ONE (1) SCPT-T1 Scripting (software option) for Touch 1
ONE (1) KSTT-T1 SCSI Terminator Kit
ONE (1) KD5P-50 50 Pin SCSI Terminator Adapter for Touch 1
ONE (1) KD5P-68 68 Pin (Female) SCSI Terminator Adapter for Touch 1
ONE (1) KD5S-68 68 Pin (Male) SCSI Terminator Adapter for Touch 1
ONE (1) KHED-50 50 Pin (Male) .1″ SCSI Terminator Adapter for Touch 1
ONE (1) KHEF-50 50 Pin (Female) .1″ SCSI Terminator Adapter for Touch 1
ONE (1) KBRM-50 50 Pin (Female) Champ Style Terminator Adapter for Touch 1
ONE (1) KBRF-50 50 Pin (Male) Champ Style Terminator Adapter for Touch 1
ONE (1) K8MM-68 68 Pos .8MM PCMCIA Terminator Adapter for Touch 1

Low Voltage Tests
What tests does the Touch 1 actually perform?

In the process of doing the standard low voltage continuity test (shorts, opens, and miswires), the Touch 1 tester will verify that the terminator has all the connections it should have as well as verify that those same connections are quality connections. This is done by checking to insure that all connections have resistance levels below the operator selected resistance threshold. During this test the term power line and the ground pins are checked to be sure they are less than the selected connection resistance value. The tester then measures the term power line capacitor value and verifies it to be within the manufacture’s specified tolerance. The wire list may also contain additional components to be tested, depending on the terminator type.

 

SCSI Terminator Script Testing

After a successful low voltage continuity test has been completed, the Touch 1 switches to a special Script file which is designed to test special features of the particular terminator type being tested (if the low voltage test fails for any reason the problems displayed by the tester will need to be corrected before the terminator script tests have begun). The terminator script test is divided into several sections. The test options applied are based on the type of terminator being tested. The tolerances specified are based on the specifications obtained from each terminator manufacture. This phase of the terminator test usually consist of:

  • Test of Mode Switching
  • Test of Voltage Outputs
  • Resistance Testing

 

Mode Testing

For those terminators that have the ability to be switched between modes, the lines that provide the mode switching are toggled and the terminator outputs are verified to have switched properly. In addition, the input voltage values are verified. If there is a manufacturer’s defined delay specified when switching between modes. That delay is verified and implimented to allow the device to properly switch the output circuitry.

 

Output voltage testing

The output voltages are verified to be within the manufacturer’s specified voltage limits. The terminator is powered and then a voltage measurement is made on each of the output lines. If the terminator switches the output voltage for each of the modes, all of the terminator outputs are verified in each mode. For LVDSE terminators this means the part is powered, the diffsense line is switched and then in all modes–LVD, SE, HVD–the output voltages are verified. For single ended terminators, all output line voltages are verified to be within the specified range.

 

Resistance Testing

After the output lines have been verified for proper voltage output, the resistance of each line is measured. To accomplish this the terminator mode is set. Then a current source is applied to the line. When the voltage stabilizes, the voltage on the line is measured. With the known current and now measured voltage, the resistance for the line is calculated, then verified to be within the specified range. This is done at 2 separate current levels to ensure accurate measurements are being taken.

 

What Terminators Can Be Tested on The Touch 1?

The table below lists the terminators for which Cirris has written tests. If your terminator hardware is not listed, we encourage you send us samples of the terminator so we can qualify it and add it to our growing list of terminators that can be tested by the Touch 1 (the highlighted numbers below are links to pictures of what those particular terminator boards look like).

TERMINATOR TYPES LIST OF TERMINATOR HARDWARE # OF POSITIONS
ADAPTER 50 pos. to 68 pos. Adapter Discrete, AN DS2107, AN UC5603. 68 POSITIONS
Differential Passive DISCRETE (Resistors Only, No IC’s) 68 POSITIONS
Differential Passive DISCRETE (Resistors Only, No IC’s) 50 POSITIONS
ForcedPerfect Resistors at 191 OHMS 68 POSITIONS
ForcedPerfect Resistors at 127 OHMS 68 POSITIONS
LVDSE ALLCHIP, AR7500, AR7519, DS2117, DS2118, DS2119, G218, G219, G218B, GMT1124, HE613, IMP5241, LX5241, ST18, UCC5510, UCC5630, UCC5680, UCC5672, 68 POSITIONS
LVD ONLY DISCRETE, DS2117, DS2118, DS2119, G218A, G219, G218B, LX5241, LX5249, UCC5630, UCC5640. 68 POSITIONS
SEActive (SingleEndActive) DS2119, GMT1121, DS2114, Discrete-50, Discrete-68 68 POSITIONS
SEACT-NEGATION (SingleEndActive Negation) MOT142237, DS2114, DS2107, IMP5115, IMP5225, UCC5619, U560 68 POSITIONS
SEACT-NEGATION (SingleEndActive Negation) IMP5115, IMP5225, DS2107, DS2105, UCC5617 50 POSITIONS
SE-PASSIVE (SingleEndPassive) DISCRETE (Resistors Only, No ICs) 50 POSITIONS