IPC-TM-650 EN 2022 试验方法.pdf - 第473页

aberrations will increase the variation in the average voltage or reflection coefficient value from which Z 0 is computed, thereby increasing the variation in the compute value of Z 0 . 4.3 Other Equipment Requirements 4…

Intermediate

values can be linearly interpolated from Table 4-I

or using:

sys

≤

For

example, if a 32 mm [1.26 in] long transmission line was

being measured, a TDR system with t

sys

≤ 80

ps should be

used. Note that, if the probe launch caused excessive ringing

in the TDR waveform, or if the launch does not repeatably

replicate the connection to the standard, then the 0.25 factor

may need to be smaller.

4.2

TDR Requirements

4.2.1 Impedance

The

impedance of the TDR unit should

be 50 Ω with an impedance uncertainty less than or equal to

± 0.5 Ω. This TDR impedance value is selected because it is

the impedance used by most high-speed/high-frequency test

instrumentation and compatibility with this instrumentation is

necessary for characterizing the dynamic TDR properties,

such as its impulse response (or transfer function). The imped-

ance of the TDR unit should be calibrated using an artifact

standard, such as an air line (see 4.3.6). However, the TDR

impedance is a function of frequency and calibration using a

fixed region of the TDR waveform (the measurement zone) will

only yield an average impedance value for the TDR unit for the

corresponding frequency range.

4.2.2

Timebase Accuracy

The

horizontal timebase accu-

racy defines how well the TDR instrument’s horizontal time

scale can display the correct length of the trace. This affects

both the accuracy of the measurement zone calculations and

any propagation delay values. The timebase accuracy should

be less than 0.25 t

sys

(see

also 4.1.2).

4.2.3

Step Aberrations

The

ability of the TDR instrument

to measure the impedance of a transmission line is related to

how well the instrument can generate a step-pulse with a

minimum of aberrations (ringing, overshoot, undershoot, set-

tling, etc.). Any ringing, overshoots, or undershoots will cause

corresponding aberrations in the TDR waveform (see Figure

4-2). These aberrations can cause significant errors in the

impedance value computed from the TDR waveform. Addi-

tionally, low frequency step aberrations may produce a ramp

in measurement zone. This ramp can cause a significant bias

in the computed impedance value. The TDR instruments step

aberrations should be less than 1% of the total step ampli-

tude. For example, the impedance error shown in Table 4-2 is

fora1mVerror of a 250 mV step. Poor settling and large

able 4-I Resolution of TDR Systems

TDR

System

Risetime

Resolution 4X Resolution

0ps 5ps/1mm[0.04 in] 4 mm [0.16 in]

20ps 10ps/2mm[0.08 in] 8 mm [0.31 in]

30ps 15ps/3mm[0.12 in] 12 mm [0.47 in]

100 ps 50 ps/ 10 mm [0.39 in] 40 mm [1.57 in]

200 ps 100 ps / 20 mm [0.79 in] 80 mm [3.15 in]

500 ps 250 ps / 50 mm [1.97in] 200 mm [7.87 in]

IPC-2257a-4-2

Figure

4-2 Potential TDR Step Aberrations

overshoot

undershoot

ringing

low frequency drift

IPC-TM-650

Number

2.5.5.7

Subject

Characteristic

Impedance of Lines on Printed Boards by TDR

Date

03/04

Revision

age4of23

电子技术应用　　　　　　　ｗｗｗ．ＣｈｉｎａＡＥＴ．ｃｏｍ

aberrations

will increase the variation in the average voltage or

reflection coefficient value from which Z

computed, thereby

increasing the variation in the compute value of Z

4.3

Other Equipment Requirements

4.3.1 Connectors

TDR

systems typically come with either

‘‘SMA,’’ 3.5 mm, or 2.92 mm connectors at their measure-

ment ports. SMA, 3.5 mm, and 2.92 mm connectors are all

50 Ω connectors and are electrically and geometrically com-

patible, therefore, they can be mated directly to each other.

However, the 2.92 mm and 3.5 mm connectors are precision

connectors (have a lower impedance uncertainty than the

SMA) and are designed to provide a more repeatable connec-

tion than the SMA connector. Therefore, for accurate mea-

surements, it is recommended that the 2.92 mm or 3.5 mm

connector be used where possible. The bandwidth of the

connectors must be great enough so that the connectors do

not affect the accuracy of the TDR measurement. The typical

-3 dB bandwidth of 3.5 mm connectors is approximately

34 GHz and of SMA connector is approximately 24 GHz. The

reflection and insertion losses of the connector should be less

than 27 dB and 0.3 dB respectively. Other connectors, com-

parable in performance to the 3.5 mm connector, may also be

used. All cable connections using SMA, 3.5 mm, or 2.92 mm

connectors should be tightened with a torque wrench to fol-

lowing specification, unless otherwise specified by the manu-

facturer of the connector or cable:

4.3.2

Cabling

All

test cables should be coaxial and have a

characteristic impedance of 50 Ω with an impedance uncer-

tainty of less than ± 1 Ω. Cables used in the measurement

circuit of the transmission line under test should have connec-

tors that are compatible with the rest of the measurement

system. The bandwidth of the cable must be great enough so

that the cable does not affect the accuracy of the TDR mea-

surement. The length of the cables should be kept to a mini-

mum. The total insertion loss (including connector loss) of the

cabling connecting the transmission line under test to the TDR

should be kept to a minimum, for example, less than

0.033 dB/cm [1db/foot] at 26.5 GHz. Table 4-4 contains sug-

gested maximum cable lengths. Faulty cables can contribute

up to a 1 Ω error. Cable connections should be tightened with

a torque wrench to ensure a good connection.

4.3.3

Probes

The

probe assembly should have a charac-

teristic impedance of 50 Ω or of approximately the same value

as that of the transmission line under test, with an uncertainty

of ±1.0 Ω or less. The probe tips should be of sufficient diam-

eter and pitch (spacing between signal and ground tips) to

accurately and repeatably probe the desired probe contact

pad geometry. (See IPC-2141 for recommended probe land-

ing layouts for TDR coupons.) Single-ended probes should

contain two tips, one each for the signal and ground lines.

Differential probes should contain two tips for contacting the

signal lines and one or two tips for contacting the reference

plane or planes. The probe tips should have moderately sharp

edges to cut through any oxides. For hand held probe assem-

blies, the probe handle should be ergonomically shaped. The

probe bandwidth should be sufficient for the desired

temporal/spatial resolution (see 4.1.2). The probe settling time

should be short so as not to affect the duration of the mea-

surement zone. The overall performance of the probe can be

incorporated into the TDR system response for computing

TDR system temporal/spatial resolution (see 4.1.2). Inconsis-

tent probe force and placement is common and can cause a

significant yet unknown error that can exceed 5 Ω. Probe

connections should be tightened with a torque wrench to

ensure a good connection.

4.3.3.1

Probes for Differential Structures

The

differen-

tial probe should be long enough to act as a transfer stan-

dard, similar to that described in 4.3.7 for testing single-ended

able 4-2 Impedance Error

Impedance

Error

Transmission Line

Impedance

0.24 Ω 28 Ω

0.4 Ω 50 Ω

0.79 Ω 90 Ω

1.23 Ω 125 Ω

able 4-3 Connector Torque Speciﬁcations

[Conversion factor is 0.1128 N-m/(lb-in)]

Connector

Type Required Torque

SMA

0.56 N-m [5 lb-in]

3.5 mm

2.92 mm, K

0.9 N-m [8 lb-in]

Table 4-4 Maximum Suggested Cable Lengths

TDR

Cable Assembly TDR Cable Length

Sampling

Unit to Static Isolation Unit 30 cm [12 in]

Static Isolation Unit to In-Line

Secondary Standard

91 cm [36 in]

In-Line Standard (such as, semi-rigid

coaxial cable)

10 cm [4 in]

IPC-TM-650

Number

2.5.5.7

Subject

Characteristic

Impedance of Lines on Printed Boards by TDR

Date

03/04

Revision

age5of23

电子技术应用　　　　　　　ｗｗｗ．ＣｈｉｎａＡＥＴ．ｃｏｍ

transmission

lines (see 5.2.1). This probe can be constructed

using a printed circuit board differential transmission line that

has a differential impedance similar to the traces being tested

or by using short identical lengths of semi-rigid cable that

connect the instrument cables to the differential probe. The

probe or semi-rigid cable should be sufficiently long to provide

an adequate duration for the measurement zone (see 4.1.2)

used during calibration and measurement.

4.3.4

Terminations

Many

instruments perform differently

under different electrical loads. If a test is performed with open

circuited lines, the calibration of the reference should also be

done using an open circuit termination.

4.3.5

ESD Protection

Static

build up on specimens prior

to test can damage the sampling heads in the TDR equip-

ment. Therefore, it is recommended that ESD protection be

used. Such protection can be supplied internally to the TDR

system or externally. If supplied externally, using a coaxial

switch for example, then the switch should be placed

between the transmission line under test and the TDR instru-

mentation. The switch, or static protection device (SPD),

should have a return and insertion loss less than 16 dB and

0.3 dB at 18 GHz. A maximum of 31 cm [12.2 in] of high

quality, high frequency cable may be used to connect the TDR

instrument to the protection switch. Samples should be

grounded to remove any residual static and/or passed

through some type of deionization device prior to testing.

Keeping the relative humidity in the test area between 45 %

and 55 % may minimize the buildup of static. Automation

software can be used to enhance the effectiveness of the

static isolation unit by switching the static isolation unit on/off

as required to minimize the amount of time that the TDR sam-

pling unit is exposed to potential ESD.

4.3.6

Calibration Artifacts (Reference or Reference

Standard)

precision coaxial air line with calibration trace-

able to a national metrology institute (such as the National

Institute of Standards and Technology), 3.5 mm or 2.92 mm

connectors at both ends, and at least 10 cm [3.94 in] long

should be used. The characteristic impedance of the air line is

based on the geometry of a coaxial transmission line using air

as the dielectric between the center conductor (signal line)

and the ground shield. The center conductor may be held in

place by glass beads located at both ends of the air line or by

the external connectors that are attached to the connectors of

the air line. The uncertainty in the nominal characteristic

impedance of the air line should be less than or equal ±

0.015 Z

ref

, where Z

ref

the characteristic impedance of the

reference air line.

4.3.7

In-Line (Transfer) Standard

The

transfer standard

is placed between the probe and the TDR unit (see Figure

4-3). This standard could be an airline, a semi-rigid coax cable

assembly with the same specifications as the flexible coax

cable assemblies, or other. The transfer standard should be of

IPC-2257a-4-3

Figure

4-3 Reference Airline and Probe Contact Pad

REFERENCE

AIRLINE

PROBE

TRANSFER

STANDARD

PROBE

CONTACT

PAD

ADAPTOR

IPC-TM-650

Number

2.5.5.7

Subject

Characteristic

Impedance of Lines on Printed Boards by TDR

Date

03/04

Revision

age6of23

电子技术应用　　　　　　　ｗｗｗ．ＣｈｉｎａＡＥＴ．ｃｏｍ