Previous Page  16 / 54 Next Page
Information
Show Menu
Previous Page 16 / 54 Next Page
Page Background

14

Chip Scale Review January • February • 2019

[ChipScaleReview.com]

ompanies developing 5G

technologies are racing to

develop the first chipsets in

order to set the standard of deployment

and be the leader. While initial standards

for 5G were set at the end of 2017 [1],

and there are ideas about the applications

of 5G (

Figure 1

), it is still unclear how

exactly it will all come together. This

uncertainty demands unprecedented levels

of collaboration and partnership. This

article explores the challenges and changes

in test methodology of 5G devices, and

showcases the results of a collaboration

with Intel [2].

5G brings three technical improvements/

enhancements when compared to the

current deployed, 4G standard:

1. G r e a t e r ava i l a bl e b a ndw i d t h ,

increasing to more than 4GB per

connection per month from today’s

less than 1.5GB per device [4].

2. L ow e r l a t e n c y , f o r c r i t i c a l

applications to be more responsive

[5].

3. The abilit y for up to 1 million

devices (such as sensors and smart

devices, per square kilometer [6]) to

be connected to the network.

Some of the resultant solutions to meet

these requirements for 5G include:

• The opening of millimeter wave

(mmW) frequencies: ~30GHz and

above;

• The increase in the number of mobile

sites to allow for more devices [7];

and

• The deployment of edge cloud nodes

so that data doesn’t always need to

go back to the central node.

Because of all of these changes, the full

network infrastructure will need to be

upgraded. Prior to 5G, the upgrades were

primarily around changing from analog

to digital in the first few generations, as

well as improving modulation techniques

[8]. “So the scale of 5G when compared to

previous upgrades to the infrastructure is

much larger than seen in the past because

of the technology extensions. However, as

Scott Fulton points out, “5G is a capital

improvement project the size of the entire

planet, replacing one wireless architecture

created this century with another one that

aims to lower energy consumption and

maintenance costs [9].”

In order to provide the chips required

for this change in the landscape, there

will be a large number of changing

requirements in wafer test that come

out of these architectural requirements.

Fo r mFa c t o r p a r t n e r e d w i t h I n t e l

t o i nve s t iga t e t he s e cha nge s , a nd

tested one such example of a new test

methodology [2].

The history of mmW testing

Historically, millimeter wave testing

of wafers was relegated to labs and

low-volume production for defense,

a e r o s p a c e a n d o t h e r s ome wh a t -

exotic applications. This is because

of the low transmission range, high

cost of generating RF signals with IC

chips, and low data rates that were

required. Therefore, wafer high-volume

manufacturing (HVM) production floors

topped out at 6GHz because mobile

phones were the devices using a majority

of the RF ICs.

Millimeter wave testing, however,

has been movi ng i nto h igh-volume

production because of automotive radar

and high-speed digital parts that require

the increased performance gained with

higher frequencies. That is either with

more accurate resolution of a nearby

vehicle or obstacle, or more data being

moved in data centers and over fiber

connections. Some of the challenges

identified in automotive test [10] include:

1. Power accuracy;

2. Maintaining RF calibration of the

test equipment (and final RF signal

path);

3. Setting appropriate test limits at

these higher frequencies;

4. Millimeter “anything” is just more

expensive; and

5. Test engineering is not familiar with

mmW testing.

C

New test methodologies for 5G wafer

high-volume production

By Daniel Bock, Jeff Damm

[FormFactor, Inc.]

Figure 1:

Usage scenarios for 5G [3].