Text preview for : 5991-3959EN Master Your MIPI M-PHY Receiver Tests - Application Brief c20140830 [7].pdf part of Agilent 5991-3959EN Master Your MIPI M-PHY Receiver Tests - Application Brief c20140830 [7] Agilent 5991-3959EN Master Your MIPI M-PHY Receiver Tests - Application Brief c20140830 [7].pdf
Back to : 5991-3959EN Master Your M | Home
Keysight Technologies
Master Your MIPI M-PHY
Receiver Tests
Application Brief
How to Verify MIPI Conformance Using the
J-BERT M8020A High-Performance BERT
Overview
The wide spread and rapid development of cell phones from simple voice-capable devices to
smart phones or even tablet PCs with the constant addition of capabilities and features came
along with a proliferation of interfaces between ASICs and "sensors" and "terminal devices"
such as microphones, cameras, loudspeakers, displays, and peripheral electrical devices as
depicted in Figure 1a, which made ASIC development and system integration an increasingly
difficult task.
Figure 1a. Smartphone with a variety of interfaces to sensors, terminal, and electrical peripheral devices. Figure 1b. Block diagram and the use of buses
standardized through the MIPI alliance.
The MIPI organization was founded in 2003 (MIPI standing for Mobile Industry Processor
Interface) in order to structure the intestines of "mobile devices ranging from smart phones,
wireless-enabled tablets and netbooks" and to "benefit the entire mobile industry by establishing
standards for hardware and software interfaces in these devices" enabling reuse and
compatibility in mobile devices making "system integration less burdensome than in the past."
"The distinctive requirements of mobile terminals drive the development of MIPI
Specifications."1
In Figure 1b the high speed interfaces are marked with red arrows. By not defining these
interfaces as monolithic blocks (such as it is done in computer standards e.g. USB or PCIe) but
instead separating the Phy- from the protocol-layer, it was possible to address the variety of
sensors and terminal devices with only two different high speed Phy-layers.
Figure 2 shows the different MIPI-standards, which protocols reside on D-PHY and M-PHY
respectively, which application they serve, and which kind of test equipment is used for either
protocol verification or validation of Phy-layer interoperability/conformance to specifications.
1. http://mipi.org/momentum
2
Application
Protocol CSI-2 DSI-1
DSI -1 CSI
CSI -3 DSI-2
DSI-2 UFS
UFS
CSI-2 DigRF SS
standard camera
Camera display
Display DigRF LLI
LLI
SS
M-PCIe
MEX
v4
v4 IC
IC
interface
Interface interface
Interface DigRF
DigRF
UniPro
UniPro
v3
v3
Physical
D-PHY M-PHY
M-PHY
standard
Protocolexerciser,
Protocol Exerciser, Scope, BERT
Scope, BERT,
Analyzerbased
analyzer Based ENA, TDR based
ENA, TDR Based
solution
Solution solution
Solution
Figure 2. Structure of MIPI high-speed digital standards with separate protocol- and Phy-layer
In M-PHY, for high speed transmission, 4 speed classes (so-called gears) have been defined/
projected with two slightly different rates per gear. Starting at 1.248 or 1.458 Gb/s for Gear 1A
or 1B respectively the data rates doubles from one gear to the next reaching to the 10 Gb/s
range with Gear 4. As a link can consist of a multitude of lanes, with data being striped across
the lanes, even higher aggregate BW can be achieved.
What needs to be tested?
MIPI does not run a compliance program because the specified interfaces are not user
accessible (other than for the above mentioned computer standards). Nevertheless,
semiconductor vendors and system integrators are interested in validating spec conformance
of ASICs and modules in order to verify interoperability and by this allow flexibility in design
and selection of parts and vendors.
The sheer amount of tests specified in the CTS (Conformance Test Suite) requires automation
for in-situ calibration and test, which Keysight Technologies, Inc. delivers with its N5990A
option 165, available for all BERTs, i.e., the heritage models as well as the new J-BERT M8020.
Throughout the remainder of this document we will now focus on one important aspect of
testing according to the M-PHY standard, i.e., the high-speed receiver (HS_RX) jitter tolerance
test. The definition and the very complex jitter cocktail required for M-PHY high speed receiver
(M_HS_RX ) test is visualized in Figure 5 (for details see application note Keysight Publication
number 5991-2848EN).
3
Figure 3. Jitter "cocktail" for M_HS_RX test.
Step Action J-BERT parameter Target value Pattern TIE-HP-filter
1 Adjust wideband RJ (>fL_RX) to achieve STRJ RJ 0.10 UIpp clk/2 (1010) 1/30 UI
7.9 mUIrms
2 Add low frequency RJ (<1/30 UI) to achieve BUJ 0.17 UIpp clk/2 (1010) fL_RX
RJ=TJ-DJ 13.5 mUIrms
3 Turn all RJ off; calibrate SJ PJ1 0.15 UI,pp CJPat Off
(f_SJ1, f_SJ2, f_SJ3, f_SJ4)
4 Turn all SJ (and RJ) off; calibrate STSJ PJ2 0.2 UI,pp CJPat Off
(240 MHz) to achieve STDJ (STSJ=STDJ-DDJ)
Table 1. Steps to achieve proper calibration of jitter for RX testing (numbers refer to M-PHY CTS rev. 1.0).
4
Solution Utilizing J-BERT M8020A
high-performance BERT
The setup for calibration and test is shown in Figure 4. It is very simple and can be created in
a short time as J-BERT M8020A with its integrated jitter sources is a perfect match for this
complex jitter cocktail.
Figure 4. Actual setup for calibration and test realized with a Keysight J-BERT M8020A and N4915-60001
ISI Trace as conformance channel.
Setting up the jitter cocktail according to Table 1 and performing in-situ calibration manually
using the M8020A's GUI [M8070A] and Keysight DSA 90000 oscilloscope connected at the rep-
lica channel (see Figure 4) is a relatively simple task because of the 1:1 match of the M8020A
impairment (jitter) sources and the M-PHY specification parameters, which will be simplified by
a future release of the N5990A automation SW.
Figure 5. Screenshots from J-BERT M8020 GUI (left) showing set up of jitter cocktail and screenshots
from DSA 91304A oscilloscope (right) showing measurement of the individual Jitter components.
5
Jitter tolerance same as most other tests are performed on a single channel. However, for
multilane realizations additional skew tests for TXs as well as for RXs are required. J-BERT
M8020 being configurable for multiple lanes will provide synchronous start of patterns with
individual skew per lane with a future SW release, and by this will as well master this task as
it is e.g. demanded for LLI (see Figure 2).
Conclusion
As shown above, the requirements for testing ASICs according to the MIPI specification and
CTS are perfectly matched by the J-BERT M8020A: Its high integration allows the test setup
to be simple despite the required complex jitter cocktail and its test automation makes in-situ
calibration and testing an easy task.
Furthermore, M8020A with its modularity provides the flexibility to adapt to test needs and
budget. For example, different speed options can be selected, depending on the speed class to
be addressed. The M8041A can also be configured as a single or dual channel module and an
additional M8051A module can be added, depending on the number of lanes to be tested. They
can also be configured as a generator only (omitting the ED), when the DUT shall not be tested
in loopback mode or doesn't support it.
Further details about features and specifications can be found in the data sheet.
Publication title Publication number
J-BERT M8020A High-Performance BERT