Views: 296 Author: Anna Publish Time: 2024-08-14 Origin: Site
Composition and structure of optical module
The optical module works at the physical layer, which is the lowest layer in the OS! model. Its function is simple to say, which is to realize photoelectric conversion. Convert optical signals into electrical signals, and convert electrical signals into optical signals. Although it seems simple, the technical content of the implementation process is not low
An optical module is usually composed of optical transmitters (TOSA, including lasers), optical receivers (ROSA, including photodetectors), functional circuits and optical (electrical) interfaces.
At the transmitting end, the driver chip processes the original electrical signal and then drives the semiconductor laser (LD) or light-emitting diode (LED) to emit a modulated optical signal.
At the receiving end, after the optical signal comes in, it is converted into an electrical signal by the photodetection diode, and the electrical signal is output after passing through the preamplifier.
Packaging of optical modules
Packaging can be simply understood as model standards. Before explaining packaging and classification, let us first introduce the standardization organizations of optical communications. Because these packages are determined by standardization organizations, there are currently several organizations in the world that standardize optical communications, such as the familiar IEEE (Institute of Electrical and Electronics Engineers), ITU-T (International Telecommunication Union), MSA (Multi-Source Agreement), OIF (Optical Internetworking Forum), CCSA (China Communications Standards Association), etc.
The most commonly used in the industry are IEEE and MSA.
MSA, its English name is Muli Source Agreement. It is a multi-supplier specification. Compared with IEEE, it is a non-official organization form of the people. It can be understood as the behavior of the enterprise alliance in the industry.
The optical module packaging classification is shown in the following table:
Classification | Type |
Package Type | 1x9、GBIC、X2、XENPAK、XFP、SFP、SFP+、SFP28、QSFP、QSFP28、CFP、CFP2. QSFP-DD、OSFP etc |
rate | 10Mbps、100Mbps、155Mbps、622Mbps、1.25Gbps、2.125Gbps、4.25Gbps、10Gbps、25Gbps、50Gbps、100Gbps、400Gbps etc |
wavelength | 850nm、1310nm、1490nm、1550nm、 CWDM、DWDM etc |
Mode | Single mode Multi mode |
Distance | 100m、300m、550m、10km、20km、40km、80km、120km、160km |
Modulation format | NRZ、PAM4、DP-QPSK/n-QAM etc |
Optical interface working mode | Deplux、BiDi |
Laser Type | VCSEL、FP、DFB、EML etc |
Optical detector type | PIN、APD |
Connector connector | FC,SC,ST,LC,MU,MTRJ |
Usability | Hot-swappable (GBIC, SFP, XFP, XENPAK) and non-hot-swappable (1*9, SFF) |
Operating temperature | Commercial grade (0~70°C), extended temperature (-20~85°C), industrial grade (-40~85°C) |
Common packages:
GBIC
GBIC, which stands for Giga Bitrate Interface Converter, was the most popular optical module package before 2000 and the most widely used Gigabit module form.
SFP
Because GBIC is relatively large in size, SFP appeared later and began to replace GBIC. SFP stands for Small Form-factor Pluggable, which is a small hot-pluggable optical module. Its smallness is relative to GBIC package. The size of SFP is half that of GBIC module, and more than twice the number of ports can be configured on the same panel. In terms of function, the two are not much different, and both support hot plugging. The maximum bandwidth supported by SFP is 4Gbps.
XFP
XFP is 10-Gigabit Small Form-factor Pluggable, which is easy to understand, that is, 10G SFP.
XFP uses a full-speed single-channel serial module connected by an XFI (10Gb serial interface), which can replace Xenpak and its derivatives.
SFP+
SFP+, like XFP-, is a 10G optical module.
SFP+ is the same size as SFP, more compact than XFP (about 30% smaller), and consumes less power (reduced some signal control functions).
SFP28
The SFP with a speed of 25Gbps was mainly developed because the price of 40G and 100G optical modules was too expensive at that time, so there was such a compromise transition solution.
QSFP/QSFP+/QSFP28/QSFP28-DD
Quad Small Form-factor Pluggable, four-channel SFP interface. Many mature key technologies in XFP are applied to this design. QSFP can be divided into 4x10G QSFP+, 4x25G QSFP28, 8x25G QSFP28-DD optical modules according to speed. Taking QSFP28 as an example, it is suitable for 4x25GE access ports. Using QSFP28, you can upgrade directly from 25G to 100G without going through 40G, which greatly simplifies the wiring difficulty and reduces costs.
QSFP28
QSFP-DD, established in March 2016, DD refers to "Double Density". The 4 channels of QSFP are increased by one row of channels, becoming 8 channels. It is compatible with the QSFP solution, and the original QSFP28 module can still be used, just insert another module. The number of electrical port gold fingers of QSFP-DD is twice that of QSFP28.
QSFP-DD
QSFP-DD uses 25Gbps NRZ or 50GbpS PAM4 signal format for each channel. Using PAM4, it can support a maximum rate of 400Gbps.
QSFP/QSFP+/QSFP28/QSFP28-DD
NRZ and PAM4
PAM4 (4 Pulse Amplitude Modulation) is a "doubling" technology.
For optical modules, if you want to achieve a rate increase, you either need to increase the number of channels or increase the rate of a single channel.
The traditional digital signal uses the NRZ (Non-Return-to0-Zer0) signal at most, that is, high and low signal levels are used to represent the 1 and 0 information of the digital logic signal to be transmitted. Each signal symbol period can transmit 1 bit of logic information.
The PAM signal uses 4 different signal levels for signal transmission, and each symbol period can represent 2 bits of logic information (0, 1, 2, 3). Under the same channel physical bandwidth, PAM4 transmits twice the amount of information as the NRZ signal, thereby doubling the rate.
CFP/CFP2/CFP4/CFP8
Centum gigabits Form Pluggable, dense wavelength division optical communication module. The transmission rate can reach 100-400Gbps. CFP is designed based on the SFP interface, with a larger size and supports 100Gbps data transmission. CFP can support a single 100G signal, one or more 40G signals. The difference between CFP, CFP2, and CFP4 lies in the volume. The volume of CFP2 is half of that of CFP, and that of CFP4 is one quarter of that of CFP. CFP8 is a packaging form specifically proposed for 400G. Its size is comparable to that of CFP2, and it supports channel rates of 25Gbps and 50Gbps. It can achieve a module rate of 400Gbps through a 16x25G or 8x50 electrical interface.
OSFP
This is a bit confusing with the OSPFQ routing protocol we often talk about.
OSFP, Octal Small Form Factor Pluggable, "O" stands for "octal", was officially launched in November 2016. It is designed to use 8 electrical channels to achieve 400GbE (8*56GbE, but the 56GbE signal is formed by 25G DML laser under PAM4 modulation), with a size slightly larger than QSFP-DD, a higher wattage optical engine and transceiver, and slightly better heat dissipation performance.
The above are some common optical module packaging standards.
400G optical module
400G is the main competitive direction of the optical communication industry. Now 400G is also in the early stage of large-scale commercial use.
As we all know, due to the large-scale launch of 5G network construction, coupled with the rapid development of cloud computing and the batch construction of large-scale data centers, the ICT industry's demand for 400G has become more and more urgent.
The early 400G optical module used a 16-channel 25Gbps NRZ implementation method and adopted CDFP or CFP8 packaging. The advantage of this implementation method is that it can borrow the mature 25G NRZ technology on the 100G optical module. But the disadvantage is that 16 channels of signals are required for parallel transmission, and the power consumption and volume are relatively large, which is not suitable for data center applications.
Later, PAM4 began to replace NRZ
On the optical port side, 8 channels of 53GbpS PAM4 or 4 channels of 106GbpS PAM4 are mainly used to realize 400G signal transmission, and 8 channels of 53GbpS PAM4 electrical signals are used on the electrical port side, using OSFP or QSFP-DD packaging.
In comparison, QSFP-DD has a smaller package size (similar to the QSFP28 package of the traditional 100G optical module) and is more suitable for data center applications. OSFP has a slightly larger package size and is more suitable for telecommunications applications because it can provide more power consumption.
Basic indicators of optical modules
The basic indicators of optical modules mainly include the following:
Output optical power
The output optical power refers to the output optical power of the light source at the transmitting end of the optical module. It can be understood as the intensity of light, in W or mW or dBm. Among them, W or mW is a linear unit, and dBm is a logarithmic unit. In communication, we usually use dBm to represent optical power.
The optical power is attenuated by half, reduced by 3dB, and the optical power of 0dBm corresponds to 1mW.
Maximum receiving sensitivity
The receiving sensitivity refers to the minimum receiving optical power of the optical module under a certain rate and bit error rate, in units of dBm.
In general, the higher the rate, the worse the receiving sensitivity, that is, the greater the minimum receiving optical power, and the higher the requirements for the receiving end device of the optical module.
Extinction ratio
The extinction ratio is one of the important parameters used to measure the quality of optical modules.
It refers to the minimum value of the ratio of the average optical power of the signal to the average optical power of the empty number under full modulation conditions, indicating the ability to distinguish between 0 and 1 signals. There are two factors that affect the extinction ratio in the optical module: bias current (bias) and modulation current (Mod), which can be regarded as ER=Bias/Mod. The larger the extinction ratio, the better the optical module is. Instead, the optical module with an extinction ratio that meets the 802.3 standard is good.
Light saturation
Also known as saturated optical power, it refers to the maximum input optical power at a certain bit error rate (10-10~10-12) at a certain transmission rate. The unit is dBm. It should be noted that the photodetector will experience photocurrent saturation under strong light. When this phenomenon occurs, the detector needs a certain amount of time to recover. At this time, the receiving sensitivity decreases, and the received signal may be misjudged and cause bit errors. It is also very easy to damage the receiving end detector. During operation, it should be avoided as much as possible to exceed its saturated optical power.