There are two WiFi6 wireless routers that use the same CPU and wireless chip, but one has a built-in “amplifier” and the other has an external “amplifier”. This article is to test the difference between the built-in FEM and the external FEM. The two wireless routers are Skyworth’s SK-WR9540X and SK-WR6640X, the end of the X represents support for 802.11ax, that is, WiFI6, if not X is the WIFI5 version of AC1200.
The same body shell.
First disassemble the machine to see the hardware chip it.
Removing the shield of WR6640X, there is no chip underneath.
Because I see no network port, I define the back of the main board in the above picture.
The WR9540X has a flash memory chip on the back of the motherboard, model number W25N01GVZEIG, with 128MB capacity.
Take off the antenna plug, turn the motherboard over to see the front of the motherboard.
Both machines have a cooling aluminum plate at the bottom of the motherboard.
Continuing with the removal of the shields on both motherboards.
The CPU and wireless chip of Skyworth’s two wireless routers are the same, and the model number is unified here.
The CPU model is ZTE ZX279128R, which belongs to the PON ROUTER integrated SOC, used here as the CPU of the wireless router, belongs to the A9 architecture, dual-core 1GHz, integrated with 4 PHY Ethernet, with 2 PCIe interfaces, can be connected to the wireless chip. Heat generation is not low, compared with the same architecture SD5117P, its heat generation is considered large, may not be the same process.
The wireless chip is shown below, MT7975DN and MT7905DAN form the “MT7915DAN”. All MTK’s AX1800 wireless specifications use this set of wireless chips.
WR9540X does not use an external FEM chip, using the MT7975DN built-in PA and LNA.
This WR6640X has an external FEM chip, the 2.4G FEM chip model is: KCT8247HE.
The parameters of KCT8247HE are as follows.
Integrated high performance 2.4GHz PA, LNA with bypass, and T/R switch
Transmit gain: 30dB at 5V
Receive gain: 15.5dB at 5V
Noise Figure: 1.5dB at 5V
+20dBm @ DEVM=-43dB, HE40/MCS11, 5V
+21dBm @ DEVM=-40dB, HE40/MCS11, 5V
+22.5dBm @ DEVM=-35dB, VHT40/MCS9, 5V
+23.5dBm @ DEVM=-30dB, HT20/MCS7, 5V
ESD protection circuitry on all PINs
Fully-matched input and output
Integrated Power Detector
Minimal External Components Required
Small package: MIS 16-pin, 3×3×0.55mm (MSL3, 260 oC per JEDEC J-STD-020)
RoHS and REACH Compliant
In WiFi5 mode, the amplification power is +22.5dBm = 178mW.
In WiFi6 mode, the amplification power is +20dBm = 100mW.
The 5G FEM chip model is KCT8575HE.
The parameters of KCT8575HE are as follows.
Integrated 802.11ax 5GHz PA, LNA with bypass and T/R switch
Fully-matched input and output
Integrated power detector
Transmit gain: 32dB at 5V
Receive gain: 16.5dB at 5V
Noise Figure: 1.7dB at 5V
+20dBm @-43dB DEVM, HE160//MCS11, 5V
+21.5dBm @-40dB DEVM, HE160//MCS11, 5V
+23.0dBm @-35dB DEVM, VHT80/MCS9, 5V
+24.0dBm @-30dB DEVM, HT20/MCS7, 5V
Integrated 2.4GHz Notch Filter
ESD protection circuitry on all PINs
Minimal external components required
Small package: MIS16-pin, 3mm x 3mm x0.55mm (MSL1, 260 oC per JEDEC J-STD-020)
ROHS and REACH Compliant
In WiFi5 mode, amplification power +23dBm = 200mW.
In WiFi6 mode, the amplification Shi rate +20dBm = 100mW.
In the -43dB DEVM amplification power, +20dBm is considered relatively high, such as SKY85755-11 and QPF4588, but also just +18dBm.
The memory chip of both wireless routers is: SCB13H2G160 (To see the full picture of the memory chip, I cut off the metal blocker.)
“SCB13H2G160EF-09N FBGA96 128*16 2G DDR3 memory chip IC”.
Memory capacity 256MB.
The flash memory models are: W25N01GVZEIG, capacity 128MB.
The hardware configuration information of the two models is compared as follows.
The difference between WR6640X and WR9540X is the red content in the above chart, that is, the difference between the independent amplifier and the wireless chip built-in amplifier. Everything else is the same!
The next is to conduct a comparative test of wireless performance, to see which of the two strong.
All the following tests, 2.4G channel 3 40MHZ bandwidth, 5G channel 44 80MHz.
Second, the wireless signal strength comparison test
Two wireless routers placed in the following chart “WIFI” location, with a wireless card AX200 computer were placed in A, B, C, D four locations for speed measurement (D point location is on the second floor, ABC and WIFI location on the second floor, there is no concrete wall). Each location recorded signal strength time of 3 minutes. In order to avoid signal interference with each other caused by errors, measured one and then another, are placed in the same position, the antenna vertical up.
The results are as follows: (in the left column, the signal is strong in the green area and weak in the red area)
At the location of point A, which is separated by a wall, the 2.4G signal strength is very close to both, and the 5G signal strength is WR6640X a little stronger.
At the location of point B, which is separated by two walls, the 2.4G signal strength of the two is still very close, and the 5G signal strength is a little stronger for the WR6640X.
2.4G or similar at the position of point C, which is 11 meters away in a straight line, and 5G or WR6640X is a little stronger.
Finally downstairs in the D point location, 5G signal strength is far apart, WR6640X obviously stronger, while the 2.4G signal is WR9540X strong so a little.
From the above results, 2.4G signal strength difference is not obvious, 5G signal strength in four different locations are WR6640X strong points.
Third, the wireless speed measurement comparison at different distances
Self-built speedtst speed measurement server connected to the WAN port of the reception test wireless router, with different wireless terminals connected to the wireless speed measurement, the placement of the devices as in the previous test signal project.
Each device is connected schematically as follows.
Each device each location six times continuously measured speed, and finally take the average value for comparison.
2.4G only in the two more distant location C and D two location speed measurement, in different time speed measurement 2.4G, the results will vary greatly, especially in the evening time, so the following 2.4G are in the morning before 11 o’clock to do the test.
1, using Intel AX200 wireless network card for speed measurement comparison
AX200 is the highest occupancy rate of WiFi6 wireless network card on the market, there is no way, now you can buy only it.
All test results are shown in the table below: (Mbps)
The results of the 5G speed average comparison are as follows.
In all four positions, the WR6640X has faster 5G speeds than the WR9540X.
The 2.4G average results are as follows.
2.4G transmission speed is not significantly different between the two.
2. Speed measurement comparison using Xiaomi 10 phone
All the results of Xiaomi 10 speed measurement are shown in the table below: (Mbps)
The results of the 5G speed measurement average junction ratio are shown in the following graphs.
Xiaomi 10 5G speed measurement in four locations, all WR6640X is faster.
The 2.4G average is as follows.
2.4G speed is about the same.
3. Speed measurement comparison using Apple SE2 cell phone
All the results are shown in the table below: (Mbps)
5G average comparison.
The WR6640X is significantly faster in the close proximity of points A and B, and about the same in points C and D.
The 2.4G average is as follows:
4, using Huawei mate20 phone for speed measurement comparison
Huawei mate20 only supports WiFi5, but also want to see how the speed measurement comparison will be under wifi5.
All the results are in the following table: (Mbps)
The average results of 5G speed measurement are as follows.
At the two locations A and B, the WDR6640X’s 5G speed is faster, and at the distant points C and D the two speeds are very close.
2.4G average comparison.
In the two positions of medium and long distance, 2.4G is WR6640X faster.
From the above test results of the four different wireless terminals, the 2.4G wireless performance of the two is basically similar, while the 5G performance is clearly WR6640X advantage some.