Previously bought a, but sent V1, said V1 and V2 mixed shipping, open the package to see V1 directly returned. A few days ago someone said V2 heat and what do not want, sent to the second-hand site for sale, I asked if it was the V2 version and then placed an order.
This year, BOM prices rose significantly, V2 is still selling at $ 65, TP-Link’s competitive market pressure is not small.
The packaging is as follows, the power supply is still 12V/2A, the router bottom version number Ver:2.0, is the V2 version.
Remove the two filaments in the label paper, directly remove the shell, you can see the motherboard:
There is also a label on the motherboard network port that says TL-XDR5430 2.0.
All antennas are soldered on the motherboard, the back of the motherboard is very difficult to flip, before the V1 BoT version more difficult, you need to solder down an antenna to do.
The back side looks like.
To look at the thickness of the heatsink from another angle.
Take the shield off, there is a piece of thermal conductive silicone gasket inside the shield, there is no major chip inside the shield, this location is directly below the CPU chip. Since there is nothing to see, the shield is installed back!
After removing the screws from the back heatsink, a small piece of the left side was found to be independent, and this location is directly below the 5G wireless chip. Probably to strengthen the wireless chip and the FEM chip heat dissipation.
Heat sink are taken off.
The bottom of the motherboard does not have any chips, so the heatsink glued back in place, flipped over to see the front of the motherboard.
The heatsink on the front of the motherboard is removed.
Continue to remove all metal shields, three shields have a piece of thermal conductive silicone gasket underneath.
There is a metal chip above the CPU chip, and then a ceramic heatsink, why this design?
In order to see the CPU chip clearly, you have to turn this iron piece upward with force, and turn over the iron. The CPU side can be seen.
The CPU of XDR5430v2 is Qualcomm IPQ5018. This is the third time I’ve seen it, the first time was on Xiaomi AX6000 and the second time was on H3C BX54.
Above the CPU is a memory chip, model M15T2G16128A, with 256MB capacity.
IPQ5018 dual-core 1GHz with integrated 1GHz NP, there is also a built-in 2.4G Wi-Fi 6 wireless chip that supports 2x2mimo with a maximum rate of 574Mbps.
There are two external 2.4G FEM chips, model number QPF4216 (same as XDR5430v1). There is also a flash memory chip here, model number MT25QU128, with 16MB capacity.
The parameters of QPF4216 are as follows
The QPF4216 integrates a 2.4 GHz power amplifier (PA), regulator, single pole two throw switch (SP2T), bypassable low noise amplifier (LNA) into a single device
• POUT = +22 dBm MCS11 HE40 -43 dB Dynamic EVM
• POUT = +24 dBm MCS8/9 VHT40 -35 dB Dynamic EVM
• POUT = +25 dBm MCS7 HT20/40 -30 dB Dynamic EVM
• POUT = +26 dBm MCS0 HT20 Spectral Mask Compliance
• Optimized for +5 V Operation
• 32 dB Tx Gain
• 1.8 dB Noise Figure
• 15.5 dB Rx Gain & 7 dB Bypass Loss
• 15 dB 5 GHz Rejection on Rx Path
In Wi-Fi 6 and 40MHz bandwidth 2.4G amplification power +22dBm = 160mW. count high power!
Now look at the 5G no chip, the model is QCN6024, to a large close-up it:
The QCN6024 supports 4x4mimo with a maximum 5G wireless rate of 4804Mbps at 160MHz bandwidth, plus 574Mbps at 2.4G, so this XDR5430v2 version is also the wireless specification of AX5400.
What is the difference between the QCN6024 and QCN9024 and QCN9074?
The main difference is the wireless band capacity and the number of OFDMA clients, and the QCN6024 does not support 4K-QAM, so I used Xiaomi 11 to connect XDR5430v2 with 5G, and the maximum wireless rate is 2401Mbps, unlike Xiaomi AX6000 and H3C BX54 which is 2882Mbps.
The main difference can be referred to the following table: (data updated on Feb 23, 2021)
Compared to the QCN9074 and QCN9024, you can see the QCN6024 as a lite or basic version.
5G wireless external with a quad FEM chip, model QPF4506.
QPF4506，I bought 6 of them before at $1.50 a piece. Article Link：
QPF4506 has two models, the other is QPF4506B, “B” is a custom model of Broadcom, different manufacturers of wireless chips for EVM requirements are a little different, some like the peak value, some require a high number of concurrent, such as MU-MIMO effect, OFDMA maximum number of clients.
XDR5430v1 version is the Broadcom program, 5G FEM chip model is also QPF4506, I do not see the QPF4506B valid mark, V1 may also be used QPF4506B, no matter how, the parameters of the two QPF is the same.
QPF4506 parameters are as follows.
• POUT = +15 dBm MCS11 HE160 -43 dB Dynamic EVM
• POUT = +21 dBm MCS9 VHT80 -35 dB Dynamic EVM
• POUT = +23 dBm MCS7 HT40 -30 dB Dynamic EVM
• POUT = +25 dBm MCS0 HT20 Spectral Mask Compliance
• Optimized for +5 V Operation
• 30 dB Tx Gain
• 1.7 dB Noise Figure
• 13.5 dB Rx Gain & 6.5 dB Bypass Loss
• 28 dB 2.4 GHz Rejection on Rx Path
If you take the H3C BX54 and Xiaomi AX6000 5GFEM to compare, the amplification power from low to high in this order.
XDR5430v2 (QPF4506) < Xiaomi AX6000 (QPF4588) < H3C BX54 (SKY85333-11)
Later in the D point location speed measurement to see.
The last thing to look at is the switch chip model is RTL8367S, as follows.
For the speed bottleneck of all LAN ports of Xiaomi AX6000, you can refer to the following article：
And XDR5430V2 this route, because there is no 2.5G network port, and RTL8367S and IPQ5018 both support SGMII + interface (2.5Gbps rate), so the transmission speed of each LAN port and wireless will not have a 1Gbps bottleneck.
Here a simple practical test to know, two computers were connected to the XDR5430v2 LAN1 and LAN2, using ixchariot while playing streams to Xiaomi 11 (5G rate 2402Mbps), the results are as follows.
The total speed of the two LAN ports to 5G wireless is 1750Mbps now, which exceeds the gigabit speed bottleneck, proving that the SGMII+ interface (2.5Gbps rate) was used to connect.
At this point, the XRR5430v2 version is disassembled and completed, you can also it do the XDR5430 Qualcomm version.
The main chips of XDR5430v2 made a summary chart as follows.
The configuration comparison table of the two versions of XDR5430 is as follows.
Finally in the downstairs D point to do a simple speedtest broadband speed test, telecom broadband 300 megabytes. The router is placed in the second gear in the following figure “WIFI” location, the phone is placed on the first floor of the D point location for speed test: (only 5G, channel 44, open 160MHz bandwidth)
I put the previous H3C BX54 and Xiaomi AX6000 results also put together to compare the look.
The speed measurement of Xiaomi 4C cell phone which only supports wifi5 is as follows.
Xiaomi 4C in the D point position, but also over 200 megabytes, more than 200 are considered relatively strong 5G, but not as strong as Xiaomi AX6000 and H3C BX54.
Then the Xiaomi 10 test results.
Also is not as good as Xiaomi AX6000 and BX54.
Finally, the Xiaomi 11, the previous Xiaomi 11 has been returned, WIFI is very weak, obviously weak, especially weak, applied for return, I subjectively think that the WIFI signal of the Su Pi is better, and it is much better. Before connecting Xiaomi AX6000 speed test also did not have 200 megabytes of speed.
Speed measurement results are as follows.
Xiaomi 11 supports 160MHz, so there is a clear advantage in the rate, and the speed test results are naturally better than Xiaomi 10, but only by 20%, perhaps Xiaomi 11, the phone’s WIFI still needs to be optimized.
On the XDR5430 V1 and V2 which signal is good, but also two or three days to do some preliminary real-world testing to clear.
Already heat is definitely greater than V1, because IPQ5018 heat than BCM6750 larger, and RTL8763S heat is also not small. The main heat is concentrated in the bottom, the old way to find a short bracket pad up to improve heat dissipation. But I do not think it is necessary, spring is here, conducive to dampness, to prevent the router “long hair”.
The V2 is not a scaled-down version of the V1, but a change from the Broadcom platform to the Qualcomm platform, with greater band capacity and MU-MIMO friendly. But the signal is likely to be weaker than the V1.
Which is better? Do you need the CPU to work better or the wireless to work better?
CPU performance is mostly reflected in NAT performance and high load performance, a test that I hope to have a way to make soon, and am preparing for.
Wireless performance can be tested in the next two days.
IPQ5018+QCN6024+RTL8367S = $65, it’s worth it!