Two Huawei Optical modems were disassembled in a row, and the wireless configuration was the same: Hi1152 plus 5G FEM chip RTC7676, looked at the pins, and found the corresponding two FEM chips, SKY85755-11 and QPF4506B, respectively.
Zoom in to see these 3x3mm sized FEM chips.
Perhaps you will ask, why is such a package? Interested can go to Google search SMT SMD, SMT process, will see the whole volume of the whole volume of chips.
Then zoom in on the front and back of the chip pins.
The QPF4506B and QPF4506 are the same parameters, with the B ending being the version optimized for Broadcom. The power is the same.
POUT = +15 dBm MCS11 HE160 -43 dB Dynamic EVM
POUT = +16 dBm MCS11 HE80 -43 dB Dynamic EVM
POUT = +18 dBm MCS11 HE80 -40 dB Dynamic EVM
POUT = +20 dBm MCS9 HT80 -35 dB Dynamic EVM
POUT = +21.5 dBm MCS7 HT80 -30 dB Dynamic EVM
POUT = +24 dBm MCS0 HT20 Spectral Mask Compliance
QPF4506 pin definition.
The other two chips did not find a diagram like the one above.
Then the power parameters of SKY85755-11.
MCS11, +18 dBm, –43 dB EVM
MCS9, +20.5 dBm, –35 dB EVM
MCS7, +22.5 dBm, –30 dB EVM
QPF4506B and SKY85755-11 both amplification power in 802.11ax mode is very similar, I first pick SKY85755-11 to replace on.
Things are bought, just do it directly, do not have to think before and after, since the Huawei HS8145x6 tossed bad radio frequency circuit, if it can still be turned on, the worst result is used as Optical modem ah (EPON).
If I had known that I could solder on the chip, I would have paid to get the phone repaired to replace it. Really, a waste of time and trouble.
Take apart the motherboard, do a good job of protecting.
Heat gun to 350 degrees blowing, holding the tweezers, blowing for 10 seconds, the surface is blown into this way or can not hold: the
Try again, tweezers a force, like the feeling of clamping the broken, the chip is considered crippled not? The legendary Huawei high-temperature lead-free tin?
Tuned to 450 degrees maximum wind speed, change the small one wind nozzle, the wind continues to blow, finally blowing down the RTC7676. (The following figure pad empty lower left corner of a small white dot, corresponding to the chip on the corner of a small dot, is the No. 1 foot)
When you master the skills, do the same thing will be very efficient, the second one did not take 10 seconds to take off. Take down with 10 seconds, soldering up will take 100 minutes, I’m not exaggerating 。。。。
Take out the SKY885755-11, first on the pins a little tin:.
Main board empty solder plate bit on the top of a little solder oil, I put a lot of oil, slippery feeling. Align the position of the No. 1 foot, open the hot air gun blowing.
I repeatedly blew and pressed and pushed numerous times, but it was still crooked.
After countless more bruises.
Turned on after making another one as usual. Huh, why only 2.4G signal, 5G is broken? No 5G WiFi ah….
This is to support 160MHz bandwidth wifi6, 5G signal to slow a while to appear. Please forgive me for being excited about my first chip change. After a while, finally out of the 5G signal, but, weak can not be weak it: -83dBm
Simply is not working wireless state, may not be soldered, some important pins are not connected to the false solder. Start again, disassemble the machine, take it down and re-tin it, and then put it back to blow, and then after 100 minutes: the appearance of.
I look very satisfied, you are satisfied with it? I do not say, you still think it is the effect of SMT patch is not.
This time not installed back into the shell, directly take the antenna removed before connected to the boot, plus a simple heat dissipation, directly boot up: the
Turn on the machine, suppress the excitement inside slowly waiting for the 5G WiFi signal to appear, just like waiting for the girl of your choice to appear when you go to school (slap, wake up)
mate20 phone connected to 5G wifi, showing a strong signal, but the rate is not 160MHz 1733Mbps, a little uneasy.
Proof that SKY85755-11 is available on the Hi1152, able to make the wireless shot, and can suck the wireless back. As for what kind of working condition can be achieved, it is another matter.
Before replacing the FEM chip, I have done a simple test of Huawei HS8145x6 in order to do a comparison in the unlikely event of a successful replacement and in the unlikely event that it works as usual. No comparison, how can it work. No comparison, how do you know your weaknesses, and then go to correct, a little progress.
I have an idea, if you change the wireless power becomes stronger, and according to the law of conservation of energy, power consumption will also increase, so it, I want to do before changing the chip three tests.
Power consumption of the whole machine.
wireless signal test at the location of point C.
Wireless speed test at point C.
Comparison of power consumption of the whole machine
USB ammeter connected to the front of the DC power supply of the wireless router, direct measurement of the DC port, to reduce the error caused by the switching power supply in different power output and have different conversion rates. Use IxChariot software to beat the current of AX200 receiving terminal for 2 minutes.
Before changing the FEM chip, the maximum power at 5G transmission was 9.38W, as follows.
Switching to SKY85755-11, the maximum power at 5G transmission is 7.04W, as follows.
More than 2W lower power than before the change.
Wireless signal test at point C location
Contrast signal as follows.
Obviously weaker than before the change ah.
Wireless speed test at point C location
Use self-built speedtest speed test server for speed testing, the network card is AX200, continuous speed test 6 times, the last one is the average, before and after the change compared to the following table: Mbps
Before and after the change, the downstream and upstream are twice as slow 。。。。。。。
Shut down, blow down SKY85755-11, ready to weld on QPF4506B, but found that there is a pin deformation, it’s okay, electrochromic iron correction on the good.
Repeat the 100-minute process to connect the two QPF4506Bs to the Huawei HS8145x6.
Turn on, the signal is very weak, as weak as the first time, turn off, take off, tin, solder again, turn on, or the same weak, try again, or weak. Forget it. Give up .
Remove the QPF4506B, not soldered, empty, off the wireless as a light cat to use. Disassemble the amplifier chip, can also be turned on, but the wireless signal is very, very weak, weak to the point of not being used. Since the wireless is not used, I pulled out the two external antennas of HS8145x6, the body is well placed ah.
Connect to the home fiber, turn on the computer, open the Optical modem interface, register the LOID online, and automatically send the configuration.
When finished, go to the Optical modem interface at
Optical power -19.75dBm, not bad.
I just want an excuse to replace the old PT923 Optical modem.
Huawei HS8145x6 in bridged mode after turning off wireless, standby power of about 4.4W.
Although the FEM chip pin and size of the same, but the pin definition may also be a little different, even if these are the same, the focus is on the circuit design requirements around the chip is different, some FEM need to have drive to work properly, PA circuit also need to do impedance matching, need to make factory power more accurate. In short, the wireless RF circuit is not as simple as imagined. mate20 even in a strong signal is not 1733Mbps rate, AX200 just connected to the 2.4Gbps rate, but a second later it dropped to 1.2Gbps or even more than 900Mbps, the signal and speed measurement is also very poor, which can prove that the previously mentioned said equal to the truth.