I saw a K6 on Xianyu that was marked as "water damage" and priced at 40 yuan, which is half the price of a used one.
However, considering that it is a scrapped machine and has been exposed to mud and water, I am still observing the situation.
However, after selling my QuanSheng machine, I hadn't found a suitable and affordable replacement for quite some time. So, I started considering buying this K6 repair unit.
I needed to update the firmware, so I compared the markings on the photos with those in a video on Bilibili. The layout of the markings was similar to that in the video, and it differed from the markings on the V3 version.
After comparing, I decisively purchased this machine. After communicating with the seller, I bought it for 20 yuan.
Images after receiving the goods:
From a visual inspection, water has already entered the internal components of the machine casing, and the situation appears to be very serious.
Preprocessing
For this type of pump, especially one used for pumping mud, I would first need to disinfect the entire machine. Mud often contains pathogens, and operating it can easily lead to infection.
The fabric has been completely saturated with mud, so it needs to be cut away. The remaining portion is then disassembled and cleaned of the shallower layers of mud.
Images that have not been cleaned after disassembly.
Based on analysis of erosion and accumulation, the sources of mud ingress to the lower layers are identified as three locations: bottom clip, Type-C penetration, and battery contact point inflow.
Mud-to-water ratio (Type-C) > Bottom clip > Battery contact points
Based on the mud flushing status, it can be inferred that when water enters, the battery installation is tight, but the Type-C charging port is not fully closed.
The top layer also has Type-C ports, and the antenna interfaces and contributions are minimal. The top layer is almost entirely composed of mud blocks, indicating that the top layer's mud supply is primarily supplemented by infiltration from the lower layers.
The speaker is not leaking, and the keyboard is slightly dirty; this is not due to water entering.
Intake water position analysis: The mud particles were observed to form a stepped pattern, and the upper screws were most severely corroded. It is inferred that the machine was positioned with its head pointing downwards and the top shell facing upwards, resulting in water ingress.
The disinfection process was not recorded. Disinfectant formula: hydrogen peroxide, mixed in a 1:100 ratio. The liquid completely submerged all components and soaked for 40 minutes.
All accessories, except for the screen itself, have been disinfected.
High-temperature baking: Use a hot air gun to bake at 120 degrees Celsius for 1 hour, removing all moisture and also achieving sterilization.
The heat sink, circuit board, and other metal components are involved in the baking process. Plastic parts naturally dry.
All equipment used in the pre-disinfection procedures should be cleaned and then sterilized together by baking.
Screen processing
The screen includes the following components: white paper pad, frame, light guide plate, diffusion paper, liquid crystal display (including FPC, LED, and polarizing film).
Due to the glass encapsulation of the liquid crystal panel, water ingress typically does not affect the liquid crystal panel itself. The primary work involved in screen repair is cleaning.
Completely disassemble the LCD screen. Note that the base plate where the light source is located has been bent; be careful when disassembling it. The LCD theme consists of black adhesive film pasted onto the frame; when separating, be careful not to tear or wrinkle the adhesive film.
Use alcohol to clean all components of the LCD screen, then quickly dry them. Try to keep all components dry and free from dust. Assemble as soon as possible after cleaning.
Incomplete drying: Black spots appear on the screen (which may disappear), and condensation appears between the liquid crystal main body and the diffuser paper.
Dust present: Black spots appear on the screen and cannot disappear on their own. The dust is located between the liquid crystal panel and the diffuser paper, or possibly on the light guide plate.
Circuit board inspection and repair
After thoroughly cleaning away the mud, I examined the circuit board to determine the extent of the damage.
First, observe the positive electrode network VBATT. It can be seen that the MLCC on the positive electrode is severely corroded and has exposed its internal electrodes.
Remove the capacitor, and when cleaning the pads, you find that the pad's plating layer is completely corroded, with only impurities attached. After removing the plating, the copper layer is exposed.
All of the pins on the magnetic bead are corroded and broken off.
Observe the GND (negative) soldering point next to it, which has loose particles on the surface that can be removed. This indicates that when water enters, a potential difference is created by the battery, forming an electrolytic cell, resulting in significant anodic oxidation corrosion. (For specific details, refer to Chapter 1 of High School Chemistry - Selective Required Reading, specifically the section on electrolytic cells.)
Replace the MLCC capacitor and add a ferrite bead.
The three current-limiting resistors on the VBATT side were also severely corroded, and the pins at the VBATT network location were completely corroded away, similar to the MLCC situation.
Plan to replace the resistor.
The operational amplifier's 6-pin, with corrosion and broken connections, along with significant oxidation and corrosion around it. A large number of resistors and capacitors fell off during cleaning.
After testing, these components were found to have inaccurate resistance values and failed soldering tests, so they were decided to be replaced entirely.
According to the schematic diagram, the 6-pin outputs are connected to the VBATTR network via a 150KΩ resistor, and the 5-pin is connected to the VBATT network via a resistor, with a voltage of 7.2V.
The output of pin 7 is connected to pin 2. It is suspected that the voltage on pin 7 is too high, creating a potential difference with pin 3, which has caused corrosion of the components shown in the diagram.
Replacing the chip, I scraped off the solder pads and, after inspection, found that the PCB had corrosion-related breaks. This was due to the effects of anodic oxidation.
The corrosion location is located between the eighth-pin resistor network and the fifth-pin network, as shown in the diagram.
Due to the presence of gaps in the network, a "fly-line" technique was used.
During the flight, we avoided using the positions that would have been occupied by modules like the 4732 once they were installed in the future.
Install the screen and verify that it powers on correctly after connecting to the network. Upon discovering that power output is not functioning normally, a major overhaul of the radio frequency (RF) system will be necessary.
During testing, there was even an instance where the signal strength exceeded the carrier level. Fortunately, the sensitivity was not a problem.
Furthermore, the frequency cannot be adjusted properly; after entering a value, it directly switches to a nearby default frequency such as 137 MHz or 450 MHz, making normal use impossible.
(This is actually a minor solder joint failure, which will eventually expose more serious problems.)
During repair, a blank screen was also discovered. Inspection revealed that a resistor had corroded and broken off its connection. Replacing the 33Ω 0402 resistor temporarily restored normal operation.
On the second day, while continuing repairs, I noticed that the screen was displaying a "snow" effect when booting up, and the system could not enter (after waiting, pressing the PTT red button did not light up, the keyboard lights flashed briefly before turning off).
Upon careful examination of the motherboard, I observed that the EPROM pins were blackened and appeared to have been desoldered. Additionally, the pull-up resistor next to them showed signs of abnormality, with the pin appearing broken.
After removing the EEPROM and resistor, I found that the pads were severely oxidized and corroded. After cleaning off the debris and re-tinning, the solder joints looked clean and shiny.
However, the solder pads for the pull-up resistor had already been broken and could not be repaired. Therefore, a 0603 10KΩ resistor was connected across GND.
Subsequently, I observed that the clock line on the screen was disconnected at the MCU pads.
After reconnecting with the cable, it should return to normal operation.
The test was performed; the frequency could be entered correctly, and the screen displayed normally.
The problem arises when the signal line has a short circuit or a broken connection. The key factor is the voltage difference created by the high-level signal, as well as the small current provided by the GPIO driver.
It is hypothesized that when water enters the machine, the signal lines with a high duty cycle and high voltage levels are more susceptible to electrolytic corrosion due to the longer time they are exposed to the influence of electrolysis.
The clock line utilization rate is 50%, while the I2C bus utilization rate, which is being pulled high, is also high in areas prone to corrosion.
The hypothesized mechanism for clock line corrosion is as follows: The MCU connects to the screen CLK pin via two terminal resistors. The screen interface has a floating/high-impedance input, resulting in a significant voltage difference at the end. When the terminal resistor near the screen corrodes and breaks, the resistance on the MCU side and the pin begin to corrode. After the battery is depleted, a conductive path forms that appears and disappears intermittently.
Next, we will proceed with the RF (radio frequency) section.
Connection test complete, transmit power remained essentially unchanged. Extending the transmission test time revealed overheating in the primary amplifier. It is hypothesized that the final amplifier has failed.
Considering the broken resistor, I decided to replace the entire amplifier tube (model HTL7G06S011P, available in the product library) at once. At the same time, I removed the resistor and directly short-circuited VBATT and VBATTR networks.
After the replacement, the VHF power was restored to normal (40 dBm = 10 W).
However, the UHF power output is 27 dBm.（0.5W）It was significantly lower than expected.
I soldered all of the circuits at every level, including those for the BK4819, and also soldered the parts of the final filter that are needed for operation.
Using comprehensive testing, power decreased instead of increasing, and VHF had no impact.
I conducted a comprehensive investigation of the UHF radio frequency current paths by comparing them to the schematic diagram. Ultimately, I found a flaw in one of the matching capacitors on the port.
One of the pins on a series-connected capacitor (470 pF capacitance) was found to be dislodged. The multimeter showed no response when testing the pin, indicating that it was open or shorted. Therefore, the capacitor was removed.
After removing the capacitor, I found that the solder pads were severely corroded and oxidized. Cleaning off the debris allows for re-soldering.
The old component pins were severely blackened and unresponsive to soldering, making them irreparable. Therefore, a new capacitor was installed.
The new image
The overall test measured power output, and UHF and VHF power levels met the standards. After tightening the enclosure, spurious emissions also met the standards.
Analysis of RF circuit failure:
The matching capacitor for the RF port is located near the port and has a direct connection to the U8C_V+ network, with a voltage of approximately 7.2V.
The U8C_V+ functions include: providing power to the operational amplifier; and ensuring that both diodes are reverse-biased during transmission, limiting the RF current path.
The voltage is generated by the circuit shown in the diagram. When transmitting, it is controlled by BK4819\_GPIO1 to enable and generate the voltage. The power source comes from the VBATT\_SW network, with a voltage of 7.2V.
Therefore, when the enabling circuit is activated, U8C_V+ creates a voltage difference that is comparable to the power supply rail of 7.2V.
It is hypothesized that an alarm or sound-activated transmitter was activated before water ingress.（VOX）Function: Initiates release underwater, causing corrosion.
After corrosion, parasitic inductance increased. This portion of parasitic parameters has little impact on VHF signals, but significantly attenuates UHF radio frequency currents.
The entire machine repair was successfully completed at this point.
Repair costs: amplifier tubes (3.5 yuan), resistors and capacitors (three types, 100 of each, 2 yuan), jumper wires, and spare 10uF and 100nF capacitors.
If you are planning to waterproof the K6, you can refer to the failure analysis section.
闽ICP备2021006864号-7
闽公网安备35020602002794号