Maintenance is crucial for keeping your mining equipment, such as the Antminer L3+, running smoothly and ensuring optimal performance. Here is a summary of the preparation and maintenance guidelines, the tools and spare parts required for repair, and the maintenance requirements for the Antminer L3+.
Table of Contents
Preparation and Maintenance Guidelines
Prior to, during, and after installation, it is crucial to properly prepare and maintain the components. This entails using thermal gel to improve heat transfer, creating air ducts to improve airflow, connecting power supplies in the proper order, fixing chips to prevent overheating, and making sure test fixtures meet production standards. These guidelines must also specify how to safely store components away from extreme temperatures and humidity levels, as well as how to clean components with approved solvents like isopropyl alcohol or distilled water. To ensure the proper operation of all system components, routine maintenance inspections should be performed every few months or at least once a year.
Preparation Requirements for Repair Platform, Tools, and Equipment
I. Platform Requirements
- The platform needs to have a workbench for fixing Antminer L3+ hash boards, which must be properly grounded. In order to avoid damaging the materials being worked on with static electricity, an anti-static wrist strap and grounding are also necessary.
II. Equipment Requirements
- soldering iron with a pointed tip and a temperature range of 350 to 450 degrees Celsius for small patches like r-c.
- Heat gun for soldering and chip disassembly; avoid prolonged heating to avoid PCB blistering.
- Hash board testing will be done using an APW3 power source with a 12V and 133A maximum output.
- L3+ hash board tester, multimeter, tweezers, and preferred oscilloscope.
- Cleansing water is used to remove residue and the appearance following maintenance, along with scaling powder, cleaning water, and anhydrous alcohol.
- Tin chips are implanted when renewals occur, along with a tin grinder, tin stencils, and tin cream.
- use black Heat-Conducting Glue (3461), after maintenance, to glue the cooling fin.
III. Test Tool Requirements
- ARC Antminer Hashboard Tester
- Lab PSU 10-30V / 1-15A
- APW9+ power supply and power patch cord for hash board power supply
- Utilize the V2.3 control board test fixture (test fixture material number ZJ0001000001).
IV. Maintenance Auxiliary Materials/Tools Requirements
- Anhydrous alcohol, flux, 138°C solder paste, and Mechanic lead-free circuit board cleaner.
- After maintenance, the flux residue is cleaned up with mechanic lead-free circuit board cleaner.
- After repair, thermally conductive gel is applied to the chip surface.
- Desoldering wick, solder balls, and steel mesh (a 0.4mm ball diameter is advised).
- The chip pins must first be tinned before being soldered to the hash board when a new chip is being installed. After evenly applying thermally conductive gel to the chip’s surface, lock the heatsink.
- code scanner for the serial port.
- RS232 to TTL serial port adapter board, 3.3V.
- Self-made short-circuit probe (use pins for wiring and welding; heat shrinkable sleeve to prevent short-circuit between probe and small heatsink).
V. Common Maintenance Spare Material Requirements
- 0402 resistor (0R, 10K, 4.7K,)
- 0402 capacitor (0.1uF, 1uF)
- Maintenance staff should be well-versed in electronics, have at least one year of experience, and be highly skilled in QFN encapsulation and soldering procedures.
- Check the maintenance work at least twice, and make sure the outcome is satisfactory each time.
- Pay close attention to the techniques used during maintenance, make sure there is no obvious PCB deformation after changing any fittings, and inspect parts for missing/open circuits and short circuits.
- The hash board tester, corresponding test software parameters, and the maintenance target should all be checked.
- Verify the testers and tools.
Overview of Antminer L3+ Components
L3+ Hashboard Structure
- Each of the L3+’s 12 series-connected voltage domains has six BM1485 chips, making a total of 72 BM1485 chips on the board.
- The BM1485 chip has voltage-reduction diodes built in, which are identified by the chip’s designated pin.
- From the first chip to the last chip, the L3+’s 25M monocrystal oscillator forms the clock.
- On the back of each chip in the L3+ are independent cooling fins. After preliminary testing, heat conducting glue is used to permanently attach the SMT paste on the front and the back to the IC. In order to fix them, apply black heat-conducting glue evenly across the back of the IC after each maintenance.
Signal Direction of L3+ Chip
The No. 1 transmits the CLK signal flow, which is generated by a Y1 25M crystal oscillator. 1 chip to the The voltages for the No. 72 chip are 0.9V in both the standby and computing modes. The IO mouth pin 11 is used to transmit the TX (CI, CO) signal flow from the No. 1 to the In the No. 72 chip, the voltage is 0V when the IO wire is unplugged and 1.8V when computation is taking place. The No. 1 transmits the RX (RI, RO) signal flow. 72 to the After leaving the IO mouth pin 12 and heading back to the control panel, No. 1 chip. Both when computing and without an IO signal plugged in, the voltage is 1.8V. The No.’s electrical level is lowered by the B (BI, BO) signal flow. 1 to the When the IO wire is unplugged or when the chip is in standby mode, the voltage is 0V, and the signal impulse is approximately 0.3 when computing. Through IO mouth pin 15, the RST signal flow transmits from the No. 1 to the When the IO signal is unplugged or when the chip is in standby mode, the voltage is 0V; however, when computing is taking place, the voltage is 1.8V.
The RST, BO, RI (RX), CO (TX), and CLK signal testing points, among other chips, are located on the front of the hash board. During maintenance, these testing points aid in fault location. Each of the 12 voltage domains on the board has six chips. Six chips connected to the same voltage domain’s power supply in parallel then connect to other voltage domains in series. The CORE voltage, LDO-1.8V, PLL-0.9V, DC-DC output, and booster voltage 14V are just a few of the testing points for the voltage domains. These test points must be tested during maintenance, and the fault points can be deduced from the circuits before and after the test point.
Testing Points Among Chips
The most direct way to find faults is to test the testing points among the chips during maintenance. RST, BO, RI (RX), CO (TX), and CLK signal critical circuits are positioned on the front of the L3+ hash board, as shown in Fig 3.
The entire board has 12 voltage domains, and as shown in Fig. 4, each domain has 6 chips that are powered in parallel before connecting to other voltage domains in series. The pin functions of the BM1485 chip are taken into account in the voltage domain single chip principle analysis, as shown in Figs. 5 and 6.
With five on each side for CLK, CO, RI, BO, and RST, ten testing points on the chip’s front and back are primarily tested during maintenance. CORE voltage, LDO-1.8V, PLL-0.9V, DC-DC output, and booster voltage 14V are additional testing points.
The DC-DC output is approximately 10V and the booster voltage output is approximately 14V during maintenance if the IO wire is unplugged and only 12V is plugged in. Among the test points, CLK must be 0.9V, RI must be 1.8V, and the voltage of all other points must be 0V.
The DC-DC and booster voltage have no voltage output when the IO wire is plugged in but the test key is not depressed. But as soon as the tool test key is depressed, the PIC kicks into action, and the DC-DC outputs the voltage that the tool test program has set. Once the tool outputs WORK and the hash board has finished computing, the booster voltage starts to function. At this point, the normal voltage of each testing point should be:
The transient voltage is approximately 1.5V when the tool only sends WORK, which has negative polarity and lowers the DC level. CO: 1.6-1.8V.
A hash board anomaly or a zero hash rate during computation will be brought on by abnormal voltage or low voltage. RI: 1.6–1.8V.
BO: 0V in the absence of computation; 0.1–0.3V impulse beat when computation is taking place.
RST: 1.8V. The output reset signal is sent once every time the tool test key is pressed.
The fault point can be deduced based on the circuits before and after the testing point if the voltage or status of any testing point is abnormal.
The pin functions of each signal, including CLK signal, TX signal, RX signal, BO signal, and RST signal, can be seen from the list above. Each signal voltage, CORE voltage, LDO-1.8OV, PLL-0.9V, etc., is measured during chip testing., should be tested. A chip CORE short circuit of this voltage domain is typically indicated by a CORE voltage of 0.8V, whereas an LDO-1.8O short circuit or open circuit of this chip is typically indicated by an LDO-1.8O voltage of 1.8V. A short circuit of the PLL-09V power supply of a chip in this voltage range is indicated by a PLL-0.9 voltage of 0.8V.
Last but not least, using the tool’s print window information, it is possible to determine the hash board’s operational state, the chip’s hash rate, and the tool’s perception of heat.
Definition of IO Pins
A 2X9 pitch 2.0 PHSD 90°dual inline package makes up the hashboard’s IO plug. The pin functions are defined as follows:
- Pin 1, 2, 9, 10, 13, and 14: GND
- The DC-DC PIC’s pins 3 and 4 (SDA, SCL) are used to connect the control panel to the I2C bus, which is how the PIC and control panel communicate. As a result, the control panel can read and write PIC data and thus regulate the hash board’s running state.
- A hash board identification signal is sent on pin 5 (PLUG0). This pin turns from being low level to high level when the IO signal is connected, raising a 10K resistance to 3.3 V.
- Signal for the PIC address pins 6, 7, and 8 (A2, A1, A0).
- The 3.3V end of the hash board’s hash rate channel is represented by pins 11 and 12 (TXD, RXD), which are converted into the TX (CO) and RX (RI) signals through resistive voltage division. All IO mouth pin ends have an electrical level of 3.3 volts, which resistive voltage division changes to 1.8 volts.
- Reset signal on pin 15 (RST) is 3.3V end-to-1.8V RST reset signal via resistive voltage division.
- Pin 16 (D3V3): The 3.3V power supply for the hash board, which receives power from the control panel and primarily provides working voltage to the PIC.
To raise the voltage from 10–10.4V to 14V, the DC–DC booster is needed. The switching power supply, specifically model number U111 RT8537, converts the 10V signal to 14V to accomplish this. L1 is an inductance component that stores the switching signal created by U111. A 14V output is then produced at the positive electrode of capacitor C1072 as a result of the boosted rectifying diode D100 charging and discharging it.
It’s important to note that any voltage anomalies in the booster circuit can easily damage the chips as well as the LDO (low-dropout) regulators of the last 4 voltage domains of the hash board. The oxidation of U111, R996, and R997 frequently results in such anomalies.
You must inspect the C948 capacitor’s two ends in order to measure the DC-DC output voltage.
Checking the consistency of the PIC voltage value and the DC-DC output voltage by looking at the tool print information is the first thing to do if the DC-DC output voltage is abnormal. Replace the low capacitance surrounding the LM27402SQ if they are inconsistent.
If there is no output from the DC-DC, you should check to see if the PIC is functioning normally, if the PIC is properly receiving the 12C signal from the control panel, and if the EN voltage of R13 and R14 is close to 1V and R11 is 12V.
Pins 1 and 3 are used as inputs and Pin 5 serves as the chip’s 1.8V output.
By voltage dividing two resistors from the LDO-1.8V, the voltage for the PLL-0.9V is produced.
Temperature sensor circuit
The BM1485’s Pins 6 and 7 are linked to the temperature sensor chip. The temperature information is obtained from the BM1485’s built-in temperature sensor and then sent through Pins 15 and 16. The RI signal is used to transmit the data back to the FPGA in the control panel.
Troubleshooting Common Miner Failures
One Or More Hashing Boards Not Being Detected
- Verify that the signal cables are securely and properly connected. Replace them with new ones if they are loose or damaged.
- Verify that the power supply is functioning properly and delivering adequate voltage and current. Replace it with a compatible one if it is defective or insufficient.
- Verify that the hashing boards’ chips are in tact and not burned or cracked. If they are broken, you might have to replace the entire hashing board or use a soldering iron and heat gun to fix it.
- Verify that the miner’s firmware is current and suitable for your hashing boards. You can attempt to upgrade it using an SD card or a web interface if it is corrupted or out of date.
Low Hash Rate
- Make sure the miner’s temperature is between the accepted range of 40 and 80 degrees. If it is too high, you might need to improve your miner’s cooling system by removing dust, changing the fans, using thermal paste, etc.
- Verify that your miner’s frequency is set correctly for your power source and environment. If it is too low, you can try gradually raising it until you find an effective and stable setting.
- Verify that your miner’s fans are functioning properly and spinning at a fast enough rate. Replace them with new ones if they are broken or slow.
- Verify the consistency and dependability of your pool connection. You might need to adjust your pool’s settings or choose a different pool if it’s unstable or frequently interrupted.
- Verify the security and stability of your network connection. You might need to adjust your network settings or use a firewall if it is being interfered with or attacked.
- Use a soft brush or a compressed air canister to remove the dust from the air vents and the fan blades. You might have to disassemble the fan and give it a thorough cleaning if the dust is too heavy or sticky.
- Check to see if the fan bearings are worn out or damaged. Put new ones in their place if they’re clunky or stuck.
- The fan wires should be connected securely and correctly, so check that. Reconnect them carefully if they are broken or loose, or buy new ones.
- Verify that the miner’s firmware is current and suitable for your fans. You can attempt to upgrade it using a web interface or an SD card if it is corrupted or out of date.
Imbalanced Impedance Among Multiple Voltage Domains
To detect open or short circuits, check the impedance of each voltage domain. Voltage that is abnormally high or abnormally low in a domain denotes an anomaly in the domain or nearby domains’ IO signal. To find the anomaly point, check voltages and signals at test points. In order to identify the source of any voltage imbalance, check the CLK and RST signals.
When performing test box checks, if only some of the 72 chips can be located, identify the anomaly chip by testing. To find the anomaly chip, rotate the TX signal of a specific chip over the terrain. Continue looking until you find it.
This signal chain will be disrupted if one chip is operational but does not transmit data from other chips. The amount of chips found is indicated by the test box. At test points before and after chip number, check voltage and impedance. 14 to locate the problem.
The test box cannot detect the chip information of the hash board and shows “No hash board”. To find the issue, look for anomalies in the voltage, chip, and Buck and Booster circuits.
Inadequate Nonce and low hashing are two causes of low hashing. Examine cooling problems and chip malfunction. Additionally, use the test box, particularly with the DC adjustable 12V power supply, to tune down and check the voltage of a specific chip.
NG of a Certain Chip
The port information indicates that a particular chip’s return’s nonce is either zero or insufficient. The chip should be replaced after checking for subpar soldering or peripheral issues. Check the target board frequently to look for burn, deformation, or displacement of the cooling fins. Check each voltage domain’s impedance and whether it reaches 0.8V with a voltage difference of no more than 0.05V. Locate the anomaly point after performing the chip’s routine inspection with test boxes. Resolder the chip, and execute two test runs on a fixed hash board using a test box. Following maintenance, record the malfunction type, and then follow up with a formal burn-in.
Cleaning and Maintenance
Regular cleaning and maintenance are necessary to keep your Antminer L3+ operating smoothly and effectively. Here are some pointers for carrying this out correctly.
Dust buildup on your miner’s fan blades, air vents, and chips can impair cooling and result in overheating. The front and back fans can be gently blown twice with compressed air or with a soft brush to remove dust. To avoid causing harm, take care not to touch the grille or the fan blades. Another option is to use an air gun to fill every gap for two to three seconds, then repeat several times. In accordance with your surroundings, you ought to perform this at least once per month and occasionally more.
Your miner’s fans are in charge of maintaining stability and stability. They may result in high temperature, a slow hash rate, or fan failure if they are defective or slow. You must turn off your miner and unplug the power cord in order to replace the fans. After that, you must remove the screws holding the fan grille in place and unplug the fan wires from the control board. The new fan must then be installed by joining its wires to the control board and fastening it with screws. Use compatible fans that meet the 12V 2.3A requirements.
You should clean and maintain your miner regularly to ensure its optimal performance and longevity
Firmware and Software Upgrade
Regular firmware and software upgrades are required to keep your Antminer L3+ secure and up to date. Here are some safe ways to go about it.
Using An SD Card
For this method, you must copy the firmware file to an SD card after downloading it from Bitmain’s website5. The power cord must then be unplugged and your miner must be turned off. The SD card must now be inserted into the miner’s control board slot. The miner must then be turned on and the power cord plugged in. Automatically beginning and lasting for about ten minutes, the upgrade process.
Using a Web Interface
For this method to work, you must download the firmware file from the Bitmain website5 and enter your miner’s IP address in your browser to access the web interface. You must then enter your username and password (the default is root) to log in. The firmware file must then be chosen from your computer by clicking Upgrade after that. You must then click the Flash image and wait for the upgrade to be completed.
You should always backup your settings before upgrading your firmware or software
Other Considerations and Maintenance Flow Chart
- Check the target board frequently to spot any burns, deformations, or displacements of the cooling fins. In each voltage domain, look for any shorts or open circuits as well. Ensure that the voltage in each domain is 0.8V and that the voltage difference is 0.05V or less. Anomalies in the voltage in nearby domains could indicate an excessively high or low voltage.
- After performing a routine inspection, test the chip with a test box to identify the problem and proceed from there.
- Check the test points on the problematic chip, such as VDD, VDD0V8, VDD1V8, and CLK IN OUT/TX IN OUT/RX IN OUT/B IN OUT/RST IN OUT.
- According to the signal flow, all other signals—aside from RX—transmit forward (No. 1 to No. 72) and in reverse (No. 72 to No. 1), respectively.) to identify the anomaly with power sequence.
- Re-solder the chip once the problematic one has been identified. Scaling powder should be added all around the chip, the chip pin should be heated until it is dissolved, and then the chip should be moved and lightly pressed. Re-grind the soldering pans and chip pins next.
- Run the test box two times on the fixed hash board. After switching out fittings, the board should be slightly cooled for the first time, and then a few minutes later, the board should be completely cooled. The time between two tests won’t have an impact on performance. Set aside the repaired board and work on another one before returning to the first one with the second one that has been fixed.
- After maintenance, record the type of malfunction, including the model, location, and cause. Input from production, CS, and R&D will be improved as a result.
- After logging, perform a formal burn-in.