Repair motherboard skills

1. A 810E motherboard is powered on automatically, and automatically shuts down after a few seconds. OK after changing the IO chip and capacitor.




　　2. CU-810 motherboard can be on CI, can not be on CII and can not turn off the machine, sometimes can not start. After changing the IO, the problem remains. Check the PW-ON line and find that the high and low levels of the circuit are not correct. After the 7407, the test board is OK.




　　3. MSI motherboard, North Bridge 693A, South Bridge 596A, boot test card does not go. On the dummy load, the CPU power supply triode voltage is measured to be 2.5V. The clock and reset on the CPU socket are normal. Replace the triode with OK.




　　4.P4 motherboard GIGA845IRA-533 DDR memory, test card running CI, but the BIOS is still sound. The memory slot is incorrect, and the voltage of the triode on both sides is incorrect. Change it to OK right now.




　　5. An elite 815EP-T motherboard, CMOS data can not be saved, battery voltage is normal. The voltage on the jumper cap is very low. Put the battery and the jumper cap, and the next few capacitors are turned off and OK. (Note: IO and diodes next to the battery may also cause such faults.)




　　6.810 motherboard, boot test card does not go away. Measure the eighth leg on both sides of the power chip, one side without voltage, and the inductor has no voltage output. There is no voltage on the two legs of the triode, and it is OK after changing the power IC.




　　7. A motherboard, at the same time connected to the keyboard and mouse can not be used, the fault remains after changing IO. After the PS/2 port power supply pin is 4. 5V, it should be 5V when it is normal, and the solution after the connected capacity is removed. (Note: This volume is the role of insurance.)




　　8. Keyboard interface can not be used, found that the keyboard interface clock voltage is 5V, the mouse interface clock is 0.3V, the mouse can be used, the problem is not here, exclusion In exchange, this motherboard does not have I/O to ground to measure the power supply and signal line resistance, power supply 300 or so, signal line 600 or so. If there is something wrong, find the connected inductor along the wrong pin to replace it




　　. MS-7021 (KT600) motherboard running FF maintenance case




　　A 7021 board, KT600 North Bridge, 8237 South Bridge, when you receive the hand is running FF, the CPU's main power supply, frequency, reset are normal, the voltage of each test point is also normal, there is no short circuit and open circuit between the North Bridge and the CPU block . It can be said that all the working conditions are satisfied. Generally speaking, this board has nothing wrong with it, the BIOS is bad, the North Bridge is welded, and the 83697 is bad. The South Bridge is broken, but these fault points have been checked. No faults. Then carefully look at this motherboard, finally hard work pays off, in the N minutes of gaze, found in the MS-7's heat sink, found a position number R115 102 resistor was knocked out. After filling this resistor, the board is OK, and then the circuit diagram is checked. It was found that this resistor is connected to the ground by the VID_SEN signal from the MS-7. This line provides the VCCA_PLL signal for the CPU through a 351 small field tube. This signal is not normal, causing the CPU to not work.




　　10.KT600 motherboard crash repair example




　　This board was originally a BIOS data error. After re-FLASHBIOS, it was sent to the test station for testing. The result was called back. The fault was entered into the system and the 3D crashed. Later, I found that not only will I run the 3D crash, but sometimes I will crash when I enter the system. Check that the power supply of the CPU is normal, and the surrounding filter capacitors are not damaged, so the main voltages are measured when the machine is down. The power supply of the memory and the reference voltage are normal. When the VDDQ voltage on the Q40 is measured, the voltage is found to be about 3V, and the normal value is about 1.5V. It is obviously wrong. Replace the 45N02 and measure the voltage after restarting. Normal 1.5V or so, I thought it was normal, but after running into the system, 3D still crashes. At this time, the VDDQ voltage has become about 3V, so the fault point is judged here. At first, it was suspected that the MS-7 that provided the feedback voltage to the Q40 was bad. It was invalid after the replacement. Later, it was found that the voltage was normal when the machine was just started, and when the graphics card was run, the VDDQ voltage suddenly increased to 3V. So I suspect that the fault is on the relevant line of VDDQ, so I look at all the relevant lines of VDDQ. When measuring Q35, I found that the diode value of the 620 tube G pole is 200 times lower than that of the good board, so I check G. Extremely related circuit, found that this is an AGP8XDET_GC# signal, and this signal is connected to the G pole of Q33, suspected that Q33 is damaged, causing Q35 to work abnormally. After replacing Q35, VDDQ voltage returns to normal, and after running the system, 3D is OK!




　　Q33, Q35 is a circuit to judge whether the VGA working state is 4X or 8X. The S35 of the Q35 provides the VREF_CG signal, and the D of the Q35 is the VDDQ voltage. This group of circuits is faulty, causing the AGP8XDET_GC# to be abnormal after detecting the VGA, making VDDQ The voltage rises to 3V, and VGA and NB will certainly work when operating at abnormal voltages.




　　11.MS-7021 (KT600) motherboard running FF maintenance case




　　A 7021 board, KT600 North Bridge, 8237 South Bridge, when you receive the hand is running FF, the CPU's main power supply, frequency, reset are normal, the voltage of each test point is also normal, there is no short circuit and open circuit between the North Bridge and the CPU block . It can be said that all the working conditions are satisfied. Generally speaking, this board has nothing wrong with it, the BIOS is bad, the North Bridge is welded, and the 83697 is bad. The South Bridge is broken, but these fault points have been checked. No faults. Then carefully look at this motherboard, finally hard work pays off, in the N minutes of gaze, found in the MS-7's heat sink, found a position number R115 102 resistor was knocked out. After filling this resistor, the board is OK, and then the circuit diagram is checked. It was found that this resistor is connected to the ground by the VID_SEN signal from the MS-7. This line provides the VCCA_PLL signal for the CPU through a 351 small field tube. This signal is not normal, causing the CPU to not work.




　　12. MSI-6712 motherboard plug VGA no reset maintenance example




　　This motherboard fault is not VGA can be normal running code, but if you plug in the VGA card, it will single-step card run single F, that is, the whole board has no reset. At first, it was thought that the tube with VDDQ voltage was abnormal, and the VDDQ voltage was pulled down after adding the graphics card, which caused the north bridge to fail to work, causing a single F. However, there was no effect after replacing the tube, and the AGP tank was replaced later. Still no effect.Later, I thought about it a bit. Only when I plug in VGA, there will be a phenomenon of running a single F, so the judgment is related to AGP. Later, I queried the circuit diagram and found that the Q30 tube was abnormal and had a slight breakdown. One end of Q30 is 3V, one end is AGPEND, and the other end is the SUSB# signal to South Bridge, MS-5. After this tube is short-circuited, when the AGP graphics card is plugged in, the 3V voltage is directly connected to the AGPEND through the AGP card, causing the SUSB# signal to be incorrect, and the south bridge is not working properly, causing the whole board to be reset without resetting. OK after replacement!




　　13.MS -7144 motherboard does not power on




　　The voltage on the PWSW is measured as normal, 5, 3, 1.5Vsb PSON# is normal, and 32.768KHz is normal. When checking ms-6, it was found that 42pin rsmrst# is low, because this letter is connected to SB, I/O and LAN, and it will not work in reset state at all times in SB, I/O and LAN. Use the disconnection method to disconnect sb and ms6 first. The phenomenon remains the same. There is no change in the I/o after disconnection. When the lan end is disconnected, the signal is normal. After replacing 8201cl, it is OK.




　　14. MS-7144 motherboard automatically power off after the maintenance of the




　　MS-7144 motherboard, K8-754 CPU architecture, the failure phenomenon is plugged into the CPU normal power, automatically power off after 15-20 seconds. The same is true after booting the dummy load. And after the power is off, the short-circuit switch can not be powered, measure all the voltages on the motherboard, and find that the 3VSB voltage is pulled down to about 1.8V, so replace the Q32 that produces 3VSB, the model is 7313, no effect, replace the 3VSB sampling voltage. MS-6, still ineffective, and then replace the South Bridge of the 8237R, still the same as before. At this time, I felt that I couldn't get started. Later, I tried not to add a dummy load and did not add CPU power. I found that there was no automatic power-off phenomenon, so I suspected that the fault was in the CPU power supply, so I did not plug in the small 12V plug power-on test. There is no automatic power-off phenomenon, generally do not plug small 12V normal, indicating that the power supply MOS or power IC is faulty, first replace the power IC, ISL6569, then power on, everything is normal, this board is repaired.




　　15.MS-6797 motherboard repair case




　　The friend brought an MSI barebone motherboard, MS-6797, 865G motherboard, just took the CPU seat air welding, he used a wind gun and tin furnace for one, but the heating is not very successful, there are a lot of air welding The place, after receiving the board, replaced a CPU seat with 935, after the cooling, power up, found that the single-step card runs single F, the whole board is not reset, CPU voltage, clock, PG, RST are not, 14.318 also It doesn't work, what's the matter? Because this board is quasi-system, the board is slender, the north bridge is far away from the CPU base, and the corner of the board is forced to change the signal and voltage of each test point. Normal, so I suspected that the North Bridge was welded, and the BGA machine heated the North Bridge. After that, the phenomenon remained the same. At this time, the motherboard was carefully observed. It was found that the tin between the PIN of the ISL6556 and the pad of the motherboard seemed to be insufficient, so the PIN of 6556 was suspected because After several heatings, the phenomenon of soldering is caused. The solder is re-added with a soldering iron. After power-on, the board is repaired.




　　16.K7 motherboard does not power a small experience




　　Sometimes, when repairing NV and VIA K7 motherboard, all working conditions are normal, the voltage is not short circuit, the crystal oscillator starts, but it is not powered, encounter this In the case, I feel that I can't get started. In fact, I can see if there is an ATTP1 8-pin IC on the motherboard. The function of this original is for system monitoring, which is often said to burn the function. The system detects that the CPU temperature is too high. After that, the IC can automatically pull the PSON# signal high. The motherboard is automatically powered off to avoid damage to the CPU. Because there is a PSON# signal on this IC, if the IC has a bad body, it will cause the motherboard to not be powered.




　　If a friend encounters a K7 motherboard that is normal but does not power up during maintenance, you can see if there is ATTP1. If so, see if the 5V voltage of the sixth leg of ATTP1 is pulled low when the switch is triggered. Action, if you can't pull down, basically can explain that this component is faulty, it will be OK after replacement.




　　17.K8 power-on immediately power off a case




　　Start a K8 motherboard, the fault is after the power-on DEBUG card flash, immediately extinguished, experience inferred, this is generally caused by a short supply of power, or the 3VSB MOS is broken 5VSB for 3VSB. Then the CMOS jumper is inserted. After eliminating the jumper and 3VSB aspects, measuring the ground diode value of 3.3V, 5V and other voltages on ATX, found that the ground value of 3.3V is only 135, which is abnormal, obviously low, so Check the IC with 3.3V input, change the network card, clock chip, sound card, BIOS block, etc., or short circuit, then prepare to kill the South Bridge, but suddenly thought, is there a MOS tube short circuit? So check all with 3.3 V related MOS tube, finally found in the memory slot to the DDR power supply 2.6V two 3055 under the tin has been out, so replace, add load after testing OK, this board repair.




　　18.865PE motherboard power-on without reset, South Bridge overheating maintenance, a




　　friend took a 865PE motherboard with a crown, was replaced by 83627HF-AW, the fault phenomenon is power-on single-step card running single F, that is, all without reset, and South Bridge overheat. Originally thought to be a phenomenon of short circuit in the internal voltage of the South Bridge. However, the measurement of each voltage point and the AD line of the south bridge did not reveal a short circuit. There is also no capacitor short circuit around the south bridge. My experience is that if the south bridge has a voltage short circuit, most of them are short-circuited by standby voltage, 1.5VSB or 3.3VSB. In this case, after the voltage is plugged in, it will not be turned on, and the south bridge should be heated, but this board is not added. In the case of electricity, it does not heat up, so the South Bridge is not necessarily damaged. In the idea of ​​the first voltage, check the power supply of each part after power-on. At this time, there are two 3055 tubes on the left side of the AGP. The lower tube looks at the line, which is the power supply of VDDQ. There is no output at all, and G has a control voltage, so it is inferred that the MOS tube is damaged, so a new 3055LD is replaced, the power is changed after the change, the VDDQ voltage is normal, and the south bridge is not heated, and the board is repaired.




　　19. Soltek's motherboard does not power




　　the computer used by the family. Recently, it often crashes. I looked at it and exploded more than 10 capacitors. I have time to change it this evening. I have not done it for a long time. ^-^ ).




　　The motherboard is the Soltek 85DR-TC. More than 10 capacitors can be fixed in 3 minutes. Try it for safety or empty board first. After plugging in the power supply, the motherboard is not powered! Halo. . . . .




　　The Nanqiao crystal shock has been measured. I/O, W83627HF 68pin PSIN and 67pin PSOUT, when the motherboard PS-ON pin is triggered, there is a voltage change, I / O should be no problem,




　　it is suspected that the power PSON trigger triode has a problem, it took more than 10 minutes to find There is no problem with the measurement, but the base has no voltage. Helpless, looking for a W83627HF replaced, still the same,




　　ready to give up, inadvertently, plugged in an AGP graphics card, you can power up, unplug the graphics card can not! After analysis, agp there is a triode tube not on the graphics card Will pull down the base voltage of the trigger tube,




　　really what TMD is designed to make me engage in one night. (The time for repairing is not much, but there is a lot of time to find things.) If




　　you hit the motherboard of Soltek, you don’t need to power it up. You have to add a video card to try it! Oh!




　　Don’t write it, the night is gone, huh, huh. . . . . . . .




　　Re: Mastertek has a lot of motherboards since 845 to plug in the graphics card to trigger! The landlord repairs the master less than the Tektronix.




　　20. Soltek MVP3 motherboard crashes during the boot process




　　Soltek MVP3 motherboard according to the user reflects the board crash in the WIN98 boot process, generally in front of the WIN98 screen just crashed. During the visual inspection, it was found that the top of the capacitors for the CPU power supply of the board swelled up. It is estimated that these capacitors may cause problems due to the increase of the internal resistance of the power supply. After removing all the damaged capacitors, the board is restored to normal after the power-on test. I have repeatedly found that Solitaire motherboards have such problems, which are "capacitance."




　　21.Aopen AX6BC Pro motherboard is not bright




　　Only the indicator light on the POST card for detection lights up at the moment of power-on. It is estimated that there may be a short circuit somewhere, resulting in power protection. Further inquiring about the user, the user has inadvertently touched somewhere when the fan is installed with power, and a spark appears. In the power check of this motherboard, it was found that the power switch tube FDB7030L and the Schottky diode 1N5817 breakdown were damaged. In the motherboard repair, the motherboard power switch tube is more damaged. Most of these switch tubes are FETs . Their parameters are close, but most of them are SMD (surface mount). Generally, they are not easy to buy in the capital cities like Harbin. To (I saw in the electronics market in Beijing that there are many businesses selling such pipes, envy, envy!). To deal with this type of SMD pipe, I have a "trick" - "no guns, no guns, and made them myself." The method is very simple, can be as follows: using ordinary TO220 package 60N06 and SMD package switch tube comparison, cutting, bending and post-generation. I just did this by becoming my own "SMD" 60N06 and replacing the FDB7030L, which repaired the board in one fell swoop. The 60N06 in the TO22O package is often used in equipment such as UPS, which is easy to buy and not expensive. The picture above shows the “Trilogy” of SMD 60N06.




　　22. Chengqi 6VIA3 motherboard is not bright




　　Visual inspection found that the capacitor near the CPU socket is bursting at the top. After the replacement, the power supply still does not work. Check the power switch tube Q14, Q15 for breakdown and replace. Power up the test machine, still not bright. Continue to check that R144 (2.7Ω) is open. The 5 pin (Vcc) of the power control IC U12 (SC1164) has no 12V, and the R160 (10Ω) connected to it is open. Replace the above components one by one, power on and try again, and R160 burns out again. After checking that there were no abnormalities in other components, I decided that U12 must be broken, because there is no SC1164 on hand and I have to stop working. Occasionally found that I have an unmodified Intel BX motherboard power control IC is SC1185, can they be replaced? I immediately found the information of these two ICs, after a comparison, found that the two are different except the sixth foot Others are nothing different. Suspend the 6th foot of the SC1185, weld it to the original U12 position, and replace the R160 again. While powering up, I pray: May my Lord bless me. The result, the result, the result - as the song sings:




　　Pulled to the hospital to sew 5 stitches - well! 23. V6931 motherboard does not shine




　　, the CPU power supply voltage is not normal, the power supply IC U6 (the above phenomenon is judged) LM2636) is damaged. At that time, because there was no LM2636 at hand, I had to consider replacing it. According to the data, the LM2635 has the same pin definition as the LM2636, except that the voltage range of the LM2635 is 1.8V-3.5V, while the LM2636 voltage range is 1.3V-3.5V, which means that the LM2635 is only suitable for the old Celeron CPU. After asking the user, I learned that the user is using the old Celeron, just LM2635. After obtaining the user's consent, replace it with LM2635, everything is back to normal.




　　24.M6CF motherboard is not bright.




　　After checking the CPU 1.5V power supply unit adjustment tube Q16 is damaged, the power-on test has been restored after the replacement. Can be found when the user picks up and does not light up, this check found that the CPU's 1.5V, 2.5V power supply voltage is abnormally low. The board power IC (U10) uses the LM2636. After comparing the typical application circuit of the LM2636, I found out the 1.5V, 2.5V power supply part of the board. This part of the working principle is as follows: The LM2636's 9-pin (Vref) provides a 1.2V reference voltage to the U11 (LM358, dual op amp) consisting of two regulator adjustment circuits, of which the 2.5V group is an operational amplifier in U11 (that is, the first, second, and third legs) and adjustment tube Q14 and feedback resistors R286, R287; 1.5V group is the other operational amplifier in U11 (that is, the 5th, 6th, 7th) and adjustment Tube Q16 and feedback resistors R282, R283. The measurement found that the 9-pin voltage of U10 is only 0.76V, which is significantly lower than the normal 1.2V. After eliminating U11 damage, it is determined that U10 is broken. Since the main part (Vcore) of this board U10 is normal, only the reference voltage of the external power supply is partially damaged, and it is a pity that it is decided to repair the 1.2V reference voltage part. First, cut the connection between U10 9 feet and U11 3 and 5 feet (both op amp non-inverting input terminals) from the back of the board. Connect one end of the 330Ω resistor to +5V and the other end to the public connection of the 3rd and 5th pins of U11. Line part. Then solder a LM385Z-1.2 between the 3rd and 5th pins of U11 and the circuit ground (1 pin grounding, 2 feet connecting U11 3, 5 feet). The LM385Z-1.2 is a 1.2V reference voltage regulator IC with high accuracy and is used in the same way as a normal Zener diode. This additional part can be soldered to the soldering surface of the capacitor CT1 (the back side of the board), which is easy to solder and aesthetically pleasing.




　　25.PCCHIPS (Kirin) 530 motherboard is not bright




　　My friend gave me a PCCHIPS530 motherboard. He told me that the board is not lit. I took out this bad motherboard when I was idle, ah, I found that the copper foil under the CPU core power switch tube was actually blown, and the core power switch tube was broken from the appearance. First remove the switch tube, connect the burned wire with the wire, and replace the core switch tube. I am on the motherboard CPU socket after the Pentium 133CPU, power up, the motherboard is on. Then change the MMX166CPU and find that the motherboard is not lit up. In exchange for 133, it is normal, such as several times. I analyzed that the MMX166 is a dual-voltage power supply with internal and external cores. The power consumption is large, and the power supply may show problems at this time. The measurement confirmed this point. The normal external nuclear voltage of the MMX166 is 3.3V, and this board is only 2V. After careful visual inspection, it is found that the external nuclear power regulating tube also has signs of overheating. Replace the original tube with FD3055, and test the machine. Back to normal.




　　26. Kirin BXCEL PC100 motherboard does not light




　　First check the CPU supply voltage, find VTT is 0V, and normal should be 1.5V. The VTT group found that the voltages of the B and C pins of Q1 (H882) were normal, and the E feet had no output. Remove it and measure the open circuit. Look at the surface for a fine crack. Replace it with D882, the board can be repaired, pay attention to the pin arrangement when replacing. The picture on the left is the removed H882. You may not see the crack. In order to shoot the crack with a digital camera, I changed the Canon A10, Nikon 2500, Nikon 950, and Nikon 775 four cameras.




　　27. ST-694XVA motherboard does not light.




　　Measure the power supply voltage of each group of CPU and find that Vcore is only 0.5V, which is obviously abnormal. Check the power switch tube Q13, Q14 is normal, use the oscilloscope to observe the U19 (HIP6021) excitation pulse output, there is an output waveform, U19 should be no problem. Careful observation revealed that the bottom of the CE35 (16V1000μ) burst, and the plate returned to normal.




　　28. Lenovo BX1Brilliant-1 motherboard can not boot




　　That is, press the power button chassis power supply fan does not turn, no output. First test the 5VSB end of the ATX power supply, the voltage is normal. Since I didn't find the problem for a while, I wanted to know if the system part was normal. Short-circuit the PS-ON terminal of the ATX power supply to the ground and force the power on. At this time, the RESET light on the POST card inserted on the motherboard indicates that the system is always in reset and cannot be initialized. After that, the value of the 5VSB terminal was measured to be low with the multimeter diode. It is estimated that there may be a component leakage short circuit. When D17 was measured it was found to have broken down. Its model is 1N5226, which is replaced by a similar 3.3V Zener diode. After power-on, not only the startup problem is solved, but also the system RESET is normal. Here, D17 and a 56Ω combination get the 3V-STBY voltage from 5VSB, which is provided to the PWRBT (power switch) circuit and affects the POWERGOOD and PWROK circuits. Forced boot can make other groups of power supply, but POWERGOOD, PWROK, RTC circuits are still not normal, so the reset is not limited.




　　29. Miscellaneous 815EP (power-on self-test display 03/01/2001-i815E-627-6A69RFGDC-00, I do not know which manufacturer) does not shine




　　a miscellaneous brand 815EP (power-on self-test display 03/01/2001-i815E-627-6A69RFGDC -00, I don't know which manufacturer is). Plug in the POST card, power on, and find that the RESET light is always on, and the PCI CLK signal is not detected. Therefore, there is a problem with the clock circuit. Measuring the supply voltage of the clock IC (U9) found that its 2.5V group was only 1.6V, abnormal. Checking 2.5V is regulated by Q16 (TL431), where Q16 is no problem. Analysis of the circuit found that 2.5V is 3.3V and is regulated by Q16 after being stepped down by D19. The pressure drop of D19 seems to be abnormal. It was found to be broken in half during the inspection. After replacing D19, 2.5V was normal. In addition, when checking the CPU supply voltage of the board, it was found that the 2.5V copper foil was blown, and everything was restored after being connected by the flying line. It is estimated that the fault is caused by a problem such as a short circuit or a fire in the 2.5V group load.