Please help, FLUKE 8920A Multimeter Repair

[daisizhou]

Dear Harry_22.
When I press the "LOW RANGE ENABLE" button, the flashing occurs again.
Please see the video I took

[Harry_22]

Hello again!

These are the unpleasant consequences of repairs when you give equipment to the customer and he returns it again. Therefore you should always find the actual cause of the malfunction and not clean the connectors or replace all the capacitors first.

Let's continue.

Push "LO RANGE ENABLE" button.
Check that TP1 voltage is around 8.6V
Measure DC voltage at R18, R27, R30, R35, R40, R41 R42 and C24. You will get two values for each component. It doesn't matter which side is where.

[daisizhou]

Hi Harry_22.
Normally, the upper shield cover is closed,Press the "LOW RANGE ENABLE" button, the 2Mhz MAX indicator will light up.
If I test TP1, only the upper shield cover can be opened,Then when I press the "LOW RANGE ENABLE" button, the 2Mhz MAX indicator does not light up,Looks like the "LOW RANGE ENABLE" feature is gone,It may be caused by too much interference,The GND connection line is also connected normally.

In other words, if the upper shielding shell is opened to measure "LOW RANGE ENABLE", it will not be possible, so how should I measure it?

[daisizhou]

Hello again!

These are the unpleasant consequences of repairs when you give equipment to the customer and he returns it again. Therefore you should always find the actual cause of the malfunction and not clean the connectors or replace all the capacitors first.

Let's continue.

Push "LO RANGE ENABLE" button.
Check that TP1 voltage is around 8.6V
Measure DC voltage at R18, R27, R30, R35, R40, R41 R42 and C24. You will get two values for each component. It doesn't matter which side is where.

Once again I tested the above test points as you requested, but unfortunately the fault disappeared as soon as I opened the upper shield cover.To be precise, when I opened the upper shield cover, the "LOW RANGE ENABLE" function disappeared and the "2Mhz MAX" indicator light went out.
How do I test next?

[Harry_22]

The most sensitive range up to 2mV is turned on only when the Low Range Enable button is pressed and the Auto Range button is not pressed.  2MHz LED is lit when this range up to 2mV is active. If the signal is more than 2mV then the range automatically switches to a higher one and 2MHz LED goes out.
When the cover is opened the noise increases and the lower range no longer turns on.
Use an oscilloscope to see what is happening at TP1.

[daisizhou]

Normally, I turn on the 8920a power supply and observe the digital tube or pointer to see if there is a flashing fault,If it is normal, I will press the "LOW RANGE ENABLE" button and observe again whether there is flickering.

Do you mean to let me directly open the upper shielding cover and directly use the oscilloscope to measure the TP1 test point?If this is the case, how can I tell if the 8920a is working properly or malfunctioning?
Because sometimes when the upper shielding shell is opened, the flashing fault disappears immediately.

[Harry_22]

Hi! Let you do two additional connections on the PCB. Please see the scheme. Thus we add  divider by 10 to reduce the signal from first OPA. I expect that "Low Range" mode will more stable and you will see 2MHz LED now.
After modification please press "Low Range Enable" and measure the signal on TP1 by oscilloscope.

[daisizhou]

Hi.
Let me tell you good news. I tested today and the fault lasted for a long time.
The fault phenomenon is that "700" continues to flash.I tested the voltage at the test point you need.

Your latest connection, I haven't connected yet, if you need I will connect to test later

In order to prevent mistakes, I took a video. The video quality is not very good. I hope you can forgive me.Thanks

[Harry_22]

HI!
I need your help. Below is a diagram from Microcap. In the ovals there is the correct voltage in rectangles under the voltages measured by you.
It is necessary to supplement missed ones and check those already measured.
Pay attention to very different values ​​like at R27 and C24.

Also measure the voltage at In+ and In- inputs.

Please use a multimeter with mandatory monitoring by an oscilloscope 8.61V at the Out.

[daisizhou]

ok .give me some time i will measure them.

Before the measurement, I compared the voltage error values of my commonly used fluke 27 and TEK TDS2012C oscilloscope.
If you need to get high precision voltage value, I would use fluke 8060a or 179C for measurement

[daisizhou]

I have a question,For example R29，The voltage on the left is -7.328，The voltage on the right side is -7.326，Do I need to measure one to one according to the schematic?Or is it still effective if left and right are reversed?


Can I measure the input interface BNC in+ and in-? Because I don’t know the pin definition of Q9A、Q9B

When measuring, I still choose to disconnect the BNC input interface as before, right? Do I need to connect a 50 ohm load?
Because there will be interference when disconnected

[Harry_22]

Fluke 27 is Ok. I don't need high precision. Concerning drawing with R29 measure collector Q13 and VR3.
Please round the values 7.326 to 7.3, 7.967 to 8.0, 12.732 to 12.7 and so on.
Measure around Q17, Q19; Q11, Q16; Q8 both, Q10, Q12. Also R33, R39, In+, In-.
Connect 50 Ohm to the input and monitor the malfunction - +8.6V at the ouput.

[daisizhou]

Hello, I found that the voltage changes and is not a stable value.I don't know which point in time test is accurate.

The voltage increased from 1.46 to 1.651 and then increased to 1.806.The same goes for other voltages.

When the final screen displays 0.01, the voltage displays the highest value.(I connected a 50 ohm load)

I don't know how to measure
Does measuring these voltage values need to be measured under fault conditions? Or is it measured under normal conditions?

[Harry_22]

Please measure all the voltages under fault conditions.

[daisizhou]

for example,For the test point on the right side of R17,Also in fault condition,The voltage changes from 1.46v to 1.651v,Should I log 1.46 or 1.651?
Usually I test one resistance and then test the next one. I cannot complete multiple resistance measurements at the same time.
When I measure the test point on the right side of R17, the voltage is 1.46,Then I measured the voltage on the right side of R13 to be 1.851,The actual voltage on the right side of R13 is 1.72.

All voltages are changing slowly and I can't guarantee whether these measurements make sense.

so i'm lost

[daisizhou]

I tried to measure,Please note that I cannot confirm exactly the left or right side of the resistor，In other words, the voltage value on the left side of the resistor may be on the right side.

If you can please tell me and I will continue, if it loses its meaning I will wait for your advice

Say one more thing,During the measurement, I found that these voltage values can sometimes be stable,If a fluctuation occurs, the 8920a is about to return to normal

[Harry_22]

Very good.
Now please measure resistance between points 1 and 2 as well as between 1 and 3. Connect the positive probe to point 1.

Measure very gently without touching PCB parts!

[daisizhou]

The forward resistance is infinite. I tested the reverse resistance. Please see the picture.

[daisizhou]

Hallo.
During my test yesterday, I found that these two chips were heating up very violently.I'm referring to the fault status that shows "700" flashing,I used my finger to touch and it was so hot that I couldn't even keep touching that kind of heat.

Today the 8920a is in normal condition, I also touch it with my fingers and they only have a little bit of heat.

I don't know if this is a useful clue, I hope it will be helpful to your analysis.Do I need to use infrared vision to shoot video? Let me know if you need to check it out, the infrared temperature shot is at a friend of mine's and I need to borrow it.
Or I have a K-type temperature probe. If you need to check the temperature, I will shoot a video for you.

[Harry_22]

Well done for finding the overheating.
These are voltage stabilizers. This is another indicator that the operating mode of OPAmp_A is changing. It can be seen in Multicap too.
 
Measure the voltage at points 1, 2 and 3 at the time of the malfunction.

[daisizhou]

I once again confirmed that the resistance values ​​of 1, 2, and 3 are the same as my last measurement.

If you need the voltage values ​​of 1, 2, and 3, you need to wait for a few days, because the 8920a is in a normal state when it is turned on today. I need to wait for the fault to occur before measuring the voltage value.

If you need other test points, please let me know and I will pre-mark them because the failure time is very short.

[Harry_22]

The next time a fault occurs measure the resistance and voltage at point 3.

[daisizhou]

Hallo.

I had some free time today, so I tested the voltages at 1, 2, and 3 in normal conditions.Please note that I am referring to the normal state, which is the state where the 8920a display shows a reading of 0.01
1:10.94v   2:-1.405v     3:8.21v

Especially when I tested the P3 test points, I saw that the 8920a display could easily display 700 and flicker.
It seems that the P3 test point is a location that induces the "700" flicker,I don't know if it's useful, I hope to tell you and give you a clue
I mean the normal state, when the multimeter red probe tests the P3 test point, 8920a shows 700 flashes

[daisizhou]

I'll add my test observations.
Under normal conditions, if you use the multimeter's red positive pole to contact test point P1,8920a displays "700" flashing,As long as the multimeter positive pole and test point 1 are separated, the flashing of "700" will immediately resume and return to the normal display state.
If you put the red positive pole of the multimeter in contact with the P3 test points,Then "700" will flash. If you separate the red positive pole of the multimeter, the 8920a "700" flash will still exist and will not automatically recover.
At this time, I measured TP1 and the voltage was 0.114v

[Harry_22]

In this case look at the signal with an oscilloscope at TP1.