Control board diagnostic techniques

[Bobeletronics]

Hello everyone, I am new to this forum.

I have been repairing cellphone, laptops, tablets on a component level for over 7 years. Recently I decided to pivot into commercial electronics as I see more opportunities to learn and grow as a technician.

I dont have formal education or training in the field, I learned how to repair portable electronics through thousands of trials and errors, I also had schematics and most of the times similar device to compare measurements to. In contrast, industrial electronics and appliances dont have schematic and the worst part is electronics components spread out throughout equipment which makes it hard to trace.
I can easily diagnose power supply board as there only few typical implementations. However, control boards are always different.

So I got couple of control boards for repair, one from welder and one from CNC machine. The boards are full of connectors, I am very curies how does one go about troubleshooting this kind of PCBs if there is no visual damage? How would

I know if I need feedback voltage from other parts of the system to activate the board or perhaps firmware doesn't boot? Or there is no universal approach?

I looked at some of industrial electronics repair website they claim they need just the board for repair, how do they repair it without schematics?

thanks in advance for any recommendation.

[Gregg]

The first thing to look for are fusible resistors; they are sometimes labeled something like 'F-1' but many times they are not labeled and appear normal.
With the boards out of the devices your job becomes a lot harder.  Maybe if you post a series of clear pictures, you'll get a lot more responses.

[Manul]

Hello,

Interesting question. I have quite a lot of experience repairing military equipment. Usually it is high-tech, complicated, no schematics and zero information on the internet. Some repairs take me months, literally.

So your question is what to do if you are faced with unknown. Well, I guess there is no magic. I don't know how these companies work which fix bare boards. I know just what I do.

First I try to gather at least some information about equipment. I talk with people, who use it. I look for information about similar equipment, scientific papers to familiarise myself with theoretical side if it is something new to me. So I read a lot.

I get back to the thing and look basic things, which are likely to fail, likely to experience stress, electrical, thermal or mechanic. Like for example power mosfets or fuses, input, output sections. If quick inspection with multimeter does not show nothing obvious then I start looking and reverse engineering by divide and conquer method. First rough, then going into more details as needed. What are inputs, outputs, power rails. After I have power rails and voltages confirmed, I can power it up. Look with thermal camera, measure voltages. If nothing is again obvious I go with more detailed reverse engineering.

I start drawing structural diagrams and collecting datasheets of all unknown components, IC's. If needed, I build some circuits to simulate inputs, write some test software. Also capturing internal board comunications, like I2C, SPI or whatever there is. Even remove and test separate components. This can really go deep and long.

This is all generalizations... You never know what will happen.

If I need to vote for the most useful tool it would be diode mode on a multimeter (it should go to at least 3V). For example, if there are many inputs like a data bus, I can go with diode mode (sometimes both polarities) to compare them. Compare other things. If there is a connector pin and I don't know where it goes, but it likely goes to an IC, I put my diode mode on that pin, get a voltage drop, go quickly with hot air over IC's and if I hit correct IC, voltage drop lowers, so it helps tracing. Also in some cases diode mode can be used as a current injector to simulate a signal. And more

Another useful thing on a multimer is LoZ mode. For example, there is 3.3V on a pin. Is it output or input with pull-up? Switch quickly to LoZ and if voltage drops a little, it is likely an input. And if pin is 0V? Is it output or pull-down? You guessed it. Diode mode injection. If 1mA can rise it, it is likely input with pull-down.

It is endless topic really. But if you like puzzle solving and don't get frustrated easily just try. Likely it will not be a one day repair, but if customer is giving you some time, go with courage and explore.

But if it is possible, always ask for the whole thing. That saves a lot of time.

[Manul]

So in your case I recommend to spend good amount of time (like one or two days) reading everything what you can find about welders and CNC machines, what are technical specifics about them, how technology changed, how it was 20 years ago, how it is now. What type of control mechanism they use, what may be different types of these devices, some theory, structural diagrams, maybe typical schematics.

[aqibi2000]

Large scale repair centres use working reference boards

[Bobeletronics]

Thank you very much, I am greatly relieved to learn that there is no magic tool and that it might take a while to figure out. Also thanks for insights into your troubleshooting technics. Diod mode is the number one tool in portable electronic repairs.

Another most important takeaways is that it is normal to ask the entire unit for repair instead of the board. No one talks about business side of things...

Hello,

Interesting question. I have quite a lot of experience repairing military equipment. Usually it is high-tech, complicated, no schematics and zero information on the internet. Some repairs take me months, literally.

So your question is what to do if you are faced with unknown. Well, I guess there is no magic. I don't know how these companies work which fix bare boards. I know just what I do.

First I try to gather at least some information about equipment. I talk with people, who use it. I look for information about similar equipment, scientific papers to familiarise myself with theoretical side if it is something new to me. So I read a lot.

I get back to the thing and look basic things, which are likely to fail, likely to experience stress, electrical, thermal or mechanic. Like for example power mosfets or fuses, input, output sections. If quick inspection with multimeter does not show nothing obvious then I start looking and reverse engineering by divide and conquer method. First rough, then going into more details as needed. What are inputs, outputs, power rails. After I have power rails and voltages confirmed, I can power it up. Look with thermal camera, measure voltages. If nothing is again obvious I go with more detailed reverse engineering.

I start drawing structural diagrams and collecting datasheets of all unknown components, IC's. If needed, I build some circuits to simulate inputs, write some test software. Also capturing internal board comunications, like I2C, SPI or whatever there is. Even remove and test separate components. This can really go deep and long.

This is all generalizations... You never know what will happen.

If I need to vote for the most useful tool it would be diode mode on a multimeter (it should go to at least 3V). For example, if there are many inputs like a data bus, I can go with diode mode (sometimes both polarities) to compare them. Compare other things. If there is a connector pin and I don't know where it goes, but it likely goes to an IC, I put my diode mode on that pin, get a voltage drop, go quickly with hot air over IC's and if I hit correct IC, voltage drop lowers, so it helps tracing. Also in some cases diode mode can be used as a current injector to simulate a signal. And more

Another useful thing on a multimer is LoZ mode. For example, there is 3.3V on a pin. Is it output or input with pull-up? Switch quickly to LoZ and if voltage drops a little, it is likely an input. And if pin is 0V? Is it output or pull-down? You guessed it. Diode mode injection. If 1mA can rise it, it is likely input with pull-down.

It is endless topic really. But if you like puzzle solving and don't get frustrated easily just try. Likely it will not be a one day repair, but if customer is giving you some time, go with courage and explore.

But if it is possible, always ask for the whole thing. That saves a lot of time.

[Bobeletronics]

Thanks, for the valuable information. I too always had working PCBs for typical laptops, phones to compare measures too. Nice to know that there is no easy way around it.