Bad SMD Cap Test

[v1.5]

hi all

i am here to ask you elders and professionals that how to test a smd cap in the circuit . Im sure that this question has been asked before many times but for upgrading and sharing new methods , i decided to ask it again.
As everyones know , smd non-polarity ceramic capacitors are the worst and more difficult to test on the circuit board. For that reason im looking for time saving methods can be apply to the capacitors without removing them all.
For that reason ,what is the best and fast method? What do you think ?

Thanks all

 

[Nigel Goodwin]

Why do you feel the urge to test them?, do they ever go faulty?.

As a professional I don't go round removing random components and testing them, I fault find and reduce the possible causes to a small and easily checked area of the board with few components on it.

I would imagine you're going to damage, or lose, FAR more ceramic capacitors than you ever find faulty.

 

[v1.5]

Why do you feel the urge to test them?, do they ever go faulty?.

As a professional I don't go round removing random components and testing them, I fault find and reduce the possible causes to a small and easily checked area of the board with few components on it.

I would imagine you're going to damage, or lose, FAR more ceramic capacitors than you ever find faulty.

hi thanks.

Yes i have found a few failed ceramic capacitor on the circuit and already replaced them with new ones. Those were especially decaupling caps which tied to ground and vcc on the circuit at the near of the ic. After replacement few boards started to work properly. But still there are lots of electronic cards that i have on my inventory waiting for check / test . I cannot lose them cause nowadays hard to find those cards on market. Because of this time saving tester would be great. Thats the reason i have asked that . So how do you think to find out a shorted ceramic capacitors which tied two power planes on the board without removing them ?

 

[Pommie]

If it's shorted then it will get very hot. Not sure how it would get shorted. When were these boards made?

Mike.

 

[v1.5]

If it's shorted then it will get very hot. Not sure how it would get shorted. When were these boards made?

Mike.

Sometimes im not able to test those boards on the field . This is why i cannot check them heat. Mostly made on 2003.

 

[Nigel Goodwin]

If they are going short, then apply an external high current power supply to the faulty rail, and see which ones get hot.

 

[v1.5]

If they are going short, then apply an external high current power supply to the faulty rail, and see which ones get hot.

thanks that will save my time. Here need to ask that , for example i will apply 1V and 5 amps current through the rails of the capacitor and will look for heat right ?
Is 1 Volt ok for testing them ? If them short voltage doesnt important right ? On this way will no risk other component from over voltage fail.

 

[narkeleptk]

I typically apply 3-5v depending on the system. 5a is a bit high. Start at 1a then work your way up. (usually you wont make it past 3a )

 

[v1.5]

I typically apply 3-5v depending on the system. 5a is a bit high. Start at 1a then work your way up. (usually you wont make it past 3a )

Thanks a lot for such a good information i ll note that

 

[Pommie]

If they are SMD, even a 1206 will get very hot with a small current. 5A will probably destroy them completely. At least you'll know what needs replacing.

Mike.

 

[Nigel Goodwin]

thanks that will save my time. Here need to ask that , for example i will apply 1V and 5 amps current through the rails of the capacitor and will look for heat right ?
Is 1 Volt ok for testing them ? If them short voltage doesnt important right ? On this way will no risk other component from over voltage fail.

If it's a 5V rail, then apply 5V - no high voltage issue if you're using the right voltage. As others have said, start with the current fairly low and work up.

 

[unclejed613]

If they are going short, then apply an external high current power supply to the faulty rail, and see which ones get hot.

since all of the decoupling caps are in parallel, this makes the best sense, and a "tester" would just read a short everywhere you check, unless it can show differences of milliohms (from the traces on the circuit board)...

one manufacturing company i worked for had a PCB with about 20 op amps on it.... one common board stuffing error was an op amp inserted backwards, and occasionally a board with one or more reversed op amps would find it's way past all of the QC stations and end up at the initial test station as a failure (supply overcurrent)... the test tech came up with a quick way to identify reversed op amps, connect it up to a supply with no current limiting, and watch for smoke.... simple yet effective...

 

[Nigel Goodwin]

since all of the decoupling caps are in parallel, this makes the best sense, and a "tester" would just read a short everywhere you check, unless it can show differences of milliohms (from the traces on the circuit board)...

one manufacturing company i worked for had a PCB with about 20 op amps on it.... one common board stuffing error was an op amp inserted backwards, and occasionally a board with one or more reversed op amps would find it's way past all of the QC stations and end up at the initial test station as a failure (supply overcurrent)... the test tech came up with a quick way to identify reversed op amps, connect it up to a supply with no current limiting, and watch for smoke.... simple yet effective...

A VERY old saying in TV servicing was "switch on, and tune for maximum smoke"

 

[unclejed613]

A VERY old saying in TV servicing was "switch on, and tune for maximum smoke"

in the audio business that can be a very expensive way of troubleshooting after spending a couple of hours rebuilding an amplifier... i prefer some method of current limiting, so i can power up the board and test it without worrying about letting out the smoke... finding backward op amps is one thing..... finding out that your bias transistor isn't working in a multi-hundred watt amplifier is another....

 

[vtech]

in the audio business that can be a very expensive way of troubleshooting after spending a couple of hours rebuilding an amplifier... i prefer some method of current limiting, so i can power up the board and test it without worrying about letting out the smoke... finding backward op amps is one thing..... finding out that your bias transistor isn't working in a multi-hundred watt amplifier is another....

Absolutely, specially dealing with pro audio with expensive, hard to reach & find matched semiconductors
Using a Variac was always the best way to bring up the rails slowly while monitoring current draw .

 

[Nigel Goodwin]

in the audio business that can be a very expensive way of troubleshooting after spending a couple of hours rebuilding an amplifier... i prefer some method of current limiting, so i can power up the board and test it without worrying about letting out the smoke... finding backward op amps is one thing..... finding out that your bias transistor isn't working in a multi-hundred watt amplifier is another....

Notice the post said "very old" (valve) and TV - no mention of modern DC coupled transistor amps - although it was just as relevant for valve amps from the same day.

 

[unclejed613]

"very old" (valve) and TV

i had a Scott radio from 1938, and it kept burning out a resistor in the power supply... it was caused by paper bypass capacitors throughout the radio shorting one by one. i finally replaced them ALL with mylars.

 

[unclejed613]

Using a Variac was always the best way to bring up the rails slowly while monitoring current draw .

even better is a series incandescent bulb... very old school but it works....

 

[Nigel Goodwin]

even better is a series incandescent bulb... very old school but it works....

It's a VERY standard procedure, far more common (and useful) than using a variac - a 60W bulb usually seemed about the best for general use - I always had one with a couple of wires soldered to it. If you used a bulb holder, someone would nick the bulb - soldering the wires to it prevented that.

 

[unclejed613]

I always had one with a couple of wires soldered to it.

i use a "furnace switch", an electrical box with a switch and an edison base fuse socket (which is identical to an edison base lamp socket), with an outlet box next to it.