Great PC Board Diagnostic Tool
Posted: 24 Sep 2010 13:02
I have been designing and fabricating PC boards from ExpressPCB with surface mount components for over 7 years using the MikroPascal compilers to program the units.
My experience doing surface mount (0805 Surface Mount is the smallest size I am currently using) has been rewarding and generally successful - however, there are times when a board comes out of the oven with a problem. Previously the effort to track down and correct such problems were really time intensive and often frustrating.
I had gradually worked my way from just using a millimeter and testing individual components to using VI curves (Curve Tracers) which reduced the time and improved the success rate, but it was still time consuming and clumsy comparing traces of circuits of a good board with a problem board. I could never afford some of the $2000-$8000 VI based testers even though I drooled reading about their capability.
It got to the point where I simply put the "bad" board away - awaiting a better method of trouble shooting them. Well, its that method has arrived.
I was recently fortunately to find in a distant land (New Zealand to be specific) a device for $299 USD (+shipping/Taxes) that just blew my socks off. It really made a huge difference in the ease and precision in diagnosing problem boards. I have since recovered over 15 boards and have identified the problem on a dozen more.
I have really be astounded at how specific (with a little practice) you can identify which component in a circuit is the problem part.
But, rather than trying to describe it (no, I am not a vendor and have no financial interest in the sale of the product), I will post the link to the webpage describing it.
Yes, it based on the VI approach but its features for the relatively low price are really what make it so awesome. You don't even need an oscilloscope, it uses the PC - Nothing new there, but what really makes it sing are the features that makes using it so productive.
Basically you import an image of your board, designate any where up to 10,000 test points on that image, capture the impedance data and scope picture of the waveform for each test point on a good board, save all of that data and even assign fault codes and description to each test point. Then fire up the ROCKKIT test device and start comparing the same points on the board in question. You will generally get a "Pass" or "Fail" assessment for each point and the shape of the VI waveform generally points to the problem component. You can even specify what percentage tolerance for the testing - so it doesn't have to match 100% to get a "Pass". In some cases, the waveform oscillators due to the test signal triggering transistor circuits and so the wave form is not fixed - but, even there you can visually match the test waveform against the "Good board" waveform.
Yes, you can also compare the results from a good board to a bad board in real time - it you are a bit more ambidextrous than I am, but there is really not much point to this method given that capturing all the data from a good board and storing it the PC is so easy - that way, you don't even have to keep one good board on hand {:>)
Based on my sales price for my units, this $299 device has already saved me $7500 in recovered boards - The more complex your board the greater benefit this tool provides.
Again, I have no interest in this product other than to share it with others of you who may have modest capital to operate on and can not afford the megabuck test equipment
In any case, if interested check it out at:
http://www.dockerty.co.nz/products.html
Best Regards
My experience doing surface mount (0805 Surface Mount is the smallest size I am currently using) has been rewarding and generally successful - however, there are times when a board comes out of the oven with a problem. Previously the effort to track down and correct such problems were really time intensive and often frustrating.
I had gradually worked my way from just using a millimeter and testing individual components to using VI curves (Curve Tracers) which reduced the time and improved the success rate, but it was still time consuming and clumsy comparing traces of circuits of a good board with a problem board. I could never afford some of the $2000-$8000 VI based testers even though I drooled reading about their capability.
It got to the point where I simply put the "bad" board away - awaiting a better method of trouble shooting them. Well, its that method has arrived.
I was recently fortunately to find in a distant land (New Zealand to be specific) a device for $299 USD (+shipping/Taxes) that just blew my socks off. It really made a huge difference in the ease and precision in diagnosing problem boards. I have since recovered over 15 boards and have identified the problem on a dozen more.
I have really be astounded at how specific (with a little practice) you can identify which component in a circuit is the problem part.
But, rather than trying to describe it (no, I am not a vendor and have no financial interest in the sale of the product), I will post the link to the webpage describing it.
Yes, it based on the VI approach but its features for the relatively low price are really what make it so awesome. You don't even need an oscilloscope, it uses the PC - Nothing new there, but what really makes it sing are the features that makes using it so productive.
Basically you import an image of your board, designate any where up to 10,000 test points on that image, capture the impedance data and scope picture of the waveform for each test point on a good board, save all of that data and even assign fault codes and description to each test point. Then fire up the ROCKKIT test device and start comparing the same points on the board in question. You will generally get a "Pass" or "Fail" assessment for each point and the shape of the VI waveform generally points to the problem component. You can even specify what percentage tolerance for the testing - so it doesn't have to match 100% to get a "Pass". In some cases, the waveform oscillators due to the test signal triggering transistor circuits and so the wave form is not fixed - but, even there you can visually match the test waveform against the "Good board" waveform.
Yes, you can also compare the results from a good board to a bad board in real time - it you are a bit more ambidextrous than I am, but there is really not much point to this method given that capturing all the data from a good board and storing it the PC is so easy - that way, you don't even have to keep one good board on hand {:>)
Based on my sales price for my units, this $299 device has already saved me $7500 in recovered boards - The more complex your board the greater benefit this tool provides.
Again, I have no interest in this product other than to share it with others of you who may have modest capital to operate on and can not afford the megabuck test equipment
In any case, if interested check it out at:
http://www.dockerty.co.nz/products.html
Best Regards
- Basic Screen with Expanded scale photo of "good" test Waveform
- RockKit11.jpg (315 Bytes) Viewed 3008 times
.