Electronics > Projects, Designs, and Technical Stuff
(ultra)acoustic in circuit component level crack detector?
<< < (4/12) > >>
rhb:
This is a very interesting idea.

Besides general reflection seismology, I have been heavily involved professionally in what is called "rock physics".  This involves among other things making measurements of propagation velocity of small core plugs (1" x 2.5") using ultrasonic transducers at 1-2 MHz.

I suggest testing the following arrangement:

A long reach caliper with tranducers on each arm, a source and a receiver with a fairly soft pad on each arm to provide good coupling.  the source on the one side of the PCB and component and the receiver on the otehr.A cracked device should show up as a large attenuation of the signal through the board.

I suggest using some cheap 40 MHz 3rd overtone quartz xtals as transducers.  At least to start with.

I have a bin full  and will see see if I can construct a satisfactory experimental setup.  I have already found that by clamping the leads in a vise and pointing a Wen hot air gun at the case I could easily desolder and remove the metal can.  The quartz xtal is fragile, so an experimental setup will not be easy.  The ones I have are 0.15 mm thick.

I'm going to start with using hot melt glue to bond a pair together and drive one while looking at the other with a scope.
cdev:
Maybe something like this would help?

https://en.wikipedia.org/wiki/Quartz_crystal_microbalance  I stumbled across the concept a ways back when I was looking into ways to make my own sensors for environmental monitoring.

There also seems to be a project involving QCMs at https://openqcm.com
cdev:
Some of the Microchip PIC chips have a functional block, which can be used to do a great many things including ultrasound, which sounds to me a lot like what you are proposing. This is from Microchip Application Note AN1375 'see what you can do with CTMU' (Charge Time Measurement Unit)

"46 Ultrasound Imaging As described in previous applications, the CTMU can
be used to measure the time between transmitted and reflected impulses.
This information can be continuously fed in real time to a graphic processor, or
processing application, to create an image. This can be
implemented directly as an ultrasonic microphone (# 2)
or indirectly, through an ultrasonic pick-up (# 15)."


--- Quote from: coppercone2 on December 20, 2018, 05:11:10 pm ---what does this multi MHz transducer look like, cost, who sells one that can be attached to tweezers? i am guessing you can't vibrate a inductor for this purpose?

or electrorestriction using a capacitor as a source and reciever. Are the amplitude levels too low to be measurable if you connect it to a high gain multi stage amplifier?

--- End quote ---
rhb:
Detecting the reflection from a cracked SMD part is going to be very difficult, if not impossible. The delay time will be on the order of a couple of microseconds or less for large parts.  The only practical way to resolve that would be to use a chirp and cross correlation.  A big problem with that technique is it requires a really fast ADC.  But the LPC4370 might manage both the ADC and the processing.  It's an M4 clocked at 204 MHz with an 80 MSa/S 12 bit ADC.

I think the concept has potential, but it's seriously difficult to design an apparatus for this sort of thing.

I got a box of parts in the post, so I have a 1.8 V logic safe continuity tracer to assemble this evening.  Real rocket science  ;)  AAA cell, 1.5 V piezo buzzer and two banana jacks.  The buzzer gives a good tone at under 1 V, so I can use dead AAA cells.

BTW If you look at the QCM sensors and look at the crystals I opened you'll see that they are identical except for the xtals I'm using are $10/100.
cdev:
Maybe you can find out some useful info from time domain reflectometry app notes from the HP archive. http://hparchive.com/appnotes
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod