IV testing of electronic components

[dren]

Hello,

I am a mechanical engineer working in packaging of electronics and as part of my project I have to do IV characteristics test on electronic components and compare them with the charts given in the datasheet. The purpose is to see if the interconnection ( wire bonds) are intact or broken.

I am not sure how to go about doing this, would be very helpful if someone can give me leads on this topic.

Thank you
 

[HKJ]

Try searching for a Keithley 2450 SMU, it is designed to do that, but it is rather expensive.
Agilent has the same type of instrument, but it is just as expensive.

[jpb]

I-V characteristics are both rate and device dependent, but fortunately you're only wanting to check if bond wires are broken or not. Does this mean completely broken or are you checking if one out of three to the same pad is broken for instance which would be much more difficult (you'd need to do some sort of inductive or accurate resistive test).

Assuming you're just looking for a connected/not connected type test then you don't need full I-V characteristics, you just need to check (for a FET for example) that the gate turns the device off and on and that current flows between source and drain.

[fubar.gr]

How about an octopus component tester? If all you need is a simple pass-fail test, maybe the old octopus could be more than enough for the job.

[SAUL BRITTO]

The octopus looks like perfect for that.I have some plans to octopus use.

[dren]

Hi,

Thanks a lot for the reply. I was reading about the octopus tester. Is there any source which tells about how to go about testing the pins? What pin combination should I use ? Pardon me, if my questions are too fundamental.

[philpem]

I've done it on a test bench with a Yokogawa GS200 voltage-current source (the older 7651 also works, and the ADCmt 6146 probably will too), an oscilloscope and a current probe (a shunt resistor is sufficient).

Set the GS200 for a sweep, whatever voltage range you want. Make the sweep fairly quick, a few hundred milliseconds. Depends how many steps you need really.
Set scope for X-Y. X axis is voltage across the component. Y axis is current.

If you can't get your hands on a GS200, DIY a ramp generator or sine oscillator and a power amplifier to replace the GS200. If you have a function generator, use it and build a power amp for it. Couple of opamps, a gain pot and a Class-AB output stage.
If you want to do better, use a high-gain instrumentation amp and a low-value current shunt resistor to reduce burden voltage in the current sense stage.

You could probably DIY most of this into a self-contained measuring instrument for not a lot of money. There was an old Maxim/Dallas appnote which explained how to turn a PC with a parallel port into an I-V curve tracer. It was as slow as molasses, but it worked reasonably well.

[dren]

Thanks for the suggestion philpem!

It was very helpful. I dont have a GS200 but will use a programmable supply that is available.

Do you know how to go about measuring the pins in the component? Like what pin should I connect to the supply and from what pin should I measure?

Thanks

[David Hess]

Sounds like you are looking for something like a Huntron Tracker which is just a dedicated two terminal curve tracer.  They are commonly used for the type of diagnostics you described.