Please help!

[Luan2002]

My friend just brought up me this topic about Capacitors. He has some problem with polarized CAP at his work place. These CAPs were installed backward on the PCB. However, the In-Circuit-Test did not detect this problem. When it come to functional test, the CAP blows up.

Please help! if anyone know how to build the circuit to detect this problem before it blows up. I appreciated it.

 

[Lewish]

IFR has an ATE system that claims to be able to detect reversed polarized capacitors. Take a look at:

http://www.ifrsys.com/solutions/ate/4200/4200_home.htm

It is the only solution I am aware of. Anyone else know of others??

 

[Luan2002]

Thanks Lewis,

That's a great resources. However, What I want a small circuit that can test the polarized CAPACITOR when the board already power on at short period of time. What I mean is that circuit is going to detect any reversed CAP on the board, before it pop up. The ATE testers are fine. However, I assume that in the worst case scenarios, if there is no testers around, then we can use that small circuit.

Please advise!

Thanks in advance! Luan

 

[buzzp]

SMT machines will test the first cap on a reel. I do not remember all of the particulars. I just know they have the capability. I would think the caps would have to be tested BEFORE the board is powered up to prevent damage to reversed caps. I would guess a small voltage could be applied to a known capacitance while measuring the current and characterized that way to make the determination if the cap is backwards or not. Just making an educated guess here. The type of capacitor will probably matter as well, ie tantalum, aluminum electrolytic, etc..
I guess the method to make this determination would depend on the mfg process. If your doing real low volumes with low component counts then maybe a visible inspection is possible. Could you explain more about what your in circuit test consist of?

 

[uWaveFDL]

If the problem is that prevalent it's more of a lack of proper QC/QA oversite. (Your Engineering and Manufacturing Engineering depts are not on the ball either.)
If this problem is after a picknplace assembly....
Get someone competent to load the parts profile.
Check your machines vision ID part profile for the problem part(s).

 

[Luan2002]

Thanks to all for replying to my post. The company has QA, QC and manufacturing Engineer. However, I am talking about the worse case scenarios. What if some one just not doing the right job on one day, then something will happen. So if we have a "detect reverse CAP" circuit in our test area, then we will catch to failure. Other than that, I think they are fine with their process.

Regards..Luan

 

[melone]

The easiest way to determine whether a cap is backwards is the leakage current (assuming that you cannot just look at the component). However, I have also heard of xray being used for this sort of thing.

Good luck and keep us posted!

 

[LewisBrien]

Is it possible to redo the ICT Program to measure resistance/capacitance between a few points around that capacitor, So if there is a difference( I.e if the CAP is in the wrong way, it might be Spotted. Not much hassle to do this.

 

[melone]

The problem is that the capacitance of a polarized cap is the same forwards or backwards. Also, depending on the amount of additional leakage paths in your circuit, you might not be able to "see" this problem using a leakage current method. The BEST solution to the problem is to prevent it from happening. Otherwise, you will be forced into stressing out the component and waiting for smoke.

Good luck, and if you figure out a method please post the solution! (This is a VERY common problem)

 

[buzzp]

I think it can be done it is just the test method involves power to the cap. If this is done with voltage near the cap voltage (or maybe not even close?) then some damage is already done. There are machines that actually do visual inspection for you using DSP.
Melone: Is the leakage current higher or lower for say an aluminum electrolytic that is inserted the wrong way? Or is it nonexistent?

 

[melone]

Leakage is higher if the aluminum electrolytic if biased in the reverse polarity. Actually, there is a 10-1000X increase in leakage on a typical lytic. Unfortunately, the small change in leakage attributed to the polarity of the lytic is usually masked by the overall circuit leakages.

 

[skenn]

Can the ICT measure resistance at a significant voltage. It's been a while since I have worked with ICT, but you should be able to detect the overcurrent in this test BEFORE blowing the cap.

 

[negativeterminal]

You can send the board through an AOI inspection machine just after it leaves SMT.

http://www.orbotech.com

has some but I think that they are high maintenance. Photon Dynamics makes a decent AOI machine, but I've heard rumors that they are no longer making new ones. Hope this helps.

 

[MaximumMax001]

we just got an AOI and it seems to be working fairly well. It is a vi-3000 if you need a name.
Hope this helps

 

[Kido]

i think your problem is that you want to test the cap before it blows up but after you have put power to it, this is not a good idea as after puting a reversed potential accross it you have already done some damage, so it is probably best that you let it blow up then go around with a schematic and visualy check all the boards from that batch and then fix them that way, because as i say you have already damaged it and it will need replacing anyway... this is what we do when we check our power supply boards for faults, check them visually then power up and hope for the best, standing clear as well also helps with litics...i hope this helps