Heat Damage to Inductors

rds have a common problem: an 0603 sized inductor, a Coilcraft 0603PS-392KL B, has been damaged. Visually, they look okay, but their value is much too low. I've confirmed that the inductors shipped to the assembler were of the correct value (I measured the spares) so I'm suspecting that the assembler may have overheated the board and the inductors were a casualty. I would n ot have expected these inductors to be a weak link on the board: it seems t o me that some of my other components (e.g., RF MMICs) would fail also, but that has not happened; only the inductors are bad.

to heat?

Update: I'm also seeing damage to another inductor; a Coilcraft 0805LS-222X GLB.

Look at the failed ones under a microscope. Chance are that the ferrite is cracked. That would cause a substantial drop in inductance.

--
Regards, Joerg 

http://www.analogconsultants.com/

How are you measuring them? In circuit?

We've used zillions of Coilcraft 0805 and 0603 parts, no problems. But not that shielded one.

The ferrite shield may be conductive, and the solder joints might be hitting the shield, which would short it out.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

My PPoE had a lot of problems with inductors during RoHS reflow. It took quite a while to get the profile right. LEDs were another major casualty of RoHS. They eventually had to buy a new oven (9-stage, IIRC) to get a usable profile.

If you haven't done so already, double-check your assembler's temperature profile against Coilcraft's.

--
www.wescottdesign.com 


----Android NewsGroup Reader---- 
http://usenet.sinaapp.com/

rds have a common problem: an 0603 sized inductor, a Coilcraft 0603PS-392KL B, has been damaged. Visually, they look okay, but their value is much too low. I've confirmed that the inductors shipped to the assembler were of the correct value (I measured the spares) so I'm suspecting that the assembler may have overheated the board and the inductors were a casualty. I would n ot have expected these inductors to be a weak link on the board: it seems t o me that some of my other components (e.g., RF MMICs) would fail also, but that has not happened; only the inductors are bad.

to heat?

Thanks for the input everyone. I inspected them with a microscope and measu red them out of circuit. Again, I could not see any damage but measurements with a LCR meter showed them way out of spec. I also compared the measurem ents with the spare inductors in case my meter was not accurate.

Darol Klawetter

rds have a common problem: an 0603 sized inductor, a Coilcraft 0603PS-392KL B, has been damaged. Visually, they look okay, but their value is much too low. I've confirmed that the inductors shipped to the assembler were of the correct value (I measured the spares) so I'm suspecting that the assembler may have overheated the board and the inductors were a casualty. I would n ot have expected these inductors to be a weak link on the board: it seems t o me that some of my other components (e.g., RF MMICs) would fail also, but that has not happened; only the inductors are bad.

to heat?

Probably not the case, but I had (big through hole) inductors that lost Q and inductance do to water that got in during the cleaning process. A soft dry in an oven cured the problem.

George H.

We've been using panasonic SMT inductors for some years with no problems. Here's the Digi-Key link:

We have used thousands of their 100, 33 and 10 uH versions of these in a batch reflow setup using a toaster oven and thermocouple. So, the thermal profile is not extrremely tightly controlled. We know what they look like when the do get cooked, as sometimes there is a load-side short.

These units are a good deal larger than 0603, though.

Jon

Is there any ways to verify what temp profiles were used by an assembler, other than what they say? Like a thermal sticker or something that records what really happened to a board?

:

oards have a common problem: an 0603 sized inductor, a Coilcraft 0603PS-392 KLB, has been damaged. Visually, they look okay, but their value is much to o low. I've confirmed that the inductors shipped to the assembler were of t he correct value (I measured the spares) so I'm suspecting that the assembl er may have overheated the board and the inductors were a casualty. I would not have expected these inductors to be a weak link on the board: it seems to me that some of my other components (e.g., RF MMICs) would fail also, b ut that has not happened; only the inductors are bad.

le to heat?

sured them out of circuit. Again, I could not see any damage but measuremen ts with a LCR meter showed them way out of spec. I also compared the measur ements with the spare inductors in case my meter was not accurate.

The ferrite doesn't care about soldering-heat--it's sintered for Pete's sake.

Assuming the core isn't cracked, with air-gaps, shorted turns, maybe?

D.C.R. comparison might be interesting, plus d(L) reportage.

Cheers, James Arthur

1: That's a damned good question.

2: I don't know.

3: Ooh! A product idea! I wonder...

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

This is kind of bloviating, but if you could squeeze the core in various ways you might see an inductance change if you either closed or opened a gap.

For that matter, if the thing is supposed to be shielded and isn't, because of core cracks, just placing ferrite material over the cracked part should lower the inductance.

For good measure (more bloviating, by the way) if you _carefully_ unsoldered a good part and a bad part, then there's a chance that the bad part would crack more easily.

Even more free bloviating: make an itty-bitty magnafluxing rig to look for cracks. Or the dye stuff that they use for non ferrous materials.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

There are gadgets that ride through with the board and log the profile. I think you can ask a serious assembler to provide a log.

--

John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  

jlarkin att highlandtechnology dott com 
http://www.highlandtechnology.com

A friend of mine makes, or made, or designed, or something, a datalogger that you could run through a solder wave machine to verify proper temperatures throughout the process (big-ass high-temp electronics!!!). Such a thing could be made to work in a reflow oven, I suppose -- but I don't know if they're out there or who to ask.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

Tempilabel, they were on the market in the 1970's, for sure. Still available from Tempil.

Jon

Worked on. He's not an employee:

So a profile can be made.

--

Tim Wescott 
Wescott Design Services 
http://www.wescottdesign.com

That's what was happening to us. Our audio transformers were shorting. The temperature peak was too high. Any lower and the QFNs didn't flow. The solution was a new oven (large aluminum caps were still a problem).

Not a full temperature profile, but tinted crayons and lacquers and labels that color change when temperature reaches a trip-point...

There are optical fiber thermometers (cannot use electricity during heating of the sensitive parts) that could be added to a reflow process oven without much inconvenience.

But, of course, inspecting the solder joints is supposed to be the final test, anyhow. It might have been easier with tin/lead (shiny was the expected luster) than with tin-silver-copper SAC (frosty is expected).

:

ote:

e boards have a common problem: an 0603 sized inductor, a Coilcraft 0603PS-

392KLB, has been damaged. Visually, they look okay, but their value is much too low. I've confirmed that the inductors shipped to the assembler were o f the correct value (I measured the spares) so I'm suspecting that the asse mbler may have overheated the board and the inductors were a casualty. I wo uld not have expected these inductors to be a weak link on the board: it se ems to me that some of my other components (e.g., RF MMICs) would fail also , but that has not happened; only the inductors are bad.

rable to heat?

measured them out of circuit. Again, I could not see any damage but measure ments with a LCR meter showed them way out of spec. I also compared the mea surements with the spare inductors in case my meter was not accurate.

Yep. If the ferrite's intact it pretty well has to be the copper, doesn't it?

Cheers, James Arthur