Printed Circuit Board Assembly & PCB Design SMT Electronics Assembly Manufacturing Forum

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Epoxied Parts Falling Off

I have a problem with 0805 caps and resistors randomly falli... - Feb 19, 2003 by RyanM  

RyanM

#23422

Epoxied Parts Falling Off | 19 February, 2003

I have a problem with 0805 caps and resistors randomly falling off in the wave solder machine. Currently I am placing single dots of epoxy under 0805's. The single dot is flattened down so I know that the part was placed. In the past when we double dotted 0805's we had problems with epoxy in the solder joints. I am using Loctite 3621 epoxy. When the board gets waved it is on it's third heat cycle. Any ideas?

reply »

#23430

Epoxied Parts Falling Off | 19 February, 2003

Consider that: * Something may be wrong with your glue. [That you have globs of flat glue on your board seems to indicate that your glue is probably cured properly. How easy is it to "pop" components from the board?] * Someone has put a goopy crud on your components. [This becomes less plausable with the increase in different part numbers affected.] * We have talked about this problem previously here on SMTnet.

reply »

#23440

Epoxied Parts Falling Off | 20 February, 2003

Interesting topic, I got a side bar question. The epoxy companies suggest a lower consistant temp (something like 1 minute at 160 degrees) for curing. But we epoxy and paste our boards that get waved. (I know, why are ya pasting a board that goes through the wave huh? Well we do, its a long story)So we use the profile for our paste. I'm sure our epoxy cures during the preheat zones, but what about the peak temp of 200 degrees? Ryan talked about 3 heat cycles, which means he heating the epoxy to this higher temp several times. Is there such a thing as too hot of temp for epoxy? What process does everyone else use to cure thier epoxy? Larry

reply »

MA/NY DDave

#23452

Epoxied Parts Falling Off | 21 February, 2003

Hi

I guess I would call Locktite.

I see David F expects that something isn't giving you a good bond strength and wants you to test it after initial application for good adherence. Check this out since his ideas are the most natural.

In addition, when talking with Locktite know your heat cycles, the dot size, and anything else about vibration in your process. My thought is that maybe something in your process has taking you beyond the adhesion limits.

YiEng, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

Jaime

#23462

Epoxied Parts Falling Off | 21 February, 2003

Call your glue manufacturer. It sounds to me that the epoxy is not curing properly. You should also test for shear strenght. How many pounds of pressure does it take to "pop" a cured component after it comes out of the oven? What type of oven do you use? Convection etc. Lenght of the oven, Speed of Oven conveyor, how many heating zones?

Just like LarryK, we also paste and glue our bottom sides, long story as well, I'm glad to hear that "we are not alone"

Our glue data sheet curing time calls for 60-90 seconds at 150 degrees. And 3 minutes at 120 degrees.

Please let us know how you solved the problem, good luck.

reply »

#23463

Epoxied Parts Falling Off | 21 February, 2003

Everyone is suspecting the epoxy. I'm just trying to understand why. Ryan stated he was having random chips falling off. If the epoxy or cure was the problem wouldn't he see the majority of the components falling off instead of just random?

reply »

#23464

Epoxied Parts Falling Off | 21 February, 2003

Because it's not limited to one part type and he's verified his profile...what else is there, unless the profile isn't what he thinks it is?

My question is, why does it stick to the boards but not the parts? Is the difference in bonding strength in those two interfaces high under normal circumstances (but still adequate for the application), or is it unusual in this case? Assuming the bonding strength is normally similar, what has effected the interface between the parts and the epoxy? Contamination from the process equipment, icky parts storage, or are all the parts from the same vendor and packaged the same way?

reply »


RDR

#23465

Epoxied Parts Falling Off | 21 February, 2003

Are you 100% positive that the parts are on before wave? I have found that asking an operator "were they there before?" does not always provide for an accurate answer. how is your handling before and after the wave process?

Russ

reply »

#23474

Epoxied Parts Falling Off | 21 February, 2003

Make sure your epoxy dispenser is repeatable. DO an experiment , inspect every board to verify epoxy is there before the part is placed... Is there an epoxy residue left on the PCB where the parts fall off after wave??? We do the epoxy paste process on a few board and it works well however, inspection after the paste and epoxy are applied is CRUCIAL... gluck

reply »

#23479

Epoxied Parts Falling Off | 22 February, 2003

In the early days of SMT we were using vibratory bulk feeders for chip components. It was quite obvious that some of the components had to try several times getting through the gravity trap or with the right side up through the light sensor. As a result body and leads were ground off and the powder passed through the feeding track. If the same feeders are used to fill the components into the tape then the components may be slightly dusted with ceramic and tin powder. You mentioned that 0805 components fall off. How about the other ones? I would not be blaming the epoxy�s exposure to temperature. The PC board and IC resist the soldering temperature very well.

reply »

#23483

Epoxied Parts Falling Off | 23 February, 2003

Larry,

Without being real clear on your process, as to whether you're: * Curing glue and reflowing paste in the same pass through the reflow oven ... OR * Using the same reflow recipe to both cure glue and to reflow paste.

We'll handle them both. We're not excited about either for the following reasons: * Cure / reflow at the same time: This is a documented disaster waiting to happen. The hot glue can absorb flux and flux res and then skin over, trapping that crud inside the glue deposit. Check the "Circuits Assembly" archives for an article by Doug Pauls and Terry Munson [sp?] at CSL. * Cure on a paste recipe: Loctite says, "Rate of cure affects strength => Higher than recommended is more brittle"

reply »

RyanM

#23492

Epoxied Parts Falling Off | 24 February, 2003

Just an update on my problem. I spoke to Loctite and they had very few solutions. One thing they suggested to try was to add a thirty-second soak time at 150 degrees to my reflow profile. The problem with that is ramp-soak-spike profiles doesn�t seem to work well with our solder paste. I am curing the epoxy as the solder paste is reflowed. I still question how the epoxy can stick to the board but not to random 0805's with different part numbers in different locations on the board. As I previously mentioned, the cured epoxy is left on the board "flattened" which I believe proves the part was on the board. When we evaluated this particular epoxy it had one of the stronger shear strengths. Plus adding reflowed solder to the end caps I don't think it's possible for the parts to be torn off of the board in handling. How long do any of you use your epoxy tubes before replacing them? Also, do you "cap" them when not in use or do you leave them on your epoxy machine?

reply »

#23493

Epoxied Parts Falling Off | 24 February, 2003

Every time this problem was brought to me, handling was the problem. Make sure to: 1- track the exact place the components are lost, i.e.: visually inspect the pcb before and after each step in your assembly line.

2- if step one does not pin point the cause, do a "rub" test just after curing the glue. A rub test consist in rubbing the cooled pcb with the bare hand to see if any component falls off. If you care about contamination, use rubber gloves.

3- That glue is not meant to resist "shocks" of any kind. If the edge of the PCB is tapped even slightly on a hard surface (like a table) then cracks may appear in the glue, moisture gets in, then pops in the wave...

reply »

#23497

Epoxied Parts Falling Off | 24 February, 2003

Dave, I went into their website. Unfortunately, they do not have a "search" option. I checked all of 2002 and Jan of 2003 archieves with out any success. You don't happen to know which month this article was featured in?

reply »

blnorman

#23498

Epoxied Parts Falling Off | 24 February, 2003

Having worked with adhesives for years, the first thing to look at if there is a failure of this type is surface contamination. The subject has been discussed here also. A structural adhesive will increase in strength with a "hotter" cure. We also paste/adhere/wave, but never 2 operations concurrently. Boards are reflowed, then adhesive applied and cured, then waved. We've been using Loctite Chipbonders for years with no real problems. We use a different screen print adhesive that we have had problems with.

reply »

#23499

Epoxied Parts Falling Off | 24 February, 2003

Larry,

CSL used to publish the whole article online. Now, they just have an abstract. Bummer. Look in Circuits Assembly magazine, 12/98 for "Microcanyons"

reply »

MA/NY DDave

#23500

Epoxied Parts Falling Off | 24 February, 2003

Hi

Claude and the next person gave some good advice I would follow. When I spoke above about vibration I was thinking of machine, transport and other, yet shock like tapping the board on it's side or otherwise is pretty much the same. You create micro fractures.

Contamination or something that prevents appropriate adhesion is also another concern.

Again before you call Locktite, it you choose, LOOK closely at your process across all shifts. Do a shaddow on every step along the way and you will learn a lot. Believe it or not it might be do to ONE human being.

YiEng, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

#23506

Epoxied Parts Falling Off | 24 February, 2003

Ryan,

You are spot-on with your analysis of the 'flattened' glue deposits.

How about this for a different approach[es] in explaining your method for removing caps from your board? * George Verboven has speculated here on SMTnet that the downward force of the wave is ~3 newton. Can the glue holding your devices stand upto this force after preheat? [Where is George, anywho? He always has good insights on this 'glued parts disappearing before coming out of the wave stuff'.] * Sometimes the chimney in the wave can scrub components from boards [Maybe it's NOT random 0805s. Maybe it's tall 0805s.]

reply »

MA/NY DDave

#23514

Epoxied Parts Falling Off | 25 February, 2003

Hi

Locktite is telling you to cure first, maybe for a good reason. So the epoxy becomes stable just before solder reflow moves stuff around and around and then settles.

A WEG yet if you did some cross sections and analysis you might find piles of fractures in the epoxy on the component side of the bond.

I think you may have found your problem.

YiEng, MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

MA/NY DDave

#23521

Epoxied Parts Falling Off | 25 February, 2003

Hi

Up above in his operation / process description after a call to Locktite and then my note and then your note, we may have found the problem. One never knows at a distance.

It appears he is striving to do soldering/reflow first and then concurrently or there-afterwards epoxy curing.

YiEng MA/NY DDave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »

#23524

Epoxied Parts Falling Off | 26 February, 2003

Dave, I had to contact Circuits Assembly, but I got the article. I'm going to forward it on to Ryan. I think you're on to his problem with flux becoming entrapped in the epoxy. It plays along with Locktites suggestion of a longer soak on his profile.

reply »


mr

#23625

Epoxied Parts Falling Off | 3 March, 2003

Hi,

We dealt with and solved this problem here at Agilent...We have been doing simultaneous reflow cure in convection reflow ovens for all of the 9 years I've been here with minor problems of chips falling off, and after switching to Loctite 3609 the problem Ryan describes completely disappeared.

The article on "microcanyons" is right on...I took many cross-sections of parts still on boards that also had missing (same) components, and found solid evidence that the adhesive we used to have allowed (if not facilitated) migration of flux and other paste binders OVER the glue dot, preventing adhesion to the part. I have many SEM photos of this phenomenon. The fact that the glue dot is flat (after the part is missing) is PROOF that it once was there. I could go on forever about this, but the short end of it was not contamination or profile related--switching to 3609 made the problem disappear, and that was two years ago. Since then, I have NEVER seen another 'flat' glue dot with part missing. If anyone wants more information, give me a shout...there's just too much to write down! MR

reply »

Mark

#23759

Epoxied Parts Falling Off | 11 March, 2003

Please inspect the pcb were the components were "lost" Check if the glue dots were lef on the PCB.If so you have a component which has been contaminated.If the dots dont appear on the PCB,which is the more common of the 2,there is an adhesion problem on the PCB.To do a non destructive test on 0805`s the component should withstand 200 cNcm. Lastly check that you dont have a water problem.H2O seriously decreases the strenght of any adhesive. Mark

reply »

Judy

#23883

Epoxied Parts Falling Off | 21 March, 2003

I look at boards coming out of the oven top side and bottom If I take that board after its been glued on bottom side out of the oven and look at the top side when I pick up that board you should see the glue parts that fell off you really have to watch the handling of a board with glue If you our putting your boards on a rack and sending them to the wave put them in every other slot if you put them in every slot the chances are you might be hitting them together So the parts are already off befoe they even go into the wave

This message was posted Add this forum to your site! Click to learn more. the Electronics Forum @

reply »


DL

#23932

Epoxied Parts Falling Off | 26 March, 2003

I Like your analysis here. I dunno the whole story, but I've tracked parts popping/falling off to the metalazation of the parts, from various suppliers, as most suppliers have some type of pre-tinned coating on the leads.(not that this means anything) I dunno if the majority of suppliers had epoxy in mind when they designed their packages. I know my buyers, bring in the cheapest bunch of crap parts, then we have to SCREAM at 'em to get good parts.

Good Luck

reply »

PCB machines

PCB equipment