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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


The glue with the holes...

Verm

#17906

The glue with the holes... | 16 October, 2001

Hi all, I have a problem and would gratefully take on suggestions, firstly let me explain the situation: We are currently screen printing adhesive, using a method similar to DEKs 'Pump Print', The adhesive we are currently using is supplied from Alpha Metals and is called 'Epibond', the problem I have been experiencing is the adhesive is not forming in the way I have observed using other adhesives. It is behaving like a sponge and has bubbles in it, a similar formation to the chocolate bar �Aero�, when the PCA encounters the wave soldering process I believe that the solder is finding a path under chips components through the sponge formation and creating a short circuit, subsequently picked up at test, upon removal of the said defective device!, behold the short through the adhesive. I believe the shear strength of the component is not only jeopardized, but now we have a realistic problem of shorts �UNDER� devices!

Some people have been suggesting solder paste in the adhesive, this is being entrapped during cure and reflowed after hitting the wave.

Has anyone seen this phenomenon, or is it just us, is anyone else using the adhesive Epibond and what are your findings, please help.

Oh, I am currently investigating the oven profile as we speak, but I feel there is nothing to find there.

(G)

reply »

#17908

The glue with the holes... | 16 October, 2001

Look at "Microcanyons", Circuits Assembly magazine, December 1998. There D Pauls & T Munson describe adhesive on boards that was cured too rapidly and formed a skin that trapped volatiles and solvents. The volatiles and solvents created long voids in the adhesive, while out-gassing. [If you don't have a copy of the magazine, check at the magazine site or CSL site.]

We talked about using solder paste profiles when curing glue previously on SMTnet.

reply »

andyc

#17909

The glue with the holes... | 16 October, 2001

Have you thought about wave solder contamination ? . We had a similar fault where glued chips were being shorted out due to "slivers" of solder dross being deposited onto the card . These dross slivers were being generated because the solder bath had not been stripped out and cleaned regularly , resulting in the wave formers beiong clogged with dross . The action of the pumps caused these dross slivers to be forced around the bath and eventually onto the card . Check with your Maintenance dept. on when your last major strip out of the wave formers took place - if over 6 months , strip out and clean . I've also used Epibond but didn't see this "sponge" effect you refer to . I'd ask Alpha to come in and look at this with you to put both your minds at ease - it could be a bad batch of Epibond (?) .

Let us know how you get on ...

Andy .

reply »

bobm

#17911

The glue with the holes... | 16 October, 2001

i am using the epibond. i had alpha come in and look at our set up. they told me that the epibond can be reflowed at regular temp profiles with no problems. because i only have one oven and two pick n place machines it was important to be able to run two jobs thru the oven at the same time. i have not seen any problems so far. i dont run alot of glue jobs but will watch more closely from now on. i would check the wave out.

reply »

Michael Parker

#17920

The glue with the holes... | 16 October, 2001

I go with DaveF's advice. You are probably curing the glue too fast.

Another phenomenon that I have experienced was during glue cure and solder paste reflow simultaeously. A good reason to run a glue cure profile seperately from solder paste reflow.

As the glue first begins to melt it creates a liquidous pool around its base. Unfortunately, I had solder paste on pads so close to the glue that when the paste slumped as it came up to temperature, two liquid masses (glue and flux) converged, dragging the solder balls into the glue. The entire glue structure was peppered with solder balls. This forced costly and time consuming rework.

Glue will cure at about 180 F, with a slower ramp rate than paste, so be careful to not overcook the glue if you are trying to kill two birds with one stone.

reply »

Verm

#17930

The glue with the holes... | 17 October, 2001

Cheers guys for your rapid response, I investigated the problem further today, eventually I cam round to curing the adhesive in a sandwich of microscope slides resting on the PCA in question (hopefully mimicking the heat transfer), the adhesive is sponge like (to say the least) I am investigating the oven cure profile further but I am using the data sheet for the material and still finding the problem, Slowing the temperature ramp rate and extending the cure time is my next step and as Dave F mentions may help. HOWEVER!! I have found the problem of air bubbles is not restricted to Alpha's epibond as I tried this technique on other adhesives we have on site they all exhibit the same problem with the exception of one. Suggesting a natural phenomenon in adhesives�.. Hmmm I think not.

reply »

#17952

The glue with the holes... | 18 October, 2001

You�re correct. Probably every supplier�s product has slightly different cure characteristics.

Let me spin this up a little differently �

There was a slight imbroglio in my office this morning involving a couple of stalwarts and the last Krispy Kream. During the scuffle, they toppled the foot and a half stack of unread mail on my desk. While transporting this mess direct to trash, a glossy brochure [�The Effect Of Humidity Voids In SMT Adhesive On Wave Soldering� Kermit Aguayo and Philip Prud'homme, The Process Circuit, Summer 2001, No. 14.] expounding on the very topic fluttered-out of the soon-to-be pizza box. Fortuitous er watt??

�The Process Circuit� appears to be an in-house publication of a consulting firm. The article is posted on their site [http://www.process-sciences.com/]. Now, here�s the weird part. They leave the figures out of the glossy brochure, but state the figures are posted on-line. [I assume they take this approach to save money and to bait readers into visiting their site.] BUTT, when you visit the site, there�s no figures there either!!! That�s �

Anywho, they say �the root cause related to moisture absorption of the glue in combination with the temperature profile. As a result, the cure temperature profile was modified to exhibit a much slower temperature ramp (< 1�C/sec vs. 3 to 4�C/sec) with 140�C peak instead of 150�C.� [These guys may tout their process knowledge, but certainly writing a cogent sentence is not one of their skills.] Here�s the story. The voids were caused by: * Water [from unfiltered the air compressor lines] being injected into the adhesive during dispensing. * Fast adhesive cure ramp skins-over the adhesive surface and seals the water in the adhesive.

reply »

#17957

The glue with the holes... | 18 October, 2001

Are you filling your own syringes?

reply »

Verm

#17961

The glue with the holes... | 19 October, 2001

No, we are not filling the syringes, to be honest the problem is not with the displacement syringes on our conventional method of placing glue, but purely with screen printed adhesive. That's the weird thing.

(G)

reply »

LW

#17962

The glue with the holes... | 19 October, 2001

I was somewhat involved with Kermit Aguayo during his work on the SMA-void problem, as I was with one of the SMA suppliers at the time. I recall the problem being related mainly to absorption of moisture by the uncured adhesive. Some SMA formulations contain ingredients that are very hygroscopic. If there is more than a short (IIRC, 2-4 minute) delay between dispense and cure, the formation of "worm-holes" results.

New formulations have been developed to avoid this problem. I now represent one of the companies who has such a product.

reply »

Facility Closure

Global manufacturing solutions provider