Homebrew PCBs

I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?

If you are in the USA, you could put in for a stimulus grant for additional 
gravity in the area of your hot plate :)

Otherwise, I'd look at the preheat.  Sounds like there is still moisture in the 
paste.

Donald.

Jim Lewis wrote:
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
>

What I do (which probably isn't best) is to put the board on a cold
hotplate, then turn the hotplate on.  It takes about 3 minutes to heat
up to soldering temperature.  I shut the hotplate off as soon as
soldering starts (it carries through anyway), then slide the board off
as soon as all the connections are soldered.

--- In Homebrew_PCBs@yahoogroups.com, "Jim Lewis" <jim.lewis@...> wrote:
>
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
>

Sounds like popcorning. It could be moisture in the chips or in the solderpaste.
If the moisture is in the chips, you could try warming them at around 150F for 30 minutes to an hour to drive out any moisture. You could also have moisture in the solder paste itself. I am not sure how how you would deal with that. Maybe a longer preheat at a lower temperature may help there as well.

----- Original Message ----- 

From: "Jim Lewis" <jim.lewis@...>
To: <Homebrew_PCBs@yahoogroups.com>
Sent: Saturday, April 04, 2009 12:12 AM
Subject: [Homebrew_PCBs] Jumping chips


>I have a problem with SMT chips jumping off the board during reflow. I 
>preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 
>minutes directly on the hot plate at 375F with a cover. Paste is from 
>Stencils Unlimited. Any solutions?

Your PCBs might have absorbed moisture if they have been left around for 
some time.

Leon

Try using an IR heat lamp like for keeping plants or reptiles warm on
the top.  You need to thoroughly heat all of the parts and make sure
the flux is hot enough to activate.

On Fri, Apr 3, 2009 at 4:12 PM, Jim Lewis <jim.lewis@...> wrote:
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
>
>

-- 
Erik L. Knise
Pacific Shipping Company
Seattle, WA

Thanks for all the tips. I pretty sure it's moisture in the paste. And it was much worse on a high-humidity day but not sure if that was a coincidence. Has anyone else noticed a connection to humidity? Does anyone know the ideal pre-heat temp and duration per the mfg?

I would check the data sheets for your paste and your parts and see
what kind of heating curve they use for re-flow soldering.  Also, make
sure you don't bake your parts to the point of no return.  There
should be a pre-heat ramp up in temperature and it usually plateaus
for pre heating then rises up to reflow temp and then tapers back off.

On Sat, Apr 4, 2009 at 5:04 AM, Jim Lewis <jim.lewis@...> wrote:
> Thanks for all the tips. I pretty sure it's moisture in the paste. And it was much worse on a high-humidity day but not sure if that was a coincidence. Has anyone else noticed a connection to humidity? Does anyone know the ideal pre-heat temp and duration per the mfg?
>
>
-- 
Erik L. Knise
Pacific Shipping Company
Seattle, WA

PCB industry glues the SMT components to the board before reflow. It's a  
pain in the butt if the components have to be replaced, but this how the problem is overcome. Finding the right cement is the key.

David.



--- In Homebrew_PCBs@yahoogroups.com, "Jim Lewis" <jim.lewis@...> wrote:

>
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
>

No they don't.

For reflow the components are rarely (if ever) glued.

For wave soldering they are glued of course.

ST

On Thu, Apr 9, 2009 at 1:36 AM, davada <davidlbarber@...> wrote:
> PCB industry glues the SMT components to the board before reflow. It's a
> pain in the butt if the components have to be replaced, but this how the problem is overcome. Finding the right cement is the key.
>
> David.
>

There is also a product by Creative Materials Inc called 122-38, a conductive epoxy which may work for this. Not sure if it will reflow but may be worth a check: Data sheet:http://server.creativematerials.com/search_results.html?cx=005626523669659259832%3Aggdtpniksno&cof=FORID%3A11&q=123-38#262
MJ  4088354979

--- In Homebrew_PCBs@yahoogroups.com, "davada" <davidlbarber@...> wrote:

>
> PCB industry glues the SMT components to the board before reflow. It's a  
> pain in the butt if the components have to be replaced, but this how the problem is overcome. Finding the right cement is the key.
> 
> David.
> 
> 
> 
> --- In Homebrew_PCBs@yahoogroups.com, "Jim Lewis" <jim.lewis@> wrote:
> >
> > I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
> >
>

This is thermally conductive epoxy, not electrically conductive.

I'm not sure how either would help with the moisture problem?

ST

On Thu, Apr 9, 2009 at 8:46 PM, michaelj22@...
<michaelj22@...> wrote:

> There is also a product by Creative Materials Inc called 122-38, a conductive epoxy which may work for this. Not sure if it will reflow but may be worth a check: Data sheet:http://server.creativematerials.com/search_results.html?cx=005626523669659259832%3Aggdtpniksno&cof=FORID%3A11&q=123-38#262
> MJ  4088354979
>

You might want to give Cash Olsen's embossing gun technique a try. His site has gone down now, but the basic idea is to use the hot plate to preheat the components to only 212 (100C) and then blast the leads with an embossing hot air gun for just a few seconds. Long enough for the solder paste to reflow.

I've had an 0805 cap or two pop up on one end while doing this technique and have actually been able to push them back down via the air pressure of the gun. An IC would be forced towards the PCB even more owing to its greater package surface area to lead surface area ratio.

Just a thought.

Regards,
Royce

--- In Homebrew_PCBs@yahoogroups.com, "Jim Lewis" <jim.lewis@...> wrote:

>
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
>

When i had "jumping chips" it was so bad they would pop right off the
board all the way across the table. This was when i tried to dilute
old no-good solder paste.
So having seen that, i don't think anything can be done once it happens.

You have to get out the moisture. if the parts are the cause, bake
them. If the paste is the cause, dispose of it and get new paste. I
expect it is the paste unless the components are stored in a wet
location.

It is very frustrating indeed, you have just painstakingly stuffed a
board, and then the buggers just hop off one after the other undoing
all your work and there is pretty much nothing you can do except
watch... That'll really ruin your day especially if you have to order
new solder paste (which in effect means even more painstaking hand
soldering with the iron for now).

ST

On Fri, Apr 10, 2009 at 6:08 PM, roycepipkins <royce.pipkins@...> wrote:
> You might want to give Cash Olsen's embossing gun technique a try. His site has gone down now, but the basic idea is to use the hot plate to preheat the components to only 212 (100C) and then blast the leads with an embossing hot air gun for just a few seconds. Long enough for the solder paste to reflow.
>
> I've had an 0805 cap or two pop up on one end while doing this technique and have actually been able to push them back down via the air pressure of the gun. An IC would be forced towards the PCB even more owing to its greater package surface area to lead surface area ratio.
>
> Just a thought.
>
> Regards,
> Royce

--- In Homebrew_PCBs@yahoogroups.com, "Jim Lewis" <jim.lewis@...> wrote:
>
> I have a problem with SMT chips jumping off the board during reflow. I preheat for 2.5 minutes 1/4" above the hot plate and then reflow 1.5 minutes directly on the hot plate at 375F with a cover. Paste is from Stencils Unlimited. Any solutions?
> -------------------------------------------------------
When I worked in the telecomms industry and we switched to SMT boards we had the same problem at first. It was at a very low level(<1% of boards) but when you are making >100k SMT boards a week it becomes a big problem!
At first we thought it was the reflow process, then the paste, but after we installed a high speed camera in the reflow oven and put our best material scientist brains to work we found out it was the homogeneity of the solder paste.
The volatiles were 'clumping' especially if it had been sitting around or in storage for a while.
The solution was to keep the paste "stirred" - continually mixed until just before use.  Also, the paste was carefully analysed in the lab for the correct consistency.
If I remember correctly the volatiles that make the solder a paste boil off. If the get too big in size(aggrigate) they explode and blow the chips all over the place.

My suggestion to you is to take your paste and remix it as best you can and see if this helps your problem.

Stephan is correct, we only ever glued when the board was put through reflow for mixed technology and SMT components on the reflow side.

Surface tension of the paste when it melts properly (controlled spike in the oven)is sufficient to hold and correctly align the components.

Components all out of alignment usually indicates poor temperature control through the oven or contaminated parts.

Hope this helps solve your problem. It is much more difficult to control all the variables for high quality and reliability in a "home bre" environment.

Regards, Geoff.