Reflow parts launching.

[itsmoked]

After my hideous oven runaway I've got things back to where they were with a new ramp and soak controller that plots on the screen and all that stuff. Fine.

What isn't fine is that the same old problem is back. A lot of boards I do are about 4 x 2 or 4 x 3". I use 0603 parts wherever I can but naturally I have to use 0805 and 1206 etc for either power dissipation needs or cost (big C values in small packages are expensive).

These bigger than 0603 parts have a penchant for bailing out. It's been so consistent that the policy (grating as it is) is to not even bother placing 1206 or larger parts because almost all of them fly. Now I'm seeing a lot of 0805 parts fly too, probably about %50. Hunting down the flyers and relocating them is painfully time consuming.

Looked at a board today thru a microscope and had the feeling I was looking at reconnaissance photos after an extensive bombing run thru a small town.

First crack at trying to logic this, it would seem like the paste has absorbed water that is flashing to steam and launching victims. So I changed the profile to soak at 50C for a while before proceeding. No real improvement was seen.

I've heard to fix this you need to "sprint from the start really fast to the peak". I haven't tried this because it seems like it would increase a rapid vaporization launching problem. On the last 10 boards I ran 3 boards, 3 boards, 2 boards, and 2 boards, and there was a marked reduction in flyers, perhaps by half on the 2 board runs. Not an insubstantial improvement! The difference would be that the temperature can probably move a bit faster on 2 board runs so maybe faster temp ramping has something to it.

Of course when you get down to tiny things that are overwhelmed by forces like capillary action it can all get 'unexpected'.

I use [URL unfurl="true"]http://www.chipquik.com/datasheets/SMD4300SNL10.pdf[/url]
paste.

We often take three days to complete placement so the paste is out for a while, however we don't really see a launched part difference between 5 hours and 72 hours.

Does anyone have suggestions.


Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

Soak at 110C ? 102C ?

You've got some kind of little roller conveyor in there, right?
Make sure the sections are aligned and the rollers are rolling, so the boards don't get jostled.

... out of ideas.


Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

Thanks Mike. No conveyor just a turbocharged modded toaster oven so no motion issue.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

Turbocharged?

As in, maybe enough airflow velocity over the boards to lift a component, or propel it sideways on a slippery pool of molten solder?


Mike Halloran
Pembroke Pines, FL, USA

 

[MacGyverS2000]

Do you have unpackaged components out for extended periods of time? Are you drying your components before use? Every single one of them should have instructions on drying temp/time. Typically it's something like heating at 60C for 24 hours, or similar, before going through reflow.

Are the components truly "popping" off, or are they merely tombstoning? Popping off would mean trapped moisture (as in, inside the component)... any moisture in the paste would slowly evaporate and dissipate rather than explode. Tombstoning would mean improper laydown of the paste or incorrectly designed pads.

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

Gotta post some pictures, Keith. Your verbal description begs for a visual confirmation.

TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[waross]

How close is your profile to the profile suggested in the Data Sheet?
What is the relative humidity in your shop?
Have you contacted the folks at Chip Quik?
They may know exactly what the cause is.
How are you with this advice from the Data Sheet?
Refrigerate at 3-8°C (37-46°F).
Do not freeze. Allow 4 hours for solder paste to reach an operating temperature of 20-25°C (68-77°F) before use.
How is your temperature controlled?
Is your air flow uniform?
Are you pre-cleaning your boards?
How about some one board test runs with shortened and extended ramp times?
Is it possible that you are getting temperature over-shooting that is not seen by you instrumentation due to thermal lag of your instrumentation?
Look at the whole process, not just the flux.
Good luck!

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[btrueblood]

Are the toaster oven filament wires exposed, or are they embedded in a metal sheath? If exposed, any chance they could be inducing voltages in the board traces?

 

[mcgyvr]

"popcorning" of parts is most often moisture problems..

Are these parts marked with a MSL level?
Obviously the "professionals" have baking/storage/dessicant procedures,etc.. that must be maintained so as a "toaster reflow" user that kind of handling/baking procedures are just as important..

even the flux..
Are you storing in a fridge?
If so are you allowing 4 or so hours after the paste is removed from the fridge,etc..

I'd try baking all the parts on your next run just to check.. Baking info should be in J-STD-033 I think..

 

[itsmoked]

Dan; Not moisture in the parts. We're talking 0805 components. There is no where for the moisture to 'hide'. No problems with leaded parts that certainly could harbor moisture.

Yes, the components are LEAVING. No tombstoning. The parts end up inches from their placements. Often remaining in the oven or landing on other boards.


IR; Pictures won't help as you'd see just an area with missing 0805 components in a neighborhood of intact 0603 parts. Like all the apartments got hit. :/

Bill;
Humidity ~ typically 60%
Paste generally kept at about 60F sealed. It's a syringe.
Toaster oven - no airflow at all, so yes, it's uniform.
No not precleaned. Boards are generally brand spanking new out of desiccant packaged bags. But it doesn't change the results if the board is ancient dirty laying around fingerprinted handled.
Process is slow enough that a small diameter T/C can run rings around it.

btrue; Whoa, that's out-of-the-box thinking. There are two radiant rods across the bottom just outside the area the boards span. They are shielded with metal U-channels with a hundred slits in them and are about an inch below the boards. There's nothing really magnetic about the subject ceramic based components though. Could be inducing voltages though most boards these included are almost completely covered with solid ground planes.

mcgyvr; Generally only the IC plastic-molded-onto-leadframes type parts seem to have industry "moisture issues" and my (possibly wrong) understanding is that the entire issue is about those parts exploding due to trapped steam, which of course, is not the issue here. All the leaded parts behave impeccably.

The paste is normally kept in a thermally massive insulated container at floor level in the work space, hence the 60F, unless it won't be needed for more than a week at which time it is refrigerated. There has been no detectable difference between right out of the overnight shipped cold bag verse paste that has sat out for 4 summer months. Same launching either way. Old paste will however cause an increase in more subtle aspects like minor tombstoning will occur. Perhaps 3 components out of 500 instead of 1 part in 500.

Thanks all for your ideas.
I'm going to run a board today. I'll let you know the results.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[IRstuff]

The level of humidity seems high to me. We're used to a much lower RH, like 30-ish, and we're always running in full ESD protection mode, but we don't reflow boards in-house.

TTFN
I can do absolutely anything. I'm an expert!
faq731-376 forum1529

 

[Compositepro]

It seems you just need to dry your loaded boards before you go through reflow. Drying can be a surprisingly slow process sometimes but heating to over 110C will speed things considerably.

 

[itsmoked]

Okay! So, because today's run was a single new proto board with a truckload of 0805s on it I decided to not change the temp profile. The board was stuffed in an exceptionally short 1.5 hours and immediately baked, this so I could bring home my enduring case of pneumonia and stay collapsed at the homestead instead of returning for a late nite finish.

I used a very bright flashlight and a lens out of a (retired) bench magnifier and stood in front of the oven door focused on the rows and rows of 0805s. Glancing to the left out the door I could see the large temp display on my office comp. I watched expecting problems around the steam temp 100C. Then 120C. Then 140C 180 200 220 225!! Nada. Not one single flyer. Clearly a very short screen time is effective. In the past probably the quickest screentime was 4 hours, which still gets you flyers.

Seems to me we're seeing moisture absorption. So I will try a very extended drying dwell in the future to see if multi-day screens can be tamed. It'll have to be once I'm well though. :/ hack hack wheeze (It sounds like I've inhaled a harmonica)



Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Seems to me we're seeing moisture absorption.

I agree.

So I will try a very extended drying dwell in the future to see if multi-day screens can be tamed.

That sounds like a good idea.
Sorry about the wheezing. There's been a bug going around our area; congestion, headaches, body aches, I'm still OK but my 3 out of 5 of my family have been sick.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[flexoprinting]

It takes a long time to place and check those little guy's, I wonder if using a dehumidifier just in that room would help.

 

[itsmoked]

I hadn't thought of that..
The space is air leaky like a sieve because it's built as a steel structure out of corrugated steel. Even though my office is 'finished' I literally feel gusts of wind coming from window areas. This pays off with stinky events. LOL

However my placement table is almost a closed space and with very little work I could make it nearly (if not air tight) air dead. Then I could keep it loaded with desiccant when doing a run. I will figure out how to do this.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[MikeHalloran]

The smallest dehumidifier you can buy will probably cost less than a big bag of desiccant, and if you install it with a drain hose, is a lot less trouble to maintain.


Mike Halloran
Pembroke Pines, FL, USA

 

[itsmoked]

I hear you Mike but mechanicals usually have a lot of heat associated with them. It's also not as clear to me how I could plumb a mechanical into my assembly table. With desiccant I have an oven I can recycle it back thru.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[Compositepro]

Mike is referring to a room dehumidifier you can get at Home Depot. Get a nice one for $250-$300 that includes a pump so you can run a 1/4" tube out a window or to a drain and don't have to empty the bucket once or twice a day. If it warms the room too much in summer a window air-conditioner will take care of that, and you will be very comfortable all year. It would not be a bad idea to use a lab desiccator over night, just for the boards.

 

[itsmoked]

Yeah, I know what Mikes talking about as I actually have one in storage. Rolls around and nicely dehumidifies a room. But, as I mentioned using one to dehumidify a room with windows essentially permanently open on adjacent walls would certainly make my power bill sit up and take notice! I don't really desire the room being less than the normal humidity because it rapidly gets one into static damage regions that I now need only pay the most minor attention to.

I have a gallon of industrial compressed air desiccant doing nothing. I figured out today that I can place a typical plastic storage bin upside down over in-process boards sitting in the stuffing table on their static mat. I can set a tray of desiccant in with them. Not only should this prevent moisture absorption I expect it should draw out any absorbed moisture that's gone into the paste. When assembly recommences the plastic bin can be removed the desiccant placed in it and the lid installed to prevent the desiccant from 'doing the room' pointlessly. The desiccant turns red when saturated and can be baked back to electric-blue ready.

Keith Cress
kcress -

http://www.flaminsystems.com