Hi, Shen Zhiwei

Share with you on this.

¡°If the ratio of component weight in grams divided by the total pad-mating area in square inches is less than 30, the components will not fall off.¡±

Actually,Many different methods have been used to avoid two-step reflow. For example, one method is to glue the bottom side components with adhesive so that both sides can be reflow soldered simultaneously. This method may save one reflow step, but it adds adhesive dispensing. Dispensing adhesive after paste printing can introduce some of its own problems.
Others have experimented with using solder pastes of different melting points for the top- and bottomsides to avoid having bottomside components fall off. Another method is to try to maintain different top- and bottomside temperatures by adjusting top and bottom panel temperatures or even blowing cold air on the bottomside while reflowing the topside. Just imagine the stress and potential warpage in the board. I do not think one has to make life so complicated. With some experience, it¡¯s easy to figure out which components can go on the bottom and which ones should go on the top. Zarrow and Belmonte1 derived a weight-of-component-to-pad-mating-area ratio to determine if components can be placed successfully on the bottomside. They established that if the ratio of component weight in grams divided by the total pad-mating area in equare inches is less than 30, the components will not fall off.
In a double-sided assembly, it is important to decide which side should House the larger components(PLCCs above 84 pins or smaller PLLCCs attached to heat sinks)that may fall off if the assembly is inverted. Obviously, larger or heavier components(those attached to heat sinks)should be placed on one side only, and this side should be reflow soldered last. There is an added benefit to this option: The smaller components mounted on the opposite side that are reflowed twice tend to self-align more during the second reflow cycle. Self-alignment is generally not seen in larger devices, with the exception of ball grid arrays (BGA). BGAs will self-align even if they are misplaced by up to 50 percent. It should also be noted that some very small components, such as ceramic crystals with only four leads, may be considered ¡°large¡± because they will fall off during a second reflow cycle if mounted on the bottomside.
Some double-sided assemblies containing only passive devices on the secondary side have both wave and reflow soldering options for that side. Wave, instead of reflow, soldering the secondary side is not and cheaper, but the former may be desirable for many reasons. For example, via holes can be fully filled only during wave soldering. The plugging of vias is necessary to achieve the vacuum required for automated test equipment(ATE) to work. If wave soldering is not used, the vias can be filled during reflow by printing solder paste over them, Screens or stencils can be ordered with this requirement in mind. Either option is acceptable, but a choice must be made at the design stage.


Best regards.
FP Deng
Celestica Dongguan
Operations- Manufacturing Engineering Department.
Tel:0769-83394581







"Shen, Zhiwei (GE Healthcare)" <[log in to unmask]>
Sent by: TGAsia <[log in to unmask]>

2007-12-29 09:22
Please respond to
Asia Committe Task Group Forum <[log in to unmask]>; Please respond to
"Shen, Zhiwei (GE Healthcare)" <[log in to unmask]>

To
[log in to unmask]
cc
Subject
[TGA] Need your help to find solution that big component drop when reflow





All
 
        Do you have experience that the big component drop when reflow?
        We have layout with big components on the bottom side, when we reflow the top side, the big component would drop or find nonwetting defect, would you like to share your experience on this?
 
    Thanks and regards!
    Jerry Shen
    GE Healthcare Wuxi
    Supplier Quality Engineering