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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Wavesoldering SPC

Sal

#5063

Wavesoldering SPC | 26 January, 2001

I am currently in the Process of setting up some SPC software on our Wavesoldering equipment. What I would like to know for each Parameter that is listed below what kind of variation from my average setpoint would cause defects.Example. If the average for my Dwell time over the Lambda wave is 2.5 secs, would setting the limits to 1.5(LSL)AND 3.5 (USL)cause defects ?

Parameters.

Max Preheat Temp Max Preheat Slope Conveyor Speed Chip Wave Dwell Time/Immersion Length Solder Wave Dwell /Immersion Length Parallelism Contact Temp

Any Opinions or responses would be gladly appreciated.

SAL.

reply »

#5071

Wavesoldering SPC | 29 January, 2001

Hi Sal,

taking the dwell time as parameter to control seems to me unessary. Dwell time is a function of other settings that can be controlled (adjusted) more easily (conveyor speed, wavehight-settings, angle). Although dwell time is essential it�s IMO better to look at and record the settings that influence the dwell time.

Btw. 1,5 s seems a bit low to me, for what I�ve learned it takes a bit more time for a good solderjoint. But as we all know it depends on the product that you are running and may differ. so I think you may need a product dependend SPC chart.

Good luck

Wolfgang

reply »

C.K.

#5091

Wavesoldering SPC | 30 January, 2001

Wolfgang is correct - Dwell Time is influenced by conveyor speed and of course, pump speed. Contact Length (the length of contact between board and wave) is something that you might want to track. Contact Length greatly influences process defects like solder shorts - the faster you run your conveyor, the MORE contact you make, and for closely pitched through-hole parts, the more contact your making. Contact length is a function of both your dwell time and conveyor speed since it is a product of the two ( Contact Length = Dwell Time * Conveyor Speed).

I'd recommend still tracking your Dwell Time, and finding that optimal value for Contact Length, which has a great influence on process defects and peel off characteristics. A good rule of thumb for dwell time is anywhere from 1.5 to 3 seconds (the process window is very large), but again, it depends on what board you're running. Typically, for a board which has lots of closely pitch connectors, I run the board very slow (around 3 - 3.5 feet per minute), to minimize contact length and at the same time run the pump speeds pretty low so that the dwell time isn't too much either....

reply »

#5092

Wavesoldering SPC | 30 January, 2001

I thought Contact Length refered to the travel distance the board stays in contact with the solder. I measured it using a tempered glass plate that also showed me if I had an even footprint of contact with the wave.

By this definition, conveyor speed changes do not affect contact length, but wave height changes do. My experience is that heavier and/or thicker boards need more time in contact with the solder since it takes longer for the solder to get up through the pth. I preferred to increase my dwell time by slowing the conveyor, but then you have to check your preheat temp's to be sure you don't burn off your flux before the board gets to the wave.

Good luck, Mike

reply »

C.K.

#5124

Wavesoldering SPC | 1 February, 2001

You are correct, but mathematically, you're able to derive your contact length (the length of contact between your board and the solder), from your dwell time and conveyor speed.

=====> conveyor speed x dwell time = contact length

and

=====> inches / seconds x seconds = inches.

First run your pump at a constant speed.....THEN,.... try running your glass plate at a very fast speed, and then you'll notice that your length of contact will be MUCH greater - because at a faster belt speed, your board (or glass plate) will tend to grab more solder at once. Then, try to run at a very slow speed, and you'll notice that smaller contact is made between wave and glass plate. The slower the speed, the less solder your glass plate will contact because it's got more time to dwell (it's contacting smaller increments of solder per unti time).

Take measurements with your glass plate - dwell time, contact length, and verify mathematically, that this is correct.

Believe me, I've tried it before. So, this is valid both in theory and in practice.

reply »

Solder Paste Inspection

Software for SMT