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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


PCB Warping.

Grant Petty

#16540

PCB Warping. | 16 March, 1998

We have just go another batch of PCB's from our loaders, and some of the boards are severely warped. The boards are PCI cards and one of the boards is 9mm out at one end. The PCB is 210mm long and 4 layers. I am wondering if anyone knows what could cause this. Is it a PCB manufacturing problem, perhaps with the inner layers that causes the PCB to warp when in the reflow oven, or is it bad handling in the reflow oven like too long at too high a temperature. There is no yellowing of the silk screen overlay though. Also we have a few chips and decoupling capacitors on the back of the board, so I think they run them through the oven on each side. I hope anyone has some advice about this!

reply »

Justin Medernach

#16544

Re: PCB Warping. | 16 March, 1998

| We have just go another batch of PCB's from our loaders, and some of the boards are severely warped. | The boards are PCI cards and one of the boards is 9mm out at one end. The PCB is 210mm long and 4 layers. | I am wondering if anyone knows what could cause this. Is it a PCB manufacturing problem, perhaps with the inner layers that causes the PCB to warp when in the reflow oven, or is it bad handling in the reflow oven like too long | at too high a temperature. | There is no yellowing of the silk screen overlay though. Also we have a few chips and decoupling capacitors on the back of the board, so I think they run them through the oven on each side. | I hope anyone has some advice about this! Grant, Check out Earl Moon's posts further on down the list in this forum. He has some good things to say about PWB warpage and where it's incurred. Regards, Justin Medernach Flextronics Int'l PIC East

reply »

Earl Moon

#16543

Re: PCB Warping. | 16 March, 1998

| We have just go another batch of PCB's from our loaders, and some of the boards are severely warped. | The boards are PCI cards and one of the boards is 9mm out at one end. The PCB is 210mm long and 4 layers. | I am wondering if anyone knows what could cause this. Is it a PCB manufacturing problem, perhaps with the inner layers that causes the PCB to warp when in the reflow oven, or is it bad handling in the reflow oven like too long | at too high a temperature. | There is no yellowing of the silk screen overlay though. Also we have a few chips and decoupling capacitors on the back of the board, so I think they run them through the oven on each side. | I hope anyone has some advice about this! This is old stuff that keeps repeating. Contact me outside the forum and I will detail.

reply »

ETS, LLC

#16541

Re: PCB Warping. | 16 March, 1998

Problem occurs because the glass transition temp for FR4 is lower than reflow temperature. Our company wrote an article which was printed in SMT magazine some years ago that deals with this exact subject. Best way to deal with this is to use a center support conveyor in addition to the edge conveyor. Perhaps your current reflow supplier can retrofit one or perhaps your company is willing to purchase a new oven. If a new purchase is acceptable I would love to compete for your business. ETS is the Manufacturer of Universal Instruments Reflow Systems, the CUREFLOW. | We have just go another batch of PCB's from our loaders, and some of the boards are severely warped. | The boards are PCI cards and one of the boards is 9mm out at one end. The PCB is 210mm long and 4 layers. | I am wondering if anyone knows what could cause this. Is it a PCB manufacturing problem, perhaps with the inner layers that causes the PCB to warp when in the reflow oven, or is it bad handling in the reflow oven like too long | at too high a temperature. | There is no yellowing of the silk screen overlay though. Also we have a few chips and decoupling capacitors on the back of the board, so I think they run them through the oven on each side. | I hope anyone has some advice about this!

reply »

Earl Moon

#16542

Re: PCB Warping. | 18 March, 1998

| Problem occurs because the glass transition temp for FR4 is lower than reflow temperature. Our company wrote an article which was printed in SMT magazine some years ago | that deals with this exact subject. Best way to deal with this is to use a center support conveyor in addition to the edge conveyor. Perhaps your current reflow supplier | can retrofit one or perhaps your company is willing to purchase a new oven. If a new purchase is acceptable I would love to compete for your business. ETS is the Manufacturer | of Universal Instruments Reflow Systems, the CUREFLOW. | | We have just go another batch of PCB's from our loaders, and some of the boards are severely warped. | | The boards are PCI cards and one of the boards is 9mm out at one end. The PCB is 210mm long and 4 layers. | | I am wondering if anyone knows what could cause this. Is it a PCB manufacturing problem, perhaps with the inner layers that causes the PCB to warp when in the reflow oven, or is it bad handling in the reflow oven like too long | | at too high a temperature. | | There is no yellowing of the silk screen overlay though. Also we have a few chips and decoupling capacitors on the back of the board, so I think they run them through the oven on each side. | | I hope anyone has some advice about this! I agree the glass transition temperature (Tg) of FR-4 ranges from 120-180 degrees C., as examples, depending on material resin formulation. This, obviously, is below soldering or reflow temperatures as are most other resin systems but does vary up to 270 degrees C. for polymide material. It should be noted that Tg may, and should, be verified using thermomechanical analysis (TMA) and/or differential scanning calorimetry (DSC) to determine material "cure" and the physical changes board constructions undergo - by board shops and laminate suppliers - on a timely basis. It also should be noted, this temperature is only exceeded for short periods during soldering and touch up operations. It is one of the reasons repeated and extended thermal excursions (poor process management due to equipment, tools, operators, solders, etc.) should be minimized. Some of this is brings up other issues. What is important is laminate integrity (the ability of the board to stay together) and dimensional stability, including warp and twist, based on board constructions selected begining at the design level. Certainly, fabricators must be capable of understanding material characteristics, selections, and relamination process management to lessen problems associated with these factors. Often they do not as many are not qualified to produce what is specified. All this means, dimensional stability will be maximized (warpage will be minimized) with effective design and fabrication process management wherein carefully specified material and construction requirements are met. Fixturing large board types is useful and important when the Tg is exceeded for long periods exceedin Tg - allowing warpage (reversing the effects of well managed relamination processing). However, dimensional stability is "built" (laminated) into the board begining, again, at the design level. Earl Moon

reply »

Facility Closure

Reflow Oven