Double sided reflow

| Hi , | I have not been able to find any good description on how to perform double sided reflow on PCBs.The best would of course be to avoid reflowing the bottom side again but in real life this is not possible. | So when double sided reflow soldering is necesary, it must be important to do it the right way. | First of all, it is necesary to be sure that the solder joints also during the second reflow is melted completely. But I have been told that it is recommended only to reach around 210degC at joints on bottom side, while the intermetalic layer should be avoided to grow to much during the second run. | Secondly I have been told that the time above 179degC should maximum be 180 seconds tatally for both reflow runs. | Is this correct and does anybody have comments to this isue? | Thanks! | Brian S. Bentzen Brian, I do double sided reflow on just about every assembly that passes through my facility. There are limiting factors to this process but you didn't address them up top. First of all, are you using 63/37 PB/SN solder? If so your eutectic point is 183 degrees C. I am not familiar with any process considerations for 179. The reason you want to limit your time over this temperature is because of component damage and second pass wetting. When a product is put through the thermal cycling gauntlet, it is subjected to various dwell times at -20 C and 100 C, typically. This stresses the overall assembly and points to any CTE mismatches. What really is not addressed by people is the thermal cycling that takes place in the assembly process. In essence, the product is cycled from 20 C to about 225C. That is significantly more rigorous than the aforementioned cycle. Because of assembly induced temperature cycles, PTVs under BGAs on thick board assemblies crack and fracture. Component functionality is sometimes compromised. Only six thermal cycles at these temperatures should be tolerated. Six may even cause failure. If we look at the typical assembly process with double sided reflow, there are two thermal cycles at reflow tempertures. Any rework performed will mean at least two more thermal cycles to reflow temperatures for each reworked component, one cycle to remove the component and one to replace it. That is four thermal cycles in a given reworked location. Those temperature cycles DO stress the assembly. The trick is to keep the rework process to a minimal. The only constraint from a process standpoint, barring component failure, is a high density double sided assembly with large ICs on both sides. Boards where surface tension of molten solder is not likely to suspend the component may require a special fixture to guard the bottom side against the convection heating current within your reflow oven. EMC global Technologies in Doylestown, PA provides this fixturing service. I have had good success with them in the past. Good luck and regards, justin medernach flextronics international

reply »

Dear Brian, I came across your question on the SMT Net and thought I better reply to this one. Perhaps you might remember me. My name is Brian Stumm and I work for ETS, Energy Technology Systems. I believe we met in Anaheim, CA during last years Nepcon show (perhaps it was one of your co-workers). Anyway ETS is the manufacturer of Reflow and Curing ovens for Surface Mount assembly. In your message you stated that avoiding a second reflow on the bottom side of a double sided board was impossible. That is not true. We at ETS have found a way to cool the bottom side of a board while reflowing the top side. With our Bottom Side Cooling feature we are actually able to create a temperature differential of over 35 degrees C from the top to bottom of the board and at the same time the uniformity for either top or bottom of the board is +/- 6 C. What this means is that while the top side of the board can reach 210 C maximum the bottom of the board will only see a 175 C maximum temperature and thus reflowing only one side of the board. Sincotron A/S is the distributor of the CUREFLOW System for Denmark and I am sure they would be happy to discuss this technology with you or feel free to correspond with me for more information about ETS's Bottom Side Cooling Feature. You can reach Sincotron at 45-86-217744 (voice). Sincerely, Brian Stumm ETS, LLC 3939 N. Freya St. Spokane, WA 99207 PH: 509-483-0900 FAX: 509-483-0331 E-MAIL: ets@eznet.com ON THE WORLD WIDE WEB: www.eznet.com/~ets Hi , | I have not been able to find any good description on how to perform double sided reflow on PCBs.The best would of course be to avoid reflowing the bottom side again but in real life this is not possible. | So when double sided reflow soldering is necesary, it must be important to do it the right way. | First of all, it is necesary to be sure that the solder joints also during the second reflow is melted completely. But I have been told that it is recommended only to reach around 210degC at joints on bottom side, while the intermetalic layer should be avoided to grow to much during the second run. | Secondly I have been told that the time above 179degC should maximum be 180 seconds tatally for both reflow runs. | Is this correct and does anybody have comments to this isue? | Thanks! | Brian S. Bentzen

reply »

Hi, I am from Celtronix Ltd. we are a provider of Odd Form assembly equipment and have gained a lot of experience in double side reflow, both the chemical side of it as well as the automatic insertion of all types of TH components in paste. Enclosed is an article on the subject I wrote. It may be usefull to you. If you have further question don't hasitate to contact me TEL | (972) 3 6485396. You are also invited to our Nepcon booth #621, where we will demonstrate the equipment and processes. best regards, Eyal Epstein | Hi , | I have not been able to find any good description on how to perform double sided reflow on PCBs.The best would of course be to avoid reflowing the bottom side again but in real life this is not possible. | So when double sided reflow soldering is necesary, it must be important to do it the right way. | First of all, it is necesary to be sure that the solder joints also during the second reflow is melted completely. But I have been told that it is recommended only to reach around 210degC at joints on bottom side, while the intermetalic layer should be avoided to grow to much during the second run. | Secondly I have been told that the time above 179degC should maximum be 180 seconds tatally for both reflow runs. | Is this correct and does anybody have comments to this isue? | Thanks! | Brian S. Bentzen

reply »

Dear Brian, Users tend to specify a req. for doublesided reflow with a cumulative dwell time above liquidious. I have seen values of max 60-90 secs for 1 one pass and 90-120 secs for the total of two passes. Intermetallics tend to be a problem when joints are kept above 183 for > 150 secs. On bilayer boards (Cu only at top and bottom) you can typically not reach differences top to bottom bigger then 35-40 C. For multilayers however, the maximum is 15-20 C. If you actively cool a multi- layer at the bottom during the 2nd pass, then a part of the joints will remelt and a part will not (no remelt will typically occur at joints which are interconnected to inner Cu layers or even the topside layer). Big companies like AT&T found this to be a reliability risk for joints as the joints which do not remelt get to take the stress/ forces which occur due to expansion differences etc. . This effect also occurs with wave soldering of boards where you already have SMT at the topside of the board.

regards, Frank J. de Klein

reply »