Profiling board

One thing I have done in your situation is ask the customer if they have any older design boards that are similar in thermal mass or if they have any from proto work, etc.... Otherwise find/make a board of similar thermal mass as best you can.

reply »

Well it depends on what kind of oven you have. If you have a short oven, you are going to have a difficult time. You will probably end up having several different profiles if that is the case. We have 3 BTU TRS 212 furnaces. They are fairly long 8 zone furnaces. By having a long oven your are able to stabilize large mass differences across the boards quite easily. The reason being is that there is a lot of preheat, and everything has time to come upto temperature before you hit reflow temperatures. We had the same question "Can we use a common profile?" My answer was yes, and the way I proved it was to take a very thin .030" and a very thick .115" thick backplane board and profile them. I also took a 4x4 .100" thick copper slug and attached probes to it. The profiles were all very similar to one another.Ramp soak and peak were right into spec. The only difference was of course the cool down. I think that there was just an article in SMT or Circuit assy. magazine about this very issue. I believe they said that you should be profiling every board, you should so you can show it to your customers, but I do 100% feel that it is totally acceptible to run a common profile. I proved it. It is going to be dependent mainly on your ovens.Break out the ol' thermo couples and check it out !

Hope this sheds a little light.

reply »

yngwie,

are you getting a wave from the customer to say it's not your fault you fried the parts? Profiling's one of those things that is a pain to dobut it's one of the most important factors in the assembly processes. I've read and seen instances where 'standard' profiles are used for all of production but I've never been convinced. One theory is that you use the mass of the card (with all the components ) to select the appropriate profile, all well and good but what if your card is say 300mm by 300mm with very few layers and a handfull of components, but by coincidence has the same weight as a sma say 100mm square rogers based PCB with lot's of chips or stuff on it, do you use the same profile?, probably not since the rodgers will sink a heck more heat. Other things to consider is the style of component you placeing in term's of the IR effect, you could have 1 PCB with all the same style components that are black boddied then another with silvber / white bodied parts, different IR uptakes and depending on the oven your using that could be anything from 10 - 20deg C of a change. I think it's important that the customer or people asking for you to manufacture without a profile card be made aware of the downsides and the lack of control that they are introducing into your process. There are system's on the market that help simulate this and get round the problem, there's oen mentioned on the forum at the moment, have a read of that but you still really want to verify that what your needing is waht your getting on the board and at critical components.

Summary, personally not something I've want to do, but if orced I'd be making sure I had a bit of paper saying they understand the risks.

JohnW

reply »

There are instruments from ECD, KIC and Datapaq that are used in addition to classic profiling for oven thermal management and they also have some very efficient prediction software for profiling. I am not an advocate of common profiles. Most boards are unique and should be treated that way and have thier own profile. Soldering is the most critical process in your operation.

reply »

One thing you should do is characterize your reflow oven, run profile boards of various densities, say 2 gm/sq.in. to 10 gm/sq.in. Adjust the oven settings for these various boards to acheive your standard profile based on your solder paste manufactures suggested profile. Once you have a standard profile for all the different weight boards (the more boards the more linear your set up chart will be) you should be able to set up one or two standard profile oven settings and then vary the belt seed to acheive the proper profile. Then set up a chart that will allow your operators to weight the board fully populated just before it goes into the reflow oven, match the gm/sq.in. on the chart, and set the belt speed according to that chart. Works great.

reply »

Hi Yngwie Have you got a reflow oven yet. If not, forget about complicated profiling with test runs or reflow trackers. We've got a nice oven with integrated temperature-feedback system. You build any profile you want and the oven will adjust its heating elements during each and every soldering cycle. Our oven "Ecolsold 350 Superior" is ideal for small to medium runs with complex boards. The problem CyberWolf mentions with the short ovens we solved with a single chamber system. As we want to control the entire reflow process as measured on the board this is the only way of securing control over time, place and temp. Have a look at our website at http://www.reflowsystems.com and especially at the second article in http://www.reflowsystems.com/Press/PressReleases.pdf

If you already have a reflow oven and want to stick with it, good luck with tracking, testing ... Because in the end PeteC and JohnW are right: "Soldering is the most critical process in your operation."

Lothar

reply »

I have the same problem. I made 3 standard profiles on a 6-zone oven, but always advice to make (or at least verify) a profile. But if customers don't want to, I point at the risks, and choose a slightly higher profile to be on the safe side. If you are using IR oven instead of a Forced convection, I would advice always to verify.

reply »