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

Printed Circuit Board Assembly & PCB Design Forum

SMT electronics assembly manufacturing forum.


Reflow oven profiling - frequency ???

Views: 3591

d0min0

#49301

Reflow oven profiling - frequency ??? | 26 April, 2007

Hi there,

in my factory we use profiling every temp. change or once per day others make 1 profile for machine per week

what is safe in your oppinion to be sure that all products made are good :) or if 1 per week is enough how to prove it to bosses :)

regards

reply »

#49302

Reflow oven profiling - frequency ??? | 26 April, 2007

If your product mix is similar in terms of thermal mass and componentry, find some standard setpoints (by doing a baseline profile) which will produce a thermal profile that's within the specs. of your solder paste and product. This approach usually works well, and once you do a profile verification on any one of those products within that product family, only minor or no adjustments will need to be made. That way, you're confident that with any new recipe you put in your oven, you can just "copy over" from another recipe.

Now, obviously the ideal situation, if you have the manpower, is to have a thermal profile on file for every assembly your company produces.

At my company, I've asked for an intern or coop to do this, and was shot down!

reply »

#49303

Reflow oven profiling - frequency ??? | 26 April, 2007

Good afternoon,

We develop reflow profile ones and control it daily or when recipe changes after that. Modern reflow ovens are usually very stable. This means that your zone temperature set points won�t deviate from initial ones. You can calculate dwell time and peak temperature Cpk to make your boss relaxed.

BR, Pavel

reply »

#49304

Reflow oven profiling - frequency ??? | 26 April, 2007

Pavel is right...Your Quality Group, customers, and Management will really appreciate Cpk's. We do Cpk's annually, or whenever a major mechanical change (motor, blower, etc) is made.

Pavel is correct in saying that Cpk's on most modern convection ovens will exceed 1.33 (6-sigma quality). Time above liquidus and peak temperatures are the main variables to study.

Here's a good link: http://www.itl.nist.gov/div898/handbook/pmc/section1/pmc16.htm

reply »

#49306

Reflow oven profiling - frequency ??? | 26 April, 2007

Hi D0m0

Can you standardize on your temps and just adjust conveyor speeds for various sizes of the boards?

Another trick is to parallel t-couple plug off the actual t-couples of the oven and mount the plug in an easy access point of the oven. Then you can just plug in a "calibrated" hand held thermocouple unit to verify they are at the correct settings.

Educate your boss on your oven a little. Most ovens have set tolerances that will alarm if the set points waiver from their indented target. Or record the data on a daily basis and show that things don�t change.

Now understand that even if you have perfect oven settings, it doesn�t guarantee a good solder joint. Inspecting the solder joints periodically is the only way to assure the solder joints are correct.

reply »

#49367

Reflow oven profiling - frequency ??? | 27 April, 2007

Profiling is product specific. An instrument called the OvenRIDER from ECD is for soldering machine quality management. See link below.

This instrument measures various parameters of your oven to include thermal transfer efficiency. The concept is, you profile a product today, you run the OvenRIDER daily and you can SPC the measurement data to see trends and find any problems with the oven. This ensures that your oven, like a year after you did the profile, is the same as it was that day.

http://www.ecd.com/emfg/instruments/ovenrider/

There are other companies that make similar products, like:

http://www.datapaq.com/Home.htm?url=http://www.datapaq.com/surveyor/datapaq_surveyor.htm

http://www.techinfocorp.com/

Also, look at these:

http://www.ecd.com/emfg/instruments/ovenwatch/

http://www.kicthermal.com/products/

reply »


GEB

#50085

Reflow oven profiling - frequency ??? | 3 May, 2007

Hi,

We run a profile check each time we load a recipe. This ensures that the correct settings have been used for the job. It also gives traceability in the event that you need to prove whether or not the oven temperature was the cause for failed products. It only takes about 15mins, and is something which the line operators can carry out, so you just need to check the graph looks correct.

reply »

#50254

Reflow oven profiling - frequency ??? | 16 May, 2007

It always surprises me how many variations there are for profiling and controlling your ovens, and the thing is most of them are fairly valid. Here's yet another option: Your board profile needs to be specific, 2 boards can look identical but have very different thermal loads due to copper mass and internal tracking etc. Gte a profile for your board that your happy with in line with the paste and components specs. OK, next what your actually worrying about is if your over has variation in it or not. Checking every profile everytime will not actually give you a real answer to this, there is too much variation int he system. You need to organize yourself a test unit, Datapaq have a good one or you can design your own, I've inlaid thermocouples in a 400mm square CDM carrier before across the width of the oven... that's important, you need to understand the side to side delta in your machines. Once you have your test pallet simply run that once a week on a program of your choice called calibration or check or bob!.... just don't ever change those settings. Now all you care about is tracking the results of your initially weekly calibration, if it hasn't drifted then all of your other profiles int hat over are fine... don't wast time and resource (and hence money!) rechecking them all. This will also allow you to build a valid Cp & Cpk measurement for your oven over time to keep people happy.

John

reply »

#50276

Reflow oven profiling - frequency ??? | 17 May, 2007

We are currently using Oven Rider (ECD) to "verify" just the conduct of the peak temperature; we define which product is the higher runner for this line and then with the settings for this product (already profiled with thermal couples attached to the board) we pass through the oven rider. This have six thermal couples, 3 of them are attached to a defined mass (aluminium cylinders) and the others are exposed to the ambient so the air temperature is measured and also the rate of heating for this mass is measured.

We only consider peak temperature for the aluminium mass (which is lower than "profiled" peak temperature) and then we just plot this temperature on a chart every first article (with +/- 10% of tolerance) and keep tracking. I really don't know how they define 10% of tolerance, it is very subjective and I think is little to high. Based on this we calculate a cpk (more for quality reports than for a problem solving tool).

There's a document on ECD webpage explaining how they calculated "oven efficiency" using ovenrider; basically they calculate the H factor (convection heat coefficient)since it is able to have fluid temperature (air) and mass temperature; I tried to implement this instead of plot just the peak temperature but quited due to lack of time.

I think there's a little of misunderstanding about "profiling" you just profile once (well or one session of profiling iterations) and then you just track behavior of the oven to see how much variation on temperatures you have inside the oven.

Regards

reply »

SMT Machines

Reflow Oven