Slope Calculation for Reflow Oven Profiles

Board Talk is presented by Phil Zarrow and Jim Hall of ITM Consulting. Process Troubleshooting, Failure Analysis, Process Audits, Process Set-up CEM Selection/Qualification, SMT Training/Seminars, Legal Disputes

Phil Zarrow With over 50 years experience in PCB assembly, Phil is one of the leading experts in SMT process failure analysis. He has vast experience in SMT equipment, materials and processes. Jim Hall A Lean Six-Sigma Master Blackbelt, Jim has a wealth of knowledge in soldering, thermal technology, equipment and process basics. He is a pioneer in the science of reflow.

Transcript

Phil
Welcome to Board Talk. This is Phil Zarro and Jim Hall of ITM Consulting, sometimes known to many in the industry as the Assembly Brothers, Pick and Place. We're here to talk about service-bound processes and equipment materials and procedures.

The question for today - I have a question concerning the gradient for reflow profiles. Is there any specification or empirical value for the time period the gradient is calculated in, one second, ten seconds, etc.. JEDEC specifies a ramp-up, but it extremely depends on the chosen time span.

Jim
It specifies three degrees per second.

Phil
We'll talk a little bit about how compatible that may or may not be, but the question is, basically, what's the right way here? How's it being calculated. Where is it?

Jim
He points out that, if he calculates it over a 10-second interval, he gets a gradient of 3.5 degrees per second, but if he calculates it over a 20-second interval, he gets a 2.8 degrees per second gradient.

Phil
This is important because a lot of things will be gradient- or slope-dependent. A lot of times we'll see, particularly with some of the lead-free solder pastes, where slope is extremely critical, and actually under that three-degree-centigrade-per-second, too. So, Jim, how should it be measured?

Jim
The first thing we have to note is that this comment comes from Germany and we have to note the use of the word "gradient". Gradient is not a word that we typically use in the United States, and it can often be confusing. Having worked for a Dutch company, we had this problem, and the word "gradient" in Europe typically means what we call a slope or ramp rate in this country. So, I will call it "slope", and it is the rate at which temperature changes over time.

This question is profound. It is one that we've dealt with for many, many years and it has never been answered successfully to my satisfaction. The idea is, when I'm calculating the slope over what interval, and typically we're talking about the minimum interval that we're looking at. And so, we get a specification for a solder paste or a component, where we're concerned about heating or cooling too fast or, like, warping a BGA or cracking capacitor or something like that. We have these slope limits that we have to stay within when we set up our profile, and this could apply to both reflow and wave solder.

So, we go to calculate at a slope and over what interval are we concerned about? This used to be a problem when the profiling tools such as Data Pack and Kick and Mole produced more sophisticated packages that calculated slope for you automatically, and they often had a feature that said, what is the maximum slope?

When they first came on, you would look at these and the maximum slope you set up a standard reflow profile and tell you that the maximum slope was 15 degrees per second. And you go and you look and it goes, "That's crazy. I don't understand that. How can that be?"

Well, when we dug into it, we found that the profile was sampling the temperature once a second and it was calculating slopes for every one-second interval throughout the profile. And so, what it found is that, due to inaccuracy of some things that we'll talk about later, that some of the slopes for one-second intervals were extremely high, and some of them were extremely low. So, obviously, that wasn't practical, but it raises the question, what interval should I calculate it over?

If I'm talking about a solder paste, let's take a simple example. I have the solder paste and I want to create a straight-ramp profile. I'm going to go from ambient all the way up to peak temperature at one continuous slope. That's my objective, and my solder-paste manufacturer says, "That's fine. The paste will work that way and the range is between 0.5 degrees and two degrees per second."

So, I set up my profile. How do I measure the slope? Do I measure it average slope from ambient from entering the oven and ambient temperature, 25 degrees up to 220 or 245 if I'm doing a lead-free profile? Just take those two temperatures, subtract the times from my profiling data and calculate the average slope, or do I need to look at smaller intervals?

Why do we have slope specifications? To protect materials and make sure they work right. So, I'm talking about a solder paste where the manufacturer's probably concerned that if I heat it too fast, and my slope is too high, I may boil the solvents and create solder balls and voids.

If I heat it too slow, I may dry out the solvents and use up the flux before I get to reflow. So, that's why they do it. But over what interval can this damage occur? Basically we're trying to prevent things from damage and enhance the proper chemical reactions.

Let's look at a classic example in real life. We have a convection oven we know that there are limits to how low a temperature I can set in the first zone that is, the zone that the board sees when it first enters the oven. And what happens is, if we measure short-term profiles within that zone, we will often see that the slope is above the maximum rated by the manufacturer.

Take my example. My maximum slope is 2 degrees per second. My board comes in at 25 degrees C. I may like to set my first zone at 75 degrees C,  but because of the thermodynamics of the oven thermal spill and so forth, parasitic heat transfer within the oven I can only set my first zone at 100 degrees.

So, what happens? The board comes into this zone. It's at 25 degrees. It gets hit by this 100-degree air and for at least the first part of the profile the slope is very high. Then, it levels off and the slope will drop off to its correct value in the next zone because I can set it at a reasonable temperature.

Should I be concerned about this short-term spike or steep part of the curve? And the answer is, I personally think so. I've never heard that data produced or discussed by any material supplier.

Likewise with BGA components we're concerned about heating or cooling too fast. I should point out another common place this exists in the oven is in the first cooling zone, where most cooling zones don't have temperature control.

So, if I'm coming out of a hot profile, particularly for lead-free, come out of the last zone, which may be operating at 260 or 270 or hotter. The board is at 240, it hits the first cooling zone, which may be at 100 degrees or less and if I look at the first few seconds or even longer than that, the slope may be steeper than the 3 or 4 degree maximum cooling slope allowed.

So, the answer to your question is, I don't have an answer. My personal feeling is that the only reasonable the minimum interval should be the time that the product is in one zone because that's all I can control. The minimum effect I can have on the profile is the length of one zone, so I typically use that as my minimum time base.

It's just the reality of what I can do with the oven or the wave machine, and nobody has ever provided better specs. I throw the question out. If anybody has any better information, please send it to us and we'll try to get it on the air.

It's been a topic talked about as long as I've been in reflow, which has been 30 years.

Phil
That's a hell of a long time. Longer than reflow itself has existed. Anyway, however you're reflowing, this is Phil Zarro and Jim Hall of ITM Consulting and Board Talk, saying whatever you do and whatever your rate of heating is, don't solder like my brother. And keep those kids away from the flux pot.

Comments