Hello, my name is Bob Willis and welcome to Defect of the Month with WNIE TV. Here is another video plus a short article to add to your collection of training material. Hopefully helping your team understand process and product failure
Now I’d like to talk about temperature profiling and mounting thermocouples. Now if you’re interested, I have created a set of colour posters which cover this particular aspect, which are FREE and they make great training aids, or they’re useful for covering up the dirty marks on your office wall. If you’d like a copy to print out all you need to do is drop us an e-mail and request a copy
Examples of profiling posters for printing also available at smta.org
Mounting thermocouples on a printed circuit board you first need a board which is the same design you are going to be processing. Normally speaking it’s fairly easy to get an example printed circuit board. There are always scrap boards produced by the PCB manufacturer, so any of your customers should be able to get one of those. It’s going to be the same metal build, same size, dimension and all of the other characteristics are going to be pretty much the same. Just because it has etching faults or electrical issues that doesn’t really matter.
Next thing is some components well I would suggest that if you go to your own rework area you’ll always find examples of components which have the same lead configuration, mass, So you have something you can solder to the board for temperature profiles. I think that’s fairly straightforward it’s in your own hands. You must make sure your sales engineers insist that the customer provides the right information/samples so you can do a professional job when you’re running your temperature profiles and being audited
Couple of the authors example board assemblies used for profiling
There are lots of great temperature profiling systems out there, but I’m not going to get into that conversation about what’s best or what you really need. I’ve used most and I’ve worked with suppliers on seminars, production lines and training sessions. As a process engineers you should know pretty much all of them which work very well. Some are over sophisticated where it’s perhaps not necessary but that’s just my opinion.
Methods of mounting leads consist of Kapton tape, which I do not like, but it still is a very popular technique because it’s too easy. High temperature, orange tape, which holds the thermocouples to the board, but does it? I would suggest not. You will normally find the thermocouple bead lifts underneath the tape, so consequently you’re not getting exactly the information you want between the first and second run.
Examples of different thermocouple mounting positions and methods
Some suppliers offer metal tape, which has good adhesion characteristics to the printed circuit board. It generally doesn’t allow the thermocouple to lift easily from the printed circuit board. You can use adhesives, but I tend to find that most adhesives after a period of time allow the thermocouple to lift and they’re going to lift just at the point you’re coming through the last reflow zone in your reflow oven. You then have to start all over again, which is really, really annoying. I personally like to use high temperature solder, I know it’s old school and it’s got lead in it. However, it works and has successfully worked for me for many, many years, there are many independent studies like NPL which have illustrated that its the best technique. There are techniques similar to this, but still it’s the best technique for surfaces and between fine pitch leaded terminations
You can also mount thermocouples into plated through holes, I normally suggest still soldering them in place. They can be mounted easily and be robust less likely to come adrift or separate from the printed circuit board. The position for monitoring is dependent on the soldering process you are monitoring which can be reflow, selective and wave soldering, robotic soldering, rework, adhesive cure or possibly curing coatings. The key thing is the thermocouple has to make intimate contact with the surface you’re trying to measure and the volume of what you’re measuring has to be consistent with the product that you’re actually going to be using.
Poor thermocouple preparation which will produce poor profiles
If you use thermocouples and for some reason leads are twisted this will give you a very inconsistent reading. The two dissimilar metal leads are going to be reading wherever they’re making contact not on the bead, what you tend to get is a spiky result. You won’t get a nice pretty graph; you’ll get a nice spiky graph which won’t look very good. Just take a careful look at some of the graphs shown in technical presentations!!! I have also seen engineers take their profile data points into Excel and smooth out the bumps and spikes
OK but not perfect profile result
If you’re using point contact correctly, then all of the wires should be insulated right up to where the two wires are welded together to create a bead. Which is where you measure the thermocouple temperature on your printed circuit board. There’s lots of other good information out there about temperature profiling
I’ve used profilers on wave soldering, selective soldering, vapor phase, convection reflow, I’ve used it for setting up rework and repair stations and of course obviously Ball Grid Arrays (BGA). One of the points covered in the poster set is how you measure temperature underneath BGAs effectively. And again I’ll show you how to do that. All you need to do is request a copy of the poster set. You can download it FREE and that will give you suggestions how to successfully profile for rework and repair. All of my Bob Willis Inspection & Defect Guide poster sets are also available from smta.org
We have many Defect of The Month videos at WNIE TV plus online articles which we hope will help you solve your process and product failures. We have listed a small selection from over 100 plus videos below created over many years with NPL, IPC & WNIE TV.
BGA popcorning
Dendrite formation
Open solder joints
Solder skips
Coating bubbles
Ultrasonic damage
Missing components
Incomplete past print
Component cracking
Tombstone chips
Solder balls
Solder bead formation
Solder shorts
Sulphur corrosion
Crimp connection failures
Our defect list grows, one month at a time