EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: Yaro on November 19, 2015, 01:41:50 pm
-
Hi all,
I've some question about reflowing. I've never used solder paste before and I wanted to try it with some PCB that I've recently ordered. I've builded a reflow station with a heatgun and a temperature sensor to try to follow reflow profile. My solder paste is with lead, Sn63Pb37 paste. MP of this paste is 185°C, I want to ask if the solder paste should melt at 185/190°C exactly or above. Becouse my paste melts at 205/215°C, but it turn back to solid at 180/185° exactly. My supposition was that temperature sensor is faster to change temperature and PCB takes more time to reach sensor temp.
Another doubt I've is final result. I've tryed 2 different brand paste and the result is the same(temperature and finish look). I've not mounted components to see how solder paste reflow. Here's some photo:
(http://i64.tinypic.com/bjbck3.png)
As you can see, there is some kind of solid clear material like wax but stronger. Using the first brand solder paste there is little more. It's more noticeable near big pads like thermals, and much less near small pads.
I want to ask, since solder on pads seems to cover well all pads and have shiny surface, is this normal? Should I just clean with isopropyl after reflow?
About reflow I let about 90-110Sec. for soaking and solder flux evaporation. Reflow profile for this paste advice 60-120Sec.
Thank you!
-
You are probably right about the temperature difference between the sensor and the PCB. During setup of soldering profile for a new PCB, the temperature is logged at several points (center, edge, small package, large package etc). Not only the temperature at the end of each heating zone is important, but also the rate of change 'C/sec
The transparent wax like stuff is flux residue. During visual inspection, the transparent residues can act as small lenses and distort the view, often to the worse. Clean or not? It depends on the type of flux and your application.
-
Relative area of copper trace leading from the pad being flowed has a lot to do with how long it will take. Typically you would run the oven hotter than the flow temp by quite a bit, but below maximum temp for the various parts. modern FR4 will begin to discolor at about the maximum temp for the parts. Also this is just above the temp where the copper will de laminate. Additionally on packages like the ones shown in your pictures, unused pads with no traces leading to them will de laminate at a significantly higher rate than ones with some copper attached. When a part is installed this will happen much less. Additionally you need to see what kind of flux you have in your solder paste. It may be water clean or RA flux that will require cleaning with a solvent of flux remover after assemble. There are also no clean fluxes. The remaining residue is relatively inert after flowing. I still will clean this type of flux anyways. It can turn a powdery white over time. Stick with the leaded solder. Its flow characteristics are much better for this kind of hand assembly and if you ever are in a situation where you need lead free, chances are you will have some quantity and can let a contract assembly house worry about it.
-
But this flux is right? I mean, this flux residual quantity is normal or there should be less or more flux? I've read that flux must evaporate in the soaking zone and seeing all this flux at the end seems strange.
-
I have seen some fluxes make an awful mess. Generally it does not evaporate with active fluxes. It states liquid at flow temperature.
-
Solder paste is a suspension ( a mixture not a solution ) of solderballs, flux and additional chemistry in a carrier medium.
Volatile chemistry evaporates, solids remain on the PCB. Reside if from a nocleanflux converts to a stable polymer after soldering process.
-
Oh, and there are also water clean or organic solders out there too.
-
I would say that the amount of residues is normal. A thin stencil and 50-60% coverage on the heat tabs would result in less residues and also lower the risk of short circuits for QFNs.
Parts of the flux will evaporate during preheat, but some of it will stay on as a protective layer against oxidation in the peak zone. If the flux cannot stand the heat, it can char and there is a risk of poor soldering quality, especially with peak temps for lead free solder. There are leaded solder pastes available that handles lead free profiles.
-
Thanks for your answers. Since there is "a lot" of flux in solder paste, I've just cleaned with isopropyl PCB board and then applied solder paste. It's correct procedure? I've read of someone that advice to apply flux before spread solder paste, but I've found that this flux create a small climb between stencil and pcb making solder paste to past pads edges and creating a bad result.
I've another question about peak temperatures. This is a table with peak values for SnPb solder paste. With 10/30Sec. adviced timing for critical applications.
(http://www.x-toaster.com/wp-content/gallery/articlesimages/dynamic/IR-Convection-Reflow-Profile-IPC-JEDEC-J-STD-020C-_-leaded3.jpg-nggid0230-ngg0dyn-0x0x100-00f0w010c010r110f110r010t010.jpg)
What I understand with this image, if I'm correct, is that peak temperature is not for solder paste but for package safty.
Considering that SMD packages unlikely exceed 2.5mm height and 350mm3 (the biggest one I've seen for MCU 100-lead LQFP with 400mm3 16*16*1.6mm, as hobbyist). And that most packages are lead-free with 260° peak temperature. Can I safty use 235/245° C as peak? (in addition that my sensor have always a temperature a little higher than PCB temperature).
-
235-240'C peak temp should be ok. Murata usually mention in their datasheets that you have to check the solderings if a "low temp profile", i.e. leaded solder is used. The reason is the Tin plating on the components terminals will melt at 232'C.