The chip that I plan to use in my design has an exposed pad on the bottom that must be soldered.
I will be hand soldering my board and will not have access to hot air (I think). I found a few videos of some being able to solder the pad by applying heat through the backside of the board, although it has to be configured a specific way. (https://www.youtube.com/watch?v=d-f-SBC0GrU)
What are my options here and how should this be approached?
This is an example of an SOIC with the similar bottom pad.
You can also use a via with a large enough hole to allow solder from the bottom to flow through to the pad.
This is my standard practice for boards which I will hand solder. Works well. I did not know anyone else did it! I guess I did not invent it. 
When I solder, I first use a flux pen (this is like a “magic marker”) to apply a thin coat of flux to the EP on the IC, then solder the IC pins to the board, then last solder the EP through the hole.
My first choice would also be to make a big THT hole through which I can poke my soldering ireon and then solder the “exposed pad (PIN 17)” last.
This method has the disadvantage that you may move the IC as it’s likely to get so hot as to melt the solder on all other pins while doing that.
Mike Harrison (youtube link in first post) is a very experienced electronics guy and his method of using a bunch of thermal vias to solder the pads and all pins from the underside in one go looks promising.
Could not find soldering info at the link. Just to say I have soldered the EP through a THT on maybe 10-12 boards and never had any issues with melting the solder to the other pins. I think my hole diameter is typically 1.0 to 1.3 mm, usually constrained by small IC dimensions. I will almost always use a solid ground plane connection to the hole without thermal relief spokes. (Thermal relief works against the goal of heatsinking the IC to the ground plane.) The one resulting issue is that high heat is required through a small soldering tip so I need high tip temperature such as maybe 425 degrees C. But generally when the solder eventually “takes” to the EP it “sucks in” suddenly. That is the time to take the soldering iron away.
I wanted to add one bit of wisdom which I learned from others. The (degree to which small amounts of flux is your friend) is surprising when hand soldering tight lead pitches or any IC. We all expect that the flux helps the solder to flow where you want it. It does that. But it also helps the solder to flow away from where you do not want it. So you get a result which is better altogether. The flux core in solder does not work as well as liquid flux pre-applied to the pads. Just beware that flux and dirt can sometimes cause boards to electrically misbehave in strange ways, so cleaning with alcohol is often needed.
I had, at best, mediocre results with variations of this approach. The really insidious thing is that I couldn’t tell the solder joint was inadequate until the IC failed after several minutes of operation. Using solder paste seemed to work better than wire solder. (Though it could be that I accumulated practice and skill with the wire solder before trying the solder paste.)
I haven’t tried “one large via”, but I DID use an array of normal-sized vias, connecting to a copper zone on the back side. That allowed me to bring the combined heat of TWO soldering irons to bear on the joint, while using solder paste on the front-side pad. As I recall, there weren’t any IC failures but the process is messy and tedious.
The solution that worked the best for me is to reflow solder the thermal pad joint, as it was designed for. My table-top reflow oven is an electric skillet. (About US$10 from a second-hand store.) Others have created effective DIY reflow ovens from table-top convection ovens. Search the web (or this forum) for more information about these projects.
Dale
The other thought I had is that this method can only work with IC packages with pins on two sides such as a power MSOP; not packages with pins on 4 sides such as an LFCSP. I cannot remember seeing packages with pins on 3 sides although it is certainly possible!
Regarding solder paste, I was always intimidated by the short shelf life (1 year in a refrigerator?). That is not practical for my situation. Is that overly pessimistic? I have learned not to leave solder flux (or myself!) in my lab during summer as the temperature can get to 50 degrees C but I have occasionally forgotten about the flux. I think one such oversight would destroy solder paste?
That sounds good if you remember to get the ICs soldered before anything else. I try to do that anyway. But it is no fun to do things as they were designed?? That rates on the bottom, like reading instructions. 
The shelf-life spec was probably developed for automated processes where everything must be kept constant and uniform, in order to get reliable and repeatable results with the machinery moving as quickly as possible.
As best I can tell, the primary mechanism for solder paste deterioration is evaporation of the solvent used as the flux carrier. Solder paste (Kester EP256 from CML Supply) stored in my desk drawer for up to two years was somewhat thicker and less spreadable than factory-fresh material, but still usable in my manual soldering process. I just had to re-calibrate my wrist as I operated the spreader.
And I have successfully rejuvenated old, thick, solder paste by squirting some liquid flux into it and mixing until it looks “about right”. (Make sure it’s true ROSIN flux, not the zinc chloride stuff from a hardware store plumbing department)
My solution - and maybe this answer is too obvious - is to solder the entire board at once. Yeah, it’s a bit tedious to apply paste to every pad, then place components, but still practical for one or two copies of a small board. But with stencils available for US$10 - US$20 apiece it’s cost effective to stencil-print solder paste onto short runs of relatively small boards.
At my age, I guess my sense of adventure is gone.
Dale
The solder/rosin manufacturers have just the right thing for rejuvenating paste (and old bottles of liquid flux that have turned to syrup): Flux Thinner.
Granted my quick check of two of the Kester options seems to indicate that the smallest increment it is sold in is a gallon jug (you can get a 53 gallon barrel if you want, and I’m sure if you contact them directly the would sell you a rail car full)… I guess they only expect industry to be interested in this product. For the home-gamer adding a compatible flux to old paste is “good enough” (even though this adds extra flux solids instead of just replenishing the volatiles). 
I would also try to make sure I’m replenishing with the same type of flux, instead of mixing in no-clean liquid flux to water-soluble flux solder paste.
On a few occasions I have had a nasty surprise and found that the solder paste that I thought was good has turned a bit crumbly and is more difficult than it should be to stencil with. Adding a few drops of liquid flux seems to work - but if it is so dried up that it is solid and lumpy I really wouldn’t trust it on anything important or fine pitched.
Adafruit sells small tubes of solder paste for USD$7, so at that size and price, short shelf life isn’t as much of an issue.
As this thread drifted to solder paste:
As a hobby I’ve been thinking every now and then of building an SMT oven and buying some solder paste “just to experiment with”, and therefore did some reading.
It seems that Henkel / Loctite GC10 is the paste with the longest shelf life. It’s officially “more than a year”.
On different forums (among which eevblog) I’ve read multiple remarks that this is one of the best quality solder pastes available.
The only solder paste I’ve personal experience with is with the “plumbers type”. This paste is meant for tinning steel plate and other large surfaces. For steel plate it has a pretty aggressive flux, and this paste is not fit for electronics.
I was able to lengthen the shelf life of my pot of solder past to a few years. Eventually the suspension broke down. A hard crust of solder particles on the bottom of the pot, and a layer thick viscous liquid on top.
I was able to use it still, by breaking up the crust and mixing the stuff, but after that the solder quickly sank to the bottom again.
It was not a biggie, because my application was not critical.
I mention it here only because it is a different method of solder paste going bad which had not been mentioned.
I would need a few trainloads every week…(not)
The small syringes of solder paste (10 - 15 grams) can be OK for placing and soldering components one-by-one , rework and repairs, etc. For stenciling, the syringes are impractical. (My stencil technique leaves several times as much unused solder paste on the stencil, compared to what gets deposited on the pads. The unused paste needs to be returned to the container, which is very difficult to do with a syringe.)
I have difficulty getting a small enough amount of solder paste out of a syringe, so I end up squeezing a short line of paste onto a piece of cardstock then putting it onto the board with toothpicks and sewing needles. (An SMT solder joint takes less solder paste than you expect, and the extra material can become a bridge between pins or traces - sometimes underneath the body of a component where it’s difficult to see.)
Unless you use it every week or two, the paste tends to dry out and clog the fine nozzle tips of syringes. And, once the paste dries out in the body of the syringe, it is almost impossible to rejuvenate it.
Dale
I could only find that material in 500g containers - can you point me to a source for smaller quantities?
I have been using a 150g jar of solder paste for almost 2-1/2 years - probably 150 - 200 boards - and its only about 30% - 40% gone. With a 500g jar, I’d have to re-write my will and decide which grandkid should inherit it!
Dale
Had a peek at Ebay…
USD 140 for a 500g container. Oops.
USD 55 for a pre-owned container with Mfg. date in 2018. Euhm, no thanks.
The only sensible way to get this stuff seems to be a group buy, and re-distribute in smaller containers.
It could be that the manufacturer is simply not interested in small scale distribution, or it’s because those larger volumes increase shelf life. (Less evaporationi because of volume/area and thick walls of the jar.)
And just to be clear:
I never used the stuff myself. It just kept popping up with people raving about it.
This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.
