Category: Guides

other things: temperature indicating crayon (watching the solder melt works, too), vacuum pickup tool for placing parts (tweezers and fingers will get the job done, too), thermocouple to measure temperature. A couple companies sell toaster oven controllers–some of which you can simply plug the toaster straight into. These are nice for being able to setup a temperature profile and log temperatures, but in our experience the ovens don’t heat up fast enough to require any throttling by a controller. They would probably be more useful for complex boards with many layers, large ground planes or ones with BGAs.

solder paste: Solder paste consists of small balls of solder floating within a gel-like flux. It’s an interesting animal that must be kept refrigerated to keep the flux from drying out. Also, the flux will slowly remove oxides from the metal–which is its job–but you don’t want it to get used up before the actual soldering operation; refrigeration also slows this process down. Some pastes have longer out-of-refrigeration shelf times than others, or may not require any at all–see the manufacturer’s recommendation.

Also note that the recommendations provided by manufacturers are geared towards industrial manufacturing with screen printing stencils and robotic dispensers. The industrial processes are designed to produce thousands of joints with minimal mistakes, so any small change in the paste rheology (how it acts under pressure) that causes an occasional defect would be unacceptable. But an occasional problem with a prototype can easily be touched up, so following the guidelines strictly may not be all that necessary.

Possible defects from dried out paste include slumping (the paste spreads out and causes bridges), solder balling, and clogging of stencils, which leads to insufficient solder coverage. Solder balling is when you find lots of tiny balls on your board that may come free and cause a short some day.

If you’re going to use a stencil, get paste in a jar instead of paste from a syringe. The syringe paste has slightly less metal to help it flow through a needle, whereas the jar paste is designed to work well with stencils. Many distributors will make you buy a plunger and needle to go along with the syringe. Digikey®, for instance, doesn’t even sell those, requires 2nd day shipping, and charges an obscene amount of money. There are lots of paste distributors; a little googling will save a lot of money. As for a needle size, I’ll recommend 22 gauge as a good starting place. You can control how much paste comes out by moving the needle over the pins faster or slower, too.

One last recommendation: get no-clean paste unless you have reason to believe your parts are very difficult to solder, and only then get water soluble paste. If you do use water soluble paste, the residues are conductive and corrosive, so be sure to clean them off with warm water.

The first side: including a 208 pin, .5mm pitch QFP

We’re going to attach components to both sides. The components on the first side will be held on by surface tension when the 2nd side bakes. Do the side with lighter components first to reduce the risk that any will fall off. PLCCs are rather heavy compared to the number of connections, so we’ll do the side without one first.

apply the solder paste: For fine pitch components, say .8mm and smaller pitch, just lay a bead of paste across the entire row. When it reflows, the paste will cling to the pads and connections and avoid bridging (for the most part). There’s a fine line between applying enough paste and adding so much that solder bridges are created. Some experimentation will probably be needed. For other components, place a small Hershey Kiss® of solder on each pad (see below for the 2nd side).

place chips on the paste: Using tweezers, fingers or a vacuum pickup tool, gently press components into the paste. If you smear the paste, wipe off the board area with alcohol and lint-free wipes and re-apply the paste. You want to avoid having bits of paste around the board since they will form into tiny balls that could come loose and cause a short.

Smaller components will snap into place when the paste melts, but larger components should be aligned as best as possible.

Baking the First Side:

what temperature and for how long? Paste and component manufacturers usually provide temperature profiles that give recommended and maximum values for the reflow process. If chips or boards are heated or cooled too fast, thermal stresses can break chips or warp boards. The paste will also not work correctly if heated too fast or too slow.

One question is just how critical the temperature recommendations really are. Most hobbyists or prototype builders aren’t trying to create 100,000 parts that last 5 years with no defects. But most industrial specifications are designed to provide that kind of performance. So, you might be able to get away with a lot of variation, but on the other hand, thermal damage may only show itself a week or month later–not on the first test. Also bad, the device might “sort-of” work, only producing bad results some of the time. So it’s probably best to follow the guidelines.

Reflow profile: As mentioned above, different components may have special requirements, but for a starting point, look at Kester’s standard reflow profile (this is for lead-based solder, they have a lead-free reflow profile, also).

There are 4 main zones:

• preheat: The assembly temperature is raised to 100-150 degrees C with a max rise of about 2.5 to 3.0 degrees C/sec. The main purpose of this stage is to evaporate solvents from the paste and slowly raise the board and components’ temperatures to avoid thermal shock and warpage. Recommended time: 60-120 seconds.
• soak: For a recommended time of 60-90 seconds, the assembly is slowly brought up to the solder melting (liquidus) temperature. This time allows the board temperatures to equalize and also for the flux to activate and clean off oxides from the metals.
• reflow: This is where the solder is actually reflowed, or melted. A time of 60-150 seconds is recommend, along with a max temperature of 240 degrees C for lead-based solder and 260 degrees C for lead-free. During this time the temperature should reach peak and come back down again to the freezing (solidus) point. You don’t want to immediately cool off the assembly as soon as the solder melts; a little bit of time is needed to allow the solder to properly bond with the connections.
• cool-down: To avoid thermal shock, keep the cool down rate below 6 degrees C / sec. Measuring board temperatures with a thermocouple, we found that everything cooled down much slower than this even with a wide open toaster oven door.

The above guidelines are based on an industry standard called IPC-J-STD-005 (not free). You can find the relevant portions of those specifications copied into this: PCB Land Pattern Design and Surface Mount Guidelines for QFN Packages. The Kester paste recommendations are slightly different. It’s probably best to start off with the lower temperature guidelines.

Again, all these parameters didn’t come into play since our toaster oven could barely heat up fast enough. We simply turned the toaster on full blast, waited for the solder to melt, counted to 15 and opened the door wide open. The board and components will have some impact on timing. Larger components will heat up slower, so be sure to make sure all the solder has melted before opening the door. If you’re using a thermocouple, place it next to the largest component. A chunk from a temperature indicating crayon will change colors permanently when its target temperature is reached. If used, place this by a large component, also.

Onto the 2nd Side:

apply paste

place components: Smaller components will self align during reflow, but larger ones should be carefully lined up.

Bake the 2nd Side:

We used clips as stilts to hold the board up so the components already on the bottom wouldn’t touch the rack. We also tried using chunks of aluminum, but these drew too much heat away from the board, and the nearby joints didn’t reflow.

Results:

First, the bottom side components didn’t fall off.

About 25 pins on our 208 pin, fine-pitch component bridged together. Besides a few other bridges, the other components turned out well.

Conclusion: If we had a stencil, this would be great, but hand soldering all these components may have taken less time than applying paste, placing components and then fixing bridges.

References:

• 10 tips for using Solder Paste from Kester A good primer for dealing with solder paste, although it’s geared towards industrial screen printing.
• Straight from the QFN package manufacturer, PCB Land Pattern Design and Surface Mount Guidelines for QFN Packages. It also contains a description of how to hand-solder QFNs, although their method doesn’t work unless the chip has connections that wrap around the bottom corner to the sides.

• flux: the key to surface mount soldering. Flux removes oxides from metal that prevent solder from bonding to it, and also helps to distribute heat. During typical soldering with flux-cored, solder wire, all the flux you need is contained in the solder. When the wire touches a hot connection, the flux flows out, cleans the joint and prevents further oxidation. However, in surface mount soldering, (brace yourself) oftentimes solder is melted on the iron, and then transferred to the joint. During this time, the flux quickly boils off and becomes useless, so additional flux is needed on the connection. If transferring solder in this manner seems questionable, bare in mind that a common process in industry, called wave soldering, is similar. Fluxed boards are slowly passed over a giant wave of molten solder that wicks into the connections.Flux comes in a large variety of different types and applicators. Our through-hole soldering guide goes over the different types, how the industry classifies them and whether or not cleaning the residues is necessary. To summarize, we recommend using a rosin-based, RMA (Rosin Mildly Activated) flux from Kester® (or anyone else that actually publishes specs on the corrosiveness of the residues). We don’t think cleaning of the residues is necessary for this type of flux for non-critical applications, but feel free to clean anyway–just be sure to do it soon after soldering because the residues quickly harden. “No-clean” fluxes have very low activation levels, and are therefore less effective than activated fluxes, but will work fine on clean parts. Use no-clean flux if you’re making circuits for NASA or are otherwise paranoid (And to feed the paranoia, note that many in industry clean no-clean fluxes). If you use water soluble flux, the residues are corrosive, and should be removed with warm water.

   

  

  The above picture shows a couple means of applying the flux: bottles with needles or brushes, a flux pen and paste flux in a syringe. It’s easier to keep flux only where it’s needed with the flux pen than the needle or brush bottles. However, if you need to add more flux on top of fine pitch pins, the bottles are better since the pen could bend pins. The tackiness of paste flux is helpful for holding components in place.
  • Flush Cutters
  • 3/8-16 Coarse Thread Tap
  • 6in Large Wire Cutters
  • Kester K100LD Lead-Free .02 1oz Solder

• tweezers:Fine-tipped industrial tweezers are essential for handling the small components. Curved tips enable you to grab a chip from the top without having your hand directly above the chip. Hemostats are useful occasionally because they open wider than tweezers, and also lock closed. A set of dental picks or a Xacto® knife is also useful for nudging components and straightening bent leads.
   
  • • IC Chip Remover / Extractor

  1/2in Kapton High Temp Tape Roll

• solder wick / braid: Used to remove bridges / shorts between pins. One popular technique is to flood the leads with solder and then wick up the excess, but this has risks associated with it–read our QFP surface mount side guide for more information. Soldering Iron Stand
• solder: .015″ or .02″ diameter, flux-cored solder are sizes we prefer. We include 63/37 (63% tin, 37% lead) in our starter kit, but 60/40 works, too. A while ago, the circuit board manufacturing industry switched to 63/37 from 60/40 because it was more effective to some degree for mass soldering of surface mount chips. For the skinny on the differences between solders see the solder section from our through-hole soldering guide. In brief, 63/37 is eutectic, which means it freezes at single temperature (like water), whereas 60/40 goes through a “plastic” state where portions of it are frozen and others are still liquid. This means 63/37 freezes slightly faster, flows slightly better and is less likely to form a disturbed joint. Theory aside, we can’t tell much of a difference with hand soldering. Food for thought: water freezes at a single temperature, but a glass of water freezes from the outside in–not all instantaneously. 

  The major decision with flux-cored solder is which flux to use, and our above recommendations apply here, too: use a RMA or RA type and “no-clean” if you’re designing pace makers. And be sure to clean residues from water-soluble flux-cored solder.

  Some people recommend silver-bearing (containing) solder for SMD soldering. Solder containing 2% silver is available but was originally added to prevent solder from dissolving silver from sliver-plated components. While silver-bearing solder may be slightly stronger and have higher conductivity, these differences are extremely small and most likely insignificant.

• magnification: With good lighting, you may be able to get by without any magnification for larger pitches (say > .8mm). A typical lighted magnifying glass on a boom arm is always helpful, but these only provide 2-3X magnification. 10X or more is useful when inspecting the smaller pitched components for sufficient solder and shorts between pins. A loupe is the cheapest route, but it can only be used for inspection after the soldering is complete. A stereo zoom microscope can be looked through during the soldering process, but these run $400+ used and much more for new scopes. Our scope recommendations are to get 30X magnification, stereo zoom, and the largest “working distance” possible. Working distance is the distance between the lenses and the work–the more room for tools and hands the better. Helping Hands





• soldering iron and tip: A lot of this comes down to personal preference, but we recommend getting a temperature-controlled soldering station with at least 50 Watts, and a 1/32″ chisel or screwdriver shaped tip.
  • the iron / station:
    

    our recommended soldering station: Weller WES51 (50 Watts, temperature controlled) or 4 Pos. Screw Terminal

    The small connections on surface mount components won’t draw that much power, so 15 Watts of power will work, but if you ever end up soldering to a large ground plane, big connector or a large wire, you’ll wish you had the power. Most hand-held irons are not temperature controlled, which means they’re constantly dumping out their rated power in heat regardless of the tip temperature. A temperature controlled station throttles the heat delivery to maintain a constant temperature. This is useful because the iron will push harder if you’re soldering a lot of joints or large components, whereas an unregulated iron will cool off and take longer to heat connections. Standard temperature recommendations are 600-700 degrees F for tin-lead solder, and 700-800 degree F for lead-free solder. Our surface mount video was shot using temperatures from 610-640 degrees F.

    On the topic of tip temperatures, here’s a youtube video showing tip temperatures of various RadioShack® irons while soldering.

  • the tip: We use a 1/32″ chisel tip even on .5mm pitch components. Smaller tips won’t work with the techniques shown in our SMD soldering 101 video because it’s difficult to hold solder at the very end of the tip. The most popular technique for surface mount soldering, called “drag soldering,” uses an even larger tip to hold a blob of solder that can then be dragged over pins. Some names for drag soldering tips include “hoof,” “mini-wave” and “bevel” tips. We use the side of a chisel tip as shown in the video. Jumper Wire Kit (350 pieces)
    

  There are also a variety of other tip shapes used for desoldering SMDs, but we don’t recommend buying a different tip for each chip–just use ChipQuik®, hot air or a skillet.

   

  A short video demonstrating desoldering a SOIC with a specialized desoldering tip.

• ChipQuik®:This is basically a low melting point solder that, when heated, stays molten long enough for all the pins to be released at once.
  

  containing alloy, paste flux in a syringe, and alcohol pads for cleaning. Click the link for a 40 second demo video.

Hot Air: Essential for Soldering / Desoldering Leadless Packages

Some chips, such as QFNs (Quad Flat No-Lead), have connections on the bottom of the chip that can’t be soldered with a typical iron. Short of getting a SchmartBoard® or drilling a hole in the PCB, hot air is required to heat the connections (a skillet or toaster oven will work, too–see below).

QFN (Quad Flat No-Lead) package upside down showing the bottom connections around the perimeter and the heat sink in the middle.

Hot air station from Madell and an embossing tool. We also recommend FM Radio Arduino Shield

A wide assortment of tips that match just about every package style can be purchased, but we’ve had luck only using a 3/16″ round tip for all but the largest chips. The picture shows a relatively inexpensive hot air station that features both temperature and flow control. It also pictures something called an embossing tool that is normally used to create decorative, raised ink patterns on invitations and other arts & crafts projects. Conveniently, the air temperature is hot enough (low 600’s F) for surface mount soldering. The hot air station goes up to about 900 degrees F (going past 800 is risky), but for standard lead-based solder (lead-free requires slightly higher temps), the embossing tool gets the job done, and you can’t beat the price: $25. Note that standard heat guns from hardware stores have nozzles that are too large; they’re usually too hot, also.

Various hot air tips

• pre-heater / hot plate:For boards with larger components or ground places, much of the heat applied by a hot air tool gets drawn away from a target component into the surrounding board. This means more time and heat is needed to raise the target part up to reflow temperature, and this always increases risk of damage. Also, most components and solder pastes have limits on how fast they can rise in temperature. If one area of a chip or board gets hotter too much faster than another, thermal expansion can warp boards or destroy chips.Enter the pre-heater or hotplate. By raising the overall board temperature to 212-250 degrees F, less heat from the high temperature hot air tool will be required, and thermal shock will be much less severe. Madell and Zephyr sell fairly inexpensive pre-heaters, but a $7 Mr. Coffee® works well for small, single sided boards. You’ll need to get a PCB holder if you use one of the hot air pre-heaters.

  Can you get away without one? Probably, but be aware that thermal damage could reduce the life-span of your board, not simply kill it immediately. Preheating is even more important for BGAs, as their thermal profiles are much more critical.

A PCB holder, pre-heater and a Mr. Coffee heater

The other route to go with leadless chips (or if you want to run small numbers of boards) is to use solder paste and a toaster oven or skillet.

Small toaster ovens can be used to reflow solder paste

• toaster oven:Ideally, a toaster oven is meant to recreate the temperature and time profile from a reflow oven in the real manufacturing process. This involves taking the board through a couple temperature stages where it’s first pre-heated, and then “reflowed” by raising the temperature above the solder’s melting point for about a minute. There are a couple companies out there that sell toaster oven controllers, namely ArticulationLLC.com and RheSiliconHorizon.com. These are neat because you can fine tune all sorts of parameters, and then just press “Start”. In our experience, however, small toaster ovens don’t heat up fast enough to require any sort of throttling by a controller. We simply turn it on high, wait for the solder to melt, and then open the door. A controller might be useful for more complex boards, though, and it’s nice to be able to log the temperatures.You want an oven that can go from 0 to 250�C (480�F) in less than 5 minutes in order to completely reflow the solder without baking the board. Our first purchase was a large, GE convection oven with 4 burners that took about 10 minutes to remelt solder (and burn the board). A $20, smaller oven with 2 burners did much better. So, if you get a large one, get the most wattage you can (preferably over 1400 Watts).

  As a later section in this guide show, you can do double sided boards in toaster ovens by propping up the PCB on stilts.

• skillet: If you have a single sided board, a skillet can also be used to reflow solder paste. Try to get one that goes up to at least 450 degrees F. The guys at Sparkfun.com have some great tutorials / blogs about their surface mount soldering experience, and highly recommend using a skillet to reflow (melt the solder paste) their boards. They apparently had trouble dealing with a mix of plastic and large metal connectors in a real oven, but were successful with a skillet. One other advantage of a skillet is that you can easily remove chips with tweezers rather than trying to reach inside an oven.

Several different brands of solder paste

• solder paste: This consists of tiny solder balls floating in gel-like flux. Once paste is applied to the pads, chips are placed on top, and the board is “reflowed” (paste melted) in a toaster oven or with hot air. Paste can be applied using the syringes shown or with a squeegee and stencil. For stencils, try StencilsUnlimited.com. Paste in syringes can be purchased from ChipQuik, Zephyrtronics, SMTSolderPaste.com and many others. Note that paste in syringes usually has slightly less metal content to help it flow through small needles. Get paste in a jar if you’re using a stencil. The main choice to make is between no-clean or water soluble paste. We recommend no-clean unless you have reason to believe your components are difficult to solder–ie, old and possibly corroded. The residues from water-soluble paste are corrosive, so be sure to clean them with warm water.Be aware that some distributors require 2nd day or faster shipping since paste lifespan decreases outside of refrigeration. If you get a syringe, you’ll likely have to buy a needle and plunger, too. A 22 gauge needle is a good starting place, and you can always lay a thicker bead of paste just by pushing more out.

Solder paste under 30x magnification

Some non-essentials that are nice to have:

• 4-40 Coarse Thread Tap and No. 43 Drill Bit : If you want to clean flux residues, use an acid brush with IPA alcohol. Be sure to wipe up the residues with a lint-free wipe (like Kim wipes) and not just move them around on the board.
• Deburring Tool  Pushing down on the top pumps up a small amount of alcohol into the dish while keeping the rest from evaporating.
• sponge with a hole: A whole in the middle gives you an edge to wipe the iron tip on, and also a place for the used solder to fall into so you’re not trying to clean the tip on older debris.
• temperature indicating crayon: Marks from this permanently change color when it reaches a particular temperature, and is useful if you’re using a toaster oven without any thermocouples to watch the temperature. We had luck just watching for the solder to melt, though. In any case, different parts of the board will reach melting temperatures at different times due to ground planes and large components that suck up heat.
• PanaVise Junior Clamp : An alternative to a sponge, it contains soft metal curls coated in flux that clean the tip without thermally shocking it. This can help to prolong tip life.
• 10-24 Coarse Thread Tap : The smaller version of the popular PanaVise has slots to hold circuit boards, and is much more stable than “helping hands.”
• Softpot Touch Sensors : This company makes a whole line of prototyping boards for surface mount components, including anything from resistors to fine-pitch QFPs to QFNs to BGAs. It’s a clever product that uses small troughs pre-filled with solder to align the chip. To solder, you simply push the solder up to each pin with a small iron tip.
• fixing mistakes: The conductive ink pen lets you simply draw traces on a board. Another option is to use tiny wire (such as 30 gauge wire wrap wire) to jumper over mistakes or lifted pads. These are called “green wires” because manufacturers initially used green colored wire to blend in with the green PCB.
• Plug-in Bread Board Dual Power Supply : (not shown) A tip tinner / cleaner (which you can still get from RadioShack®!) has a more aggressive flux in it that can help clean stubborn residues off a tip. You can also get a polishing bar to refresh a dirty tip. Polish lightly, though–once the protective outer plating is pierced solder will quickly dissolve the copper inside.

The leads on a PLCC are called J-leads due to their shape. A brief summary of the steps is to flux the pads, tack two opposing corners, add more flux and then solder the rest of the pins.

First, flux the pads.

Then align the part, keeping in mind that pin 1 on PLCCs is in the middle of a row, not a corner.

Place a small amount of solder on a clean tip, and then touch a corner pin and its pad while holding down the chip. You may want to add some flux to this pin beforehand. The next step is to tack an opposite corner, but be sure to re-check the alignment first. Once two pins are soldered you’ll need to completely remove the chip to make any adjustments.

Add additional flux to the pins of the PLCC. The flux from the wire solder may be enough to skip this step, however.

Lay a piece of solder against the bottom of the leads and the board; .02″ diameter solder works well and is shown on the left. Now successively press the solder into each one of the pins. Heat each joint long enough to allow the solder to completely wick around to the back of the pin, but no longer.

Finished Results:

The goal is to have the solder wick all the way around to the back of the pin, and have a smooth filet (or curved ramp) connecting the pin to the pad in the front. The most important aspect is that the solder looks like it wetted (adhered to, clung to) the metal. Lead-free solder and some types of flux with lead-based solder will leave duller joints; this is OK.