Technical Article

Introduction to the Manually-Controlled Toaster Oven Reflow

December 14, 2015 by Robert Keim

With the help of a DIY thermocouple measurement system, you can use a cheap toaster oven to accurately reproduce a reflow-soldering temperature profile.

With the help of a DIY thermocouple measurement system, you can use a cheap toaster oven to accurately reproduce a reflow-soldering temperature profile.

The Reflow Quandary

In a previous article (Make an EFM8-Based System for Monitoring and Analyzing Thermocouple Measurements), we developed an EFM8-and-Scilab-based system for collecting, recording, and displaying thermocouple temperature measurements. That article tells you how to build and use the system, and it provides links for downloading the microcontroller code and the Scilab script. In short, the system displays the current temperature and a graph representing the temperature history, as follows:

This temperature feedback will enable you, the assembly technician, to accurately control the oven’s heating profile via simple, real-time adjustments.

As you may know, surface-mount assembly can be a serious impediment for students, hobbyists, entrepreneurs, or anyone else who wants to design and test high-performance circuit boards without paying for professional assembly. Hand soldering is becoming increasingly burdensome—sometimes impossible—owing to the proliferation of miniscule components, leadless packages, and ball-grid arrays. Professional assembly is rarely a feasible option: it is surprisingly expensive, especially for small quantities, and coordinating with the assembly house can be awkward if you are trying to use whatever stray components you have lying around your lab/office/garage/bedroom. Also, let’s say you decide to splurge on the initial assembly job—what happens when you need to replace that 100-pin microcontroller? Even if you have a hot-air soldering station to help you with this kind of rework, before long you will probably be wishing for a tube of solder paste and a reflow oven.

Fortunately, low-cost, simple, DIY reflow is perfectly achievable. It’s not easy, especially with tiny or small-pitch components, and it’s not an automated, defect-free, assembly-line sort of operation—your brain is needed to help the toaster oven control its temperature, and your hands may be needed to remedy a few solder bridges or “tombstoned” capacitors. But again, DIY reflow is achievable with little more than a cheap toaster oven, a tube of solder paste, and a system for measuring and recording temperatures.

Toast, PCB—What’s the Difference?

There are a few steps—along with plenty of tips and techniques—involved in the entire DIY PCB reflow process. But you won’t get anywhere until you have a hot container in which to melt your solder paste, so first we need to focus on pressing an ordinary toaster oven into service as a reflow appliance extraordinaire. In this article we will cover preliminary information and techniques, and the next article will present a detailed procedure for compelling a toaster oven to obey the temperature profile of your choice. Here is a photo of my oven:

This oven sells for as low as $30. It is nothing fancy—no convection, no infrared, no digital display. Maybe that stuff helps, but it’s not essential. Actually, there a few things I like about this oven:

  • The metal tray that slides in underneath the rack helps to distribute the intense heat generated by the lower heating element.
  • It’s small—no point in wasting electricity on a big oven when all you need to cook is a PCB.
  • Temperature is controlled by a mechanical dial rather than buttons—this allows for rapid, intuitive, pleasantly tactile adjustments.
  • The timer’s old-fashioned “dingggg” sound reminds me of home . . .

The Profile

Manufacturers do not formulate and publish reflow temperature profiles simply to entertain themselves. I’m sure it is possible to successfully reflow a PCB without adhering to the profile, but I highly doubt that the results are consistent or of particularly high quality. So I think it’s well worth the time and effort to develop a procedure that allows you to replicate a profile with reasonable accuracy. The first step, of course, is deciding which profile to follow. Sometimes the datasheet for an individual part will include a recommended profile. Obviously, if four different parts on the same board all come with different profile recommendations, you can’t please them all. If your board includes one component that is more important or temperature sensitive than all the others, perhaps you should follow the recommended profile for that component. In general, though, it makes sense to use the profile that accompanies your chosen solder paste—your PCB has a variety of components but only one type of paste, and the manufacturer knows what sort of temperatures will make the paste do its thing.

In this article we will use the profile for a leaded, no-clean solder paste manufactured by MG Chemicals. (Lead-free solder requires higher temperatures that are more stressful for the components and the PCB, and the environmental impact of using leaded paste on a few prototype PCBs is probably not something to worry about.)

Here is a brief description of the three primary temperature phases:

  1. First, we need to bring PCB, components, and solder paste from room temperature to something much higher than room temperature. This is called the “preheat” phase.
  2. The “soak” phase, which in this profile is included as part of the preheat phase, uses a relatively stable temperature so that PCB, components, and paste can reach thermal equilibrium.
  3. The “reflow” phase takes the PCB from thermal equilibrium to temperatures above the solder’s melting point. The solder paste melts and “flows,” wicking away from non-metallic areas and forming bonds with copper PCB pads and component leads/lands/balls.

As indicated in the profile, the critical characteristics are the following:

  • the rate of temperature increase during the preheat phase
  • the duration and temperature for the soak phase
  • the duration of the reflow phase, which begins when the temperature exceeds the solder’s melting point and ends when the temperature falls below the melting point
  • the peak temperature during the reflow phase

Preparing Your Hardware

Here are two photos of my setup:

The thermocouple enters through the crack between the glass door and the metal frame, and I bent the tip of the thermocouple so that it stays close to the surface of the PCB. You may want to experiment with different PCB locations. My placement is based on the following considerations:

  • The lower rack position is preferable to the upper rack position because I have the metal tray to protect the board from the lower heating element. Less separation between the PCB and the upper heating element might contribute to problematic temperature gradients.
  • Though we want to avoid gradients, we also want good response to the heating element (primarily for the steep transition from soak phase to reflow phase). Thus, I keep the PCB approximately aligned with the lower heating element.
  • We will need to open the door to facilitate heat removal, and a central placement (relative to the front and back of the oven) seems to provide an acceptable rate of temperature decrease.

Conclusion

Now we understand why we are trying to assemble a printed circuit board in a toaster oven and how to go about it, and we have a temperature profile in hand. The next article will show you how to turn a low-precision, low-cost kitchen appliance into something that can actually function as a reflow oven.

2 Comments
  • P
    PhilTilson December 24, 2015

    Interesting!  I have been successfully manufacturing small numbers of SMD PCBs using a slightly more sophisticated oven for some time now.  My particular setup uses a microwave/fan/grill oven - though not on microwave, of course!  (I once inadvertently hit Microwave rather than Grill, and within a second or two all the ICs had ‘popped’, with an exciting display of sparks!)

    In my case, I pre-heat the fan oven to 180ºC, then put in the boards for two minutes.  I then switch to Grill and heat the boards for around another two minutes until the solder paste starts to flow.  I cut the power after about 10 seconds and open the door for 15 seconds or so, then remove the boards and place them on a cool granite surface.  The theory is that the thickness of the PCB slows the cooling to the cold granite.

    With this technique I am now able to produce batches of PCBs, four at a time, with very few problems; most of the boards check out perfectly.  Occasionally the solder paste has not flowed properly onto a component, easily remedied with a fine-tipped soldering iron.  So far (touch wood!) I have not damaged any components by overheating using this method and I’m not aware of any ‘in the field’ failures of soldered joints.

    I shall look forward to the following article(s) to see how I SHOULD have been doing it!

    Like. Reply
  • S
    Steve M December 24, 2015

    For the best setup build a test (profile) card that is fully populated and set a thermocouple in the most densely populated area of the card. You want to find the coolest part of the card. When setting thermocouples, you should have the them touching the board. I used Kapton tape or conductive epoxy to hold them to the card. Depending on the size of the card, multiple samples would be needed.
    You need to ensure that the coolest areas of the card reach reflow.  45-60 secs liquidus in cooler areas should get good results.

    Like. Reply