DIY hot-air iron
A while back, I wrote a column on fine-pitch soldering . Several commenters noted that a hot-air soldering iron was the hand tool of choice when it came to fine pitch soldering by hand. So I decided to look up the tools and found that they were out of my price range. I still wanted one. After more research, I found this site  which describes how to build such a tool from a soldering iron. Based on that, I came up with my own set of instructions  for putting together a hot-air iron. The instructions are complete and include where to buy the iron. I'd like to give special thanks to Nelly Cabrera of the Stimpson Eyelet Company  who sent me samples of the final piece required to complete the design. The nozzles (eyelets).
The development of the hot-air iron was not smooth sailing despite what seemed like adequate instructions.
Final testing of the hot-air iron had me pulling my hair out because I couldn't make solder melt with it. I was using the smallest pump I had, an AQT3001, which was a remnant from previous acquarium adventures - my thinking was that if another builder has to buy a pump, a small one will cost less. It didn't work. I tried different eyelets (nozzle tips), and nothing seemed to help. The narrow steel nozzle (Stimpson A1735) had a fairly high temperature output, but so did a more open brass nozzle (Stimpson GS5-12). Well, I had theories. One was: "too much air flow". So I used an air valve to decrease the air flow. It made things worse. Was the steel losing less temperature because it was less thermally conductive? I tried using the brass eyelet as a guard around the steel eyelet in the hopes that I'd get some gain. Bigger losses.
I was at my wits end. I thought, "I'll try something that shouldn't work, but maybe I'll learn something." Did I ever. What I tried was adding a second pump, a Whisper 10, that I had on hand to my setup. I used a "T" and some tubing to connect it to the setup. In theory, more flow should have decreased the temperature output. But it didn't. The temperature went up. And now I could melt solder. Quite nicely with less than a ten second dwell.
Under what conditions does increasing the flow through a heat exchanger increase the heat transfer? That happens if the increase in flow causes a change from a laminar flow regime to a turbulent flow regime. Increasing the flow causes the flow to became turbulent. A very good thing in a heat exchanger. It is my opinion that the real purpose of stuffing the heat transfer tube tube of the iron with a metalic foil (see construction article ) is not to directly increase heat conduction from the heater walls but to increase the turbulence of the airflow through the soldering iron body to increase the heat transfer. If I hadn't started my experiments with an inadequate air pump because I wanted to keep costs down I would never have learned that.
To get deeper into this aspect of heat transfer, you can look up Reynolds number . In fact, don't fly without it.
All that got me to thinking further about what one more pump in tandem with the other two would do. The first mate and I decided to make a tour of the pet stores looking for pumps. We ogled the fish at one store but I didn't see any pumps that interested me. So we went to the Petco  on State Street in Rockford. I didn't want to buy the smallest size in their pump line, the AC-9901, although it probably would have been adequate for what I had in mind. And they were out of the AC-9902 that I wanted. Dang. So I reluctantly paid $22 plus tax for the AC-9903. Am I ever glad that it did. By itself it has plenty of pressure (which equates to flow) for this application and also has a control knob on top which controls the pressure/flow. It can create enough flow so that the iron output air temperature actually decreases at the highest flow rates.
What I have to do now is to learn how to use the tool. If I turn the flow up high enough I can blow components (0603s) all over the universe. And that is no where near the highest setting.
You will notice that I only looked at Petco pumps. There is a reason for that. Petco gives the actual specs of its line of pumps on the box. Prominently, on the outside. For instance, the AC-9903 I used needs 4 watts to deliver 4.5 liters a minute (70 gph), with a maximum pressure of .014 MPa (about 2 psi). No doubt you don't get both at the same time.
Now I should have known, from my Naval Nuclear Power Training, that increasing the flow in a heat exchanger can increase the heat transfer significantly if the flow regime changes. But I had never run into it as a practical problem because I never had to design anything where air flow mediated heat transfer (other than laminar flow convection cooled semiconductors/heat sinks) was my direct responsibility. Until this little project. Moral of the story? It is a complicated universe. Sometimes first order approximations are not even good for the first order.
M. Simon's e-mail can be found on the sidebar at Space-Time Productions .
Engineering is the art of making what you want from what you can get at a profit.