For my Modding my EEE PC to do Awesome Things project, I've found that I need a way to desolder really tough components from a board. I've seen articles online where the author took an electronic desoldering iron, added an aquarium pump, and turned it into a pretty effective hot air pencil.
Just by chance, we have started selling those desoldering irons in work for under £10 (Maplin: N37CH) . I went to the local pet shop and bought an aquarium pump for about £10, and 1m of tubing for 40p. Total cost - < £20 (with my staff discount)
The desoldering iron is basically a desoldering pump with a heated metal tip. There is an element somewhere near the tip which heats the whole thing, and a hollow tube all the way down the middle into the suction chamber.
Using the aquarium pump, I plan to convert that sucking action into a steady stream of air blown through the tip, which should be heated, then expelled onto the board in order to melt the solder.
Step 1 - Strip down the desolder iron
The first step is to strip the desolder iron down into its individual parts. With my iron, all I had to do was stick a screwdriver round the back of the yellow chamber and pry it out. It slid out nicely and allowed me access to the little black bit which guides and holds the plunger.
The plunger really isn't needed any more so I removed mine. This left a nice hole to stick the end of the pipe into. I had to drill mine out to about 6.5mm dia. before I could fit the pipe in, and then used copious amounts of superglue to hold it in place.
This pretty much completes the mods to the desoldering iron. Just put it back together with the pipe sticking out the back.
Step 2 - The Pump
Nothing really needs modded here. Just attach the end of the pipe to the pump's outlet. In my case the pump has two outputs. I assume this comes from one motor, which would mean that all my air would escape out the unsealed hole. I used a little piece of pipe, sealed one end, and put it on the other outlet to plug it up.
And that's it - a (hopefully) finished hot air pencil. Now to try it out.
Step 3 - Testing
Obviously I have to test this sucker now. Below are the two test subjects - a pre-soldered circuit board and a drink bottle with a plastic wrapper.
Well what do you know? It didn't work! At least not on the solder on the circuit board. I did manage to melt a line down the wrapper on my drink bottle, but stopped before actually piercing the bottle as I planned to drink it later. Anyway the important point is it failed to melt the solder on my board.
Temperature of iron with air: 352°C
Temperature of air at 1cm: 220°C
According to the manual of my solder station lead free solder melts in the range of 350-380°C. Obviously my hot air flow is too cold to melt the solder.
With the air on my soldering iron can just about get up to the required temperature . However the tip was able to melt the solder with ease whenever it touched it directly. I have to assume that the tip is a lot hotter but the air flow through the iron is cooling the probe and messing with the results.
Step 4 - Perfection
I need to increase the temperature of my airflow by about 100°C. There are two ways I can think of doing this. The first is to place a piece of copper braid inside the barrel of the iron. This copper braid will heat up, providing a larger surface area for the transfer of heat to the air on its way through the iron.
Thermometer readings (with solder braid):
Temperature of iron without air: 396°C
Temperature of iron with air: 381°C
Temperature of air at 1cm: 300°C
Well look at that! An increase of 80°C already! In order to get these I had to strip the whole thing down again and superglue some of the seals. (I also had to burn my fingers a few times on the hot air jet)
And the PCB melting test result is: fail. I have a feeling we're very close, but it just wont melt.
Step 5 - Perfection II
Maybe some extra air flow will improve things. I decided to open up the pump and see if theres any way to increase the flow.
This pump actually has a really clever design! What you see at the back is most of a transformer, which they are using as an electromagnet. As the AC magnetises the metal core the permenant magnets on the arms of the pump are attracted, squeezing the bladder about half way down the arm and expelling the air inside. Somewhere inside there are a few one-way valves which mean the expelled air goes in the right direction and isn't sucked up again.
More importantly for this hack though, both outputs are completely independent, which means that if I combine them I will get double the airflow.
Ok, even for me this is a bit of a hashed up job, but with a mixture of sticky tape and superglue I managed to get a pretty airtight seal. And the tests?
Thermometer readings (without solder braid, with double air flow):
Temperature of iron with air: 340°C
Temperature of air at 1cm: 190°C
Thermometer readings (with solder braid and double air flow):
Temperature of iron with air: 310°C
Temperature of air at 1cm: 150°C
Hmmm alright... Something has seriously gone wrong here. The solder braid is actually taking the temperature down significantly, and the temperature of the iron is dropping. I think the extra airflow is actually cooling the iron too much and the 30W element can't keep up. Oh well, back to the drawing board.
Step 6 - Perfection III
Ok double airflow didn't work, how about a reduced airflow?
After dismantling my homemade T-junction (and supergluing my fingers together with un-dried superglue), I went back to the single air flow. I set up a rig to hold the temperature probe in front of the nozzle and squeezed the pipe to try and find the highest temperature I could obtain.
Results (with solder braid):
Temp @ 100% airflow: 200°C
Highest attainable temp: 290°C
Two observations here. First, I can only get 200°C here! If you remember from earlier, I was able to get 300°C the first time I tested this. This highlights the servere un-repeatabilty of these results. Secondly, I was able to get a 90°C increase by squeezing the tube. This was with about 75-90% of the airflow restricted! I still want to test this on the board though:
Aw damnit! Still didn't work!
Well, it seems that my 30w soldering iron is just not strong enough to stay warm while the airflow is going through it. Guess I'm going to need a stronger soldering iron to continue this.