For many years, my main soldering iron had been a Weller clone. Changing between tips was a hassle and it was lacking power for larger ground pours. It was due time for an upgrade. Well-known brands such as Weller, Ersa, JBC, and Metcal are expensive, often costing over $3000. Furthermore, they pretty much force you to stay in their ecosystem. There is no good way to switch once you decide on one and buy all their tools and tips.
Luckily, there exists a solution to this problem. It is called Unisolder. A soldering iron controller, designed to support as many different soldering iron handles as possible. It is not the cheapest controller out there but it is certainly the most versatile one I know. Really universal soldering controller (dangerousprototypes.com)
Building the controller
This controller you can’t just buy. It has to be assembled by hand and let me tell you, it is not as trivial as it sounds. The schematics, bill of materials and gerber files can be found on the forum. The rest is up to you. You have to source all the components and order the PCBs. I had some trouble with finding some components because the original BOM was created in 2015. When everything had arrived I started with the assembly. It was one of the hardest boards I ever assembled. I spent weeks getting the controller to boot and to pass the initialization stage.
I made a solder bridge between 2 pins of an IC. The program thaught it was a problem with the I2C bus, causing me to hours of debugging and finally replacing all I2C ICs. It was none of them. When I finally found the issue, I was so pissed off that i gave in and bought a microscope.
After finding the soldering mistake, the controller would finally turn on. But as soon as the handle with the tip was connected it would reboot. After more long hours of troubleshooting, I found a 3V zener diode which was measuring 4.2V. I was sure the zener I ordered was correct. No matter… I ordered two more batches of Zeners from Mouser, and out of all of them, only one was measuring true. This experience made me question my trust in Mouser. The datasheet wasn’t very clear, but it did state it was a 3-volt Zener. It’s crazy that you can order three times and get the wrong part. If this had been a more complex circuit, debugging would have taken forever.
Once all the issues were solved, my soldering iron was finally working. I built a quick enclosure to test it and ran it as my daily driver for a couple of months. It was awesome. The soldering iron just melted everything effortlessly. I was even able to solder a radiator copper pipe, which was crazy to me.
Adding more tools
After some initial testing and determining the quality and usability of the station, I bought a desoldering iron and tweezers, along with their respective tips. I also got some additional tips for the soldering iron itself, including a big chunky one and a fine tip for detailed assembly work. The problem with the current setup is that while the Unisolder is a versatile controller, it’s primarily designed for soldering irons. Not for switching between them. It’s not intended to be a full-blown multi-tool station.
What I envisioned was a modular system that could support various tools like soldering irons, desoldering tools, tweezers, and switch between them. I wanted it to be customizable so other people with different tools than me could benefit from it. The goal was to be able to add and try out new tools, like a hot air handle.
Because of the similarity of the handles I decided that all the modules will be the same and configurable. You would press a button to select a module, and every other module would be deselected. Someone else had tried something similar, but with less success. They made a YouTube video showing their attempt, where the temperature spiked up to around 500 degrees or so. From what I gathered, the issue seems to be related to noise, as noted by someone on the forum. They mentioned that the problem was due to the messy wiring and high noise levels. I wasn’t entirely clear on what they meant by “noise” or where it originated from, and I couldn’t find a detailed list of the specific issues they faced. This left me with a bit of uncertainty about how to avoid the same problems in my design.
My only option was to try it and see what happens. I built a small relay board to handle the module switching. I used SPDT relays. It turned out I needed four of these relays to properly handle the module selection. There are eight wires (excluding ground) required for the most complicated module, therefore every board has to have the ability to detach 8 wires.
I connected the relay board from the handle to the controller, and when I switched on the relays, everything seemed to work as expected. I also checked the waveform graphs of the soldering iron to see if there was any noticeable difference or increased noise, but everything looked pretty much the same to me. I decided to go ahead and build it, but I was still worried about whether it would work.
Module design
If I wanted everything to be modular I had to design the most involved module first. If my PCB included everything for that module all others would work. I also had to move everything around so it would fit into the enclosure. The most complicated is the desolder module. Solder and tweezers modules are simpler. Yet the board still had to work with all different combinations of connecting the tool.
I drew up a circuit inspired by the old car radios where pressing one button would pop out all the others. I designed a similar system for my project. It was quite fun to design and implement. I used a breadboard (link) and 74HC logic family components for the circuit. The design worked as intended, allowing me to switch modules in and out as needed.
PCB
The PCB I designed serves a crucial role in managing the connection between the various tools and the controller. Essentially, the PCB handles connecting or disconnecting the tool modules via relays. Here’s how it works:
Relay Control: The PCB controls relays that switch the connection between the tool modules and the controller. When a specific module is selected, the PCB disconnects all other modules first and then, after a brief delay (around half a second), reconnects them. This ensures that only the selected tool is active at any given time.
Modular ID Configuration: To make the PCB compatible with various controllers, it includes an ID selection feature. You can configure the ID to match the station’s requirements, allowing the controller to recognize and manage different tools effectively.
Button Detection: The PCB also features button detection for module selection and other functions. This detection is a bit unconventional; it involves pulling certain pins to ground using 10-mega-ohm resistors and transistors that act like switches. The design is intended to detect when a button is pressed by checking if the pin is pulled to a low logic state.
During assembly and testing, I faced a few unexpected issues:
Button Detection Problems: Initially, the button detection didn’t work as expected. Pressing the button didn’t trigger the desoldering iron as it should. When I started measuring with the oscilloscope, the issue seemed to resolve as soon as I touched the ground with the probe. This behavior was puzzling and seemed to coincide with the desoldering iron cartridges glowing red.
Intermittent Problems: After some time, the button detection issue fixed itself, but I was still unclear about the root cause. I suspect the issue might have been related to a damaged or leaking component, possibly the reflector, causing some electrical interference or noise.
These problems highlighted the importance of thorough testing and debugging. Despite these challenges, the system eventually started working reliably, which was a relief.
Overall, the PCB effectively manages the tool connections and has proven to be a flexible solution for my soldering station.
Enclosure & cassettes
After getting the circuit working, I turned my attention to making an enclosure. I had faced many problems with enclosures before, particularly with my high-voltage power supply. The process was cumbersome: finding a large and heavy enclosure, drilling holes, and painting it was a lot of work. The result never looked as good as something I could have bought ready-made. In the end, I probably spent more time and money on it than if I had just purchased an enclosure.
I was looking through Mouser to find the best case for my project but couldn’t find anything suitable. I remembered seeing a case in a video from Marco Reps, which looked very interesting. Insert link here. It was a Euro-style desktop enclosure that cost around 100 euros. The great thing about it was that it used standardized 100 x 160 millimeter PCBs and had interchangeable cassettes that fit inside. This made it easy to build and customize your setup.
I ordered the case but ran into trouble finding the correct cassettes. Despite their standardized size, I couldn’t locate the right cassettes on major distributor sites like Mouser or Digikey. It turned out that I needed to order directly from the manufacturer to get the correct ones. Specifically, I needed the 14 HP or 850 cassettes, but finding these proved to be quite challenging.
Then it hit me: instead of waiting for the cassettes from Conrad, which didn’t arrive on time, I decided to 3D print them myself. I downloaded the 3D models from Fischer and customized them to fit my needs. I made modifications for the front and back panels and adjusted the size of the PCB to fit since I didn’t have the standard connectors or backplane. I only used the Eurorack standard because I needed ready-made enclosures that looked good. Not because I necessarily needed to switch modules in and out and change their positions on the fly.
By the way, for the enclosure, I also replaced the sides. The original plastic sides were quite unattractive, so I swapped them out for some nicer, more aesthetically pleasing panels. This made the enclosure look much better.
Insert image here.
I printed the cassettes in plastic and am now sending the files to JLCPCB for the front panels. The cost for printing these at JLCPCB is very reasonable, especially compared to doing them by hand or on a breadboard. Plus, their anodizing options are quite affordable, making it a great solution for future projects.
Vacuum
This was my first experience with vacuum work, and I quickly discovered that it’s not as complicated as it might seem. I sourced most of the vacuum components from AliExpress, and the quality turned out to be quite good. The metal parts can handle up to 20 bars of pressure, while some plastic parts are rated for up to 10 bars. However, I’ve been running everything at 5 bars because that’s the pressure level I found to be effective.
The vacuum system for the desoldering handle uses a simple design: an air block with three openings. The air moves through the pipe, creating a vacuum on one side of the block. Before the air reaches the block, it passes through a solenoid controlled by a button on the handle. When you press the button, the air rushes through, creating negative pressure that pulls the solder from the tip.
There’s also a simple valve at the front to regulate the pressure, but I found it isn’t very useful. I plan to keep it open most of the time. However, if I ever design a hot air system, precise flow regulation will be essential to avoid blowing components away.
I also plan to add a reduction valve to limit the maximum input pressure to the soldering station to 5 bars. I intend to connect everything to a compressor and distribute compressed air throughout the workshop using copper piping. I might document that process in the future.
The initial idea was to use a different type of valve, but it turned out to be just a regular valve that could be adjusted to let more or less air through. If you set it to 1,000 bars and open it halfway, it still provides 500 bars of pressure. Instead, a reduction valve will ensure that the pressure never exceeds 5 bars.
[Insert image of the desoldering cassette with arrows indicating airflow direction here.]
Rotary encoder
Additionally, instead of using the standard buttons for the controller, I added a rotary encoder. This upgrade wasn’t too difficult; I just followed the diagram and made the necessary connections. [Insert image here.]
Connectors
Then I assembled everything and made the connectors. I also want to discuss the different styles of connectors I used, particularly their pinouts. I plan to include diagrams for the pinouts of Binder and Hirose connectors, covering all three types I used. Since these connectors are not widely known and their part numbers aren’t well-documented, I think it’s important to provide this information. Proper documentation can be very helpful for others who might need it.
Glowing problem
During assembly, I ran into a major issue. I wasn’t entirely sure if everything would work, but it seemed promising. However, when I first connected the system, the tweezers began to glow unexpectedly. At first, I thought it might be a wiring issue with the tweezers or some problem with the device itself. Unfortunately, the same issue happened with my desoldering iron.
I started documenting all the connections, what I tested, and the results. I was getting worried because I couldn’t figure out why this was happening. I feared I might not be able to fix it. Then, the worst happened: the system began to work intermittently without me changing anything. An intermittent problem is always challenging because it’s hard to pinpoint the cause.
For now, it appears to be working, but it’s frustrating because the tips aren’t cheap. For instance, the heater cartridge for the desoldering iron costs 150 euros, and each cartridge for the tweezers is 50 euros. Only one of the tweezers was malfunctioning, which adds to the concern.
If this problem happens again, I’ll need to review the code and diagnose why the heater might be overheating. It could be related to the feedback loop in the temperature controller, where it mistakenly thinks the temperature is very low (possibly even zero degrees) and tries to supply excessive power to the heater. This shouldn’t happen if the controller is functioning properly, as it should alert me with an “open heater” or “open sensor” message if there’s a disconnection.
Overall, it’s worrying that this issue is occurring, and I’m concerned about the reliability of the system.
In the next section, I’ll describe what the PCB does. Essentially, the PCB’s primary function is to connect or disconnect the tools from the controller using relays. These relays are controlled by a mechanism that handles the switching between all the cassettes or tools. The setup ensures that when one cassette is selected, all others are first disconnected. After a brief delay of about half a second, the other cassettes are reconnected.
Additionally, the PCB is designed to be universal for any controller or tool configuration. It allows you to set an ID for the station, so the controller can recognize which tool is being selected. I also integrated sleep mode and button detection into the PCB. The button detection works in a somewhat unusual way: it involves pulling a pin to ground. This is achieved with a 10-megaohm resistor and some transistors acting as switches to detect when the button is pressed, which results in a logic low signal.
I encountered a problem where, for some reason, pressing the button wouldn’t trigger the desoldering iron. I started troubleshooting with an oscilloscope and found that the system only started working correctly when I touched the ground. I still don’t fully understand why it behaved this way, but it was happening simultaneously with the issue of the cartridges glowing red.
After the issue seemed to resolve itself, the button detection also started working correctly. One possibility I’m considering is that a component might have been damaged or leaking, causing some erratic behavior. This could have led to spikes or other issues, but for now, the problem seems to have cleared up.
Conclusion
In conclusion, this project has been an incredible journey, and I truly enjoyed working on it. It took me about a year to complete everything, and I’m excited to see how well it serves me moving forward.
One of the aspects I’m particularly proud of is the modular design of the cassette system. It allows for any handle and configuration, and with the same PCB, you can use various tools such as soldering irons, desoldering tools, and hot air tools. You can even stack them in parallel. While not infinitely, the system is designed to handle multiple modules with minimal noise and maintain functionality.
I also want to touch on the costs and compare them to a commercial JBC station. One of the major benefits of this DIY system is its affordability. For instance, even though genuine JBC tips can be expensive, I found that tips from AliExpress are quite affordable, with some costing as little as $10. The original tips are great, but the cheaper alternatives work just as well for my needs.
Regarding other high-end systems like Metcal, while they offer excellent performance with their frequency heating technology, they lack the configurability that my DIY system provides. Metcal tips might perform better, but you’d be stuck within their ecosystem. JBC, on the other hand, is a fantastic system. Despite the need to change tips with two hands, it’s incredibly fast and efficient. The quick tip change feature is one of my favorite aspects.
I’m also pleased that Sparky’s design allows for such flexible tool management, and I’m looking forward to integrating everything into a fully functional setup where I can easily switch between the desoldering iron and other tools. The modular and customizable nature of this system has been a joy to build and use.