Are there any special requirements for soldering an Artemis module to another board? It seems to me like it could be more complicated than a typical part being soldered to a board.
I use a toaster oven reflow machine, and my control of the temperature profile will not be as good as a real reflow machine that you must be using in your production flow. Mostly, I am worried that my sub-optimal reflow temperature profile will potentially mess up the solder joints between the Apollo3 BGA pads and the Artemis module itself. If that happens, then both boards become trash. Also, I put down the solder paste by hand. How sensitive are the module pads to having variations in the amount of solder on the pads?
I figure that the Sparkfun folks must have gained some good experience constructing the Redboards. Are there any tricks to be aware of?
I’ve done a fair bit of module reflow soldering with a toaster oven and a pid controller and have found no complications. I don’t let the temp go above 217C and hold it there for about 10 seconds. The modules I’ve used are cypress PSOC ble modules. I’m about to do some with Artemis modules in the next 10 days or so.
I use a stencil for anything this small for sure. Others might not need it. I do a lot of reflow in my little ~$50 toaster oven and have always been amazed at how easy it is. It is much easier than hand soldering fine pitch through hole not to mention a heck of a lot faster. And stencils are so cheap from China it’s practically free (I use JLCPCB)
I use JLCPCB too, but have not graduated to their stencils yet. I still use a solder syringe and a microscope. It works, but it is gets tedious when I want to make more than one copy of a board.
I use a Rocketscream PID controller for my toaster oven. Their documentation said that the default PID parameter settings were not too sensitive. I decided to log the output of the controller’s decisions during a run and discovered that “not too sensitive” was an overstatement: the PID functionality is essentially unused, at least for my toaster oven. It’s heaters are on 100% for the whole time until the interior reaches the soak temp, then basically off for the entire time during soak, then on again the whole time when reaching reflow temps, then off for good as soon as it hits reflow. There was almost no time during the whole run where the PID controller made any decisions that would cause the heating elements to cycle in any fashion to maintain a temp: just all on or all off. It seemed to me that the controller software could have been replaced with about 7 lines of Arduino software to turn a relay on or off with time delays in between. The bottom line is that the oven works, but I’m not convinced that its temp profile is really all that well controlled.
Again, I would be really interested in knowing if your soldering experiments turn out well!
I’ll keep you posted. My PID controller isn’t much better than yours but seams to have the opposite problem. It is far too gentle, I tend to control the temp by hand (opening and closing the door slightly and just watch the set temp…
Just to close this thread out: I have now soldered down two Artemis modules using a toaster oven. Both tries were successful. I placed the solder by hand for the first module using the tip of xacto knife and a microscope. That was a difficult job, but not impossible. I guess it might have taken me 30 mins to do all the Artemis pads. The second module, I used a stencil. It took about 1 second to wipe on the solder. The result looked great. The stencil sure seems the way to go.
I did alter my toaster oven software before doing the soldering. I got rid of all the PID control because it seemed to be a lot of complication that didn’t work very well. The PID controller was letting the interior start cooling again before starting the reflow phase, which seemed bad. What I found worked well for my 4-element toaster oven using a no-clean leaded solder formula paste:
- apply full power until the interior hits 100F, then shut the power off
- Let the oven temp coast upwards as the elements cool down
- When the interior temp coasts up to 120F, I turn on full power again
- keep power on until the temp hits 220F, then power off and open the door
I am attaching a pic of the resulting reflow profile.
Bottom line: Artemis modules appear to be solderable just fine using a toaster oven.
Hi all, I’m wondering if anyone’s done this w a handheld heat gun?? I’m wondering if the inside pads get hot enough to reflow or if the fr4 acts as a heat shield… I basically already have too much stuff on my lab bench and am trying to avoid a toaster…
Your toaster oven should live in the garage anyway, especially when it is running!
I made an Artemis Redboard shield that implements a toaster oven controller. My actual goal was to use the proven design to teach a class on how to make circuit boards (https://www.island-underground.com/pcb-class). The class project would be for students to design their own reflow controllers based on the original known-working design, which would leave them with a reflow machine for their own future projects. And then, the pandemic. But I do have a working controller design and some spare PCBs. If you want one, I’ll send you one. Come to the dark side Luke…
hmm, thx robin- I just talked w the guy I work for and asked him about buying a toaster, but he just went ahead and bought a full reflow oven instead… anyways thx for the advice…
also- if you’re still having issues w that lora antenna, I have access to vna’s and stuff to match it…
For the record - I use a cheap toaster oven without a controller (it’s got a temperature setting knob which makes me feel slightly better) and it does the job.
There is some valuable discussion on soldering the Artemis here (including by hand with a desoldering heat gun)
Per the link above, I’ve done plenty of manual Artemis swapovers (although at this stage mostly with leaded solder) - the amount of heat that needs to go in will guarantee that everything on the module melts, so you need to be very careful not to disturb anything. I’ve only lost 2 (out of probably 20) though.
