My first idea when i saw this: "why not add an extra layer to any pcb to make it reflow itself". Turns out you already thought of it. Great video. Super interesting concept.
Haha technically it was also how the idea started but I decided to first valid the concept with a hotplate.. But now that it worked I 'm just to excited for the self Reflowing pcbs! 🙃
@@CarlBugeja Well, you made useful experiences. I am interested to see the use case for the self reflowing PCBs. Surely it can not be sustainable for large scale manufacturing, but maybe a good way to spend a unused free layer at PCBway ;)
@@danielaaroe The application is that as a hobby player you might not have a reflow oven of your own for that one pcb per year you might want to make. Which is a pretty serious limitation, for many components it's the only reliable way to solder them. LGA, BGA packages, you can try zapping them with a heat gun and try eyeballing it, but the success rate is not good. In professional setting it's not useful at all of course, you'll just buy the tools you need. Even in home setting if you are willing to buy any supplies getting a bottle of vapor phase reflow fluid would probably solve your issues. But if all you have is some paste, board and components, then it would be pretty neat if the board would solder itself.
If you're making the pcbs separate then connect to the "heater" with screwed-in wires rather than needing to solder onto it, it'll eliminate the chance of that solder melting too!
Don't sleep on the possibility of an edge connector. No soldering on the hot board required, easy attach/detatch. Could even have a few extra pins on it to include the thermocouple and specify heater type and have the µC adjust its voltage/current parameters for each heater.
Exactly what I was thinking. Let's heat the control circuit so it doesn't work anymore!! Most semiconductors fail catastrophically at 150c. They stop working around 100c junction temp. Defo a stupid idea to have the control electronics on the hotplate !!
U should get rid of the holes. They are counterproductive since they lower thermal mass and heat is distributed unevenly to reflow PCB. Anyway good project as always !
@@micultimy91 instead, maybe create a lever that lifts the board up after heating: it could poke the board up from a few tiny holes from below (little metal toothpicks that push thru the heating element).
Set Zones In the next iteration. So you're controller can use only smaller sections depending on size of PCB to be reflowed. You might be able to use a PCB on edge as Underside Supports For your "heater" pcb. Past all that this is such a simple design I hope it works out.
A thick layer of anodizing is often used to create the dielectric layer of an aluminum PCB.... which is something I find to be pretty interesting. The copper traces can also be added to that dielectric layer via either electroplating or sputtering techniques to avoid having any polymer resins that would lower temperature ratings.
Great idea! A cool addition would be to split the board into heating zones. You could control which zones receive power using transistors or something like that. That way you don't have to heat the whole thing for a tiny PCB. It would both speed up the heating speed, limit the power usage and make the thing a bit safer.
Thanks for including the part number of that thermocouple amplifier IC! Haven't used thermocouples in any of my projects yet, but thanks to you I now have a starting point for putting something together 🙂
Hey Carl, great video. This went exactly where I thought you were taking it in the last episode. I'm glad it worked, and agree with your planned upgrades. The warping could be mitigated with pilot holes for bolting it to a test bed (akin to a motherboard test bench) that way the middle can have some support while heating. Keep up the great work, and I am glad you are doing well.
Me, who can't even solder two wires with a soldering iron: "Huh, interesting." Edit: Carl, your videos are super cool. Could you make a video about basic electronic engineering or electronic engineering tips? Thanks for the heart btw.
Soldering is almost an art, and there are tricks to do it well. If you are going to start, I reccomend you dont use one of those just plug in ones. Get a variable temperature iron, you wont regret it. When I first tried, I was absolute crap. I didnt clean the boards and every single solder joint was a cold one, and I kept lifting the pads because the temperature was too high, but I couldent adjust it because I was using a cheap plug in one!
@@justin.campbell exactly ! temperature control is a biiig deal, also for me soldering flux changed absolutely everything, I went from struggling to solder some wires to soldering up a perfect PCB with just the right variable iron and a jar of soldering flux.
I'm not sure if anyone suggested it but add PCB legs. You can plot small notches for the legs to slide into with a small amount of room for expansion. I would put them in a cross configuration to support the center. This should help with the sagging. It will also get the heat off the surface you are working on and gives you a place to mount the electronics. Maybe a faceplate for temp control and screen. Then you can also plot a small hole in the center of the heating surface to put a thermocouple into the center from the bottom. Additionally, with the 3d concept, add a PCB arm on the top with a screw pivot and a point like a finger that aligns with the center. Add ample areas on the pivoting finger to blob on solder for weight then you don't have to tape the PCB down.
Idea for version 2: Insulation. Add some refactory material and a housing around the PCB heater so heat radiating off it is reflected back like a toaster oven.
Excellent heating tool is the plasic pipes heat welder plumbers use.Desoldering led from backlight stripes,reballing bga chips and reflows small sizes pcb.And its price is very affordable.
Suggestion #1: Build an FR4 Frame around it to mount the components on and suspend the Al-core in the center. [X] #2: the slots in the Al-core are not necessary, the heat conductivity of the aluminium is good enough to cool it from below, mount a fan underneigh. #3: make your layout more flexible, to compensate for manufactoring errors or to allow different supply voltages you could allow to modify the resistance by separating/shorting tracks via (high temp) solder-jumpers
Hello Carl, Nice project. 1. My idea is to implement fan/s to cool the PCB board after reflowing. You can installed from below and connect to controller who can turn On the fan/s after finishing the reflowing. After the dropping the temperature to about 30°C it can stop automatically. Just nice future which I will like it to have. If the fan/s is installed from bellow is good to have more holes on the heater ( like many bigger vias). 2. I think that you holes are too wide and too long and heat maybe will not distribute evenly. Maybe is good to be 1-2 millimeters wide and about 10 millimeters long. 3. Other think is I like to have real time reflow profile on the computer screen or similar. I like to have loading/edit/run reflow profiles thru the network LAN or Wi-Fi. 4. I s good to have two surfaces for small PCB and big PCB boards. Lets say into the big surface you can separate one area about 4x4 inches for small PCBs. You can switch between areas small or full. This is not too much related to the saving of the electricity but to protect your finger from burning when handle the boards:). If you have cooling fan not burnet fingers:)). Let me know what do you think about it.
Bit of an off-topic comment, and I know this is months old now, but this reminds me of that PCB CNC machine. I don't remember the name of it or who makes it, and I'm not having luck finding that information currently. But it 'print' a circuit, using a proprietary paste, onto a PCB blank, typical fiberglass or flexible materials, and would bake the paste into traces; iirc, it could also drill for vias, but I would have to fact check this. I remember arguing that the machine was overpriced, the proprietary paste was most definitely overpriced, and that it would be fairly easy to DIY such a machine, assuming you could attain a similar paste, as the paste extruder was just a plunger mechanism, iirc, or possibly a blocked screw extruder. This reflow plate, and using a 3D printer as the CNC, and with a different tool being the paste extruder, proves that my concept of a DIY variant of this machine is, in fact, possible, and quite easy to accomplish; only problem, again, being the paste. The only difference, iirc, in both the original and DIY machines, would be that the pasted board needs to be facing the hot plate, with a holding mechanism that also forms a cavity, so that it creates what's essentially an oven, so that the paste properly bakes. Of course, these machines would not be close to a true copper trace PCB, but it would be better than, say, a conductive 3D printer filament, as it's a more permanent, baked-on solution. I believe this PCB-making machine would also be able to reflow solder, as the baking temperature is much higher than solder reflow temperatures, making it a multi-use machine. Further this by also incorporating a pick and place machine, have a tool changer, and a separate non-heated bed for component layout, though this would be a hybrid manual-automated machine, as the machine wouldn't be able to select components without a predetermined layout or without machine learning and artificial intelligence to pick random components from random placements to be placed in correct orientations and board placements, or have laser cut part arrangement boards to make pick and placing simpler; alternatively, for home production machines, the possibility of a tape reel dispenser in the pick and place head. Essentially, one idea could turn into a home factory in one machine, after initial prototyping is done of course. An ambitious idea, yes, but it also seems entirely possible to do, as each separate component does already exist, it would just be combining them all into a single machine, as tool changing in this form can exist, and the 'dual bed' design would just be an extended axis, which does also exist; so what would stop this from not being feasible, other than the solder paste issue? I've seen so many project products fail to come to fruition, or become a very niche product at most, because everything is hand assembled, and in the orders of hundreds of products, this takes a long time with a single person or small team. A combination machine like this, a desktop factory, would expedite this process, at least in bulk time as it can run continuously, and a print farm would further expedite the process. This would also be a direct link between rapid prototyping and production, something not often seen in the CNC world. It makes sense for a machine like this to exist for small-scale production, done out of someone's hobby room or small business.
If you make a self-reflowing PCB, you could have a tiny LED or something on the end that only lights up if the entire circuit has reflowed properly. Or maybe even a resistor connected to the thermal regulator so it knows to switch to the main microcontroller from now on. That way it automatically reflows the first time you turn it on and then it's done. You can pass power through the circuit when heating too, so maybe it's possible to do an "If isComplete applyCoolingAndEnd(); Else keepHeating(); waitUntilReflowComplete();" so that you can automatically stop it at just the right time before you get bridges
You know what they say, you learn something new every day. I learned today that the guy from the cap and ball channel has a second channel and likes electronics.
For rigidity I would add something mechanical as a cross brace, to give it either a U or a T shape cross section. thin aluminum in an L shape and rivets would be a very stiff and low effort solution, but would wick away some of the heat. If its too much heat you'd have to find something that could be attached (rivets or otherwise) with a low heat transfer coefficient.
It might be interesting to have several heating elements that could be controlled individually so that you could optimize a consistent temperature across the whole board.
Or other thermal insulation that can tolerate the heat and is handy. (I hate working with glass wool.) This is assuming the system can match the thermal profile with the insulation. Too much insulation, and it won't cool fast enough. If it doesn't cool fast enough, then you have to add a cooling fan.
If you make a self reflowing PCB and add some sort of vibrating actuator (that is glued on and controlled away from the self-reflowing section) you could have a self depopulating PCB. Might be useful for semi-secure devices.
You could divide the long track into maybe 4 tracks and make a pad pattern that lets you change the wiring (all in series/ all in parallel/ series-parallel combination) so you can tune the resistance and experiment better!
Consider also that aluminum and pcb have different expansion coefficients... Which might lead to bending. (I think that removing holes facilitates the rise in temperatuee as it looks less like a radiator)
i use an electrical induction stove and an steel plate and works marvelous it has temperature sensor and is quite accurate maybe u can make a similar concept but in small format
with such a slow profile, you could probably use just a proportional heater, maybe with a bit of differential control, but that should be fine, and reduce the whole PID tuning, and code size !
I think the aluminium backing is acting as a Big heatsink, pumping and dissipating most of the heath on the back of the plate. Maybe try removing the slots and put the pcb to be reflowed on the aluminium side? And add some insulation on the other side to focus the heath to the reflowed pcb?
the spec sheet said Vg threshold is 3-5 , so the mosfet will turn off at a max voltage of 5 V and a min voltage of 3V, so 5 V logics is unliekly to work , you could get better mosfet like IRF150 or IRF540N , because 2-4 Vg threshold wil work better with 5V logics.There are also mosfet with logic input like FQP30N06L or IRL serie. the could go as low as 1 - 2.5 , which could also work with 3.3V logics. Use the right compoenet could save so much trouble
My first impression... Get some nichrome wire and lay it out in a nice spiral on a copper plate maybe 3mm thick... Use a layer of fiberglass and epoxy between the copper and nichrome...
You could put a few small holes in the pcb and design a small clip or use wire to hold the thermocouple onto the pcb. Then you dont have to use the tape.
Probably it's a bad idea but i wonder if it's possible to make a PCB with a grid of very fine heat cells, like a TFT screen but cells produces heat in stead of light? I don't know if it's possbile or achievable but i think it would be great to aim directly to the soldering areas with a high accuracy. It would prevent components from damage and it makes the process more managable.
You could take a look at the E3D high temperature heatbeds, they go to 200°C with a max of 250°C. Maybe cheaper to just make control for that than trying to remake the huge PCB as well.
Try an alu (or copper if you can afford it) plate on top as a heat spreader. Should allow the heat to spread across the current gaps and provide more thermal mass. Just be sure to tweak your PID values after! :)
How would it behave if you put 2 panels face to face, and powered both (using the aluminum back as the heater surface) Wouldn’t that be the same, resistance-wise, as cutting the board in half?
It would be interesting to integrate a thermocouple in the PCB or even use the PCB as one of the 2 metals. A possible option would be a type T (copper-constantan).
Like your bench & table set up! Great video! Have you ever tried to make a flow/reflow surface using quartz heaters, such as found in some toaster ovens? Thanks
If you change from the default (for aluminum backed PCBs) ENIG to HASL (lead free), it's cheaper on PCBWay's website. I'm wondering if it would be better to specify black soldermask, given this is essentially a heating element.
Temperature sensing suggestion- could you use heating element resistance as a temp sensor? Will probably need 4 terminal sensing, but should be possible. Also - split the heater into sections with separate PWM. Will make it easier to compensate for etching variability across the heater.
I thought about this - monitoring the current and measure the resistance change with temperature.. In theory it should work but you have to consider other stuff like variations in track resistance.. But I think it still can be done if the current change is constant.. So I might consider this for v3 😜 once v2 works haha
@@CarlBugeja Might be possible to measure track resistance at room temperature per-board, and then assume a fixed temperature coefficient. Track width may vary, but aluminium expansion coefficient and copper tempco should be consistent across PCB batches. Best of luck for v2!
If your firmware uses floating point math, you can regain a lot of memory by switching to integer math and avoiding the floating point library. It's a pain but it works, and once it works, it's done.
Is this the same Altium that every time you try to use the help file, you either get information from an old version of the program which doesn't work like that anymore, or you get a message saying the help for that feature is 'coming soon'? The same software that won't even let you do a simple print of the current view without compiling an output instruction, even if you only want a simple print?
Allium designer is fantastic and my preferred tool but industry standard not quite. I'm stuck with Pads standard plus it's alright I would kill for an Alltium licence...
If you mount your self-reflowing PCB upside down you can make it drop the (heavier) components if the circumstances require it. No, I have no idea why the circumstances would require it. A stupid kind of fuse maybe. :D
Could patrons eventually have access to the finish schematic to make it themselves? Because, an typical hot plate for reflow is more expensive and much much bigger.
Another fantastic idea! Good use of inner layers & (soon to be) ground "planes". Ever figure out how the resistance was so far off of your estimate? Was it the sharp u-turns?
@@CarlBugeja I think you liked that X-ray machine and want to get to the bottom of this {}. Best of luck with your quarantine. Thank you for actually doing it. Oh, if you can get sides on your heater, it might help keep convection from "reaching" the target (perhaps slots in PCB for walls).
Ohh ohh ohh, make a circuit board that is self soldering. Have coil layers under the components that you power up to melt the solder. I know it's useless for mass production shiz but it would be good for small runs and one offs for people who don't have a reflow oven. Oh wait, just finished the video, your already on it. Good job.
Hi! Very good idea!. What you need to do is a PCB with a printed pancake coil. Then, make a driver so that everything works as a heat inductor. Place the PCB side with the printed coil facing down, and attach a thin stainless steel sheet connected to the ground on the top side. This will heat up and dissipate electromagnetic waves, thus preventing the components you are trying to weld from being affected. Hola! Muy buena idea!. Lo que necesitas hacer es una PCB con una bobina tipo panqueque impresa. Luego, fabricarle un driver para que funcione todo como un inductor de calor. Coloca la cara de la PCB con la bobina impresa hacia abajo, y anexa del lado de arriba una hoja de acero inoxidable delgada conectada a masa. Ésta se calentará y disipará las ondas electromagneticas, evitando asi que sean afectados los componentes que intentes soldar.
My first idea when i saw this: "why not add an extra layer to any pcb to make it reflow itself". Turns out you already thought of it. Great video. Super interesting concept.
Haha technically it was also how the idea started but I decided to first valid the concept with a hotplate.. But now that it worked I 'm just to excited for the self Reflowing pcbs! 🙃
@@CarlBugeja Well, you made useful experiences. I am interested to see the use case for the self reflowing PCBs. Surely it can not be sustainable for large scale manufacturing, but maybe a good way to spend a unused free layer at PCBway ;)
Imagine a chip dies and you just power the reflow heater and replace it with no tools needed
@@alexstone691 of course, especially if there were reflow-zones(to avoid reflowing all the other joints and creating unnecessary oxidization)
@@danielaaroe The application is that as a hobby player you might not have a reflow oven of your own for that one pcb per year you might want to make. Which is a pretty serious limitation, for many components it's the only reliable way to solder them. LGA, BGA packages, you can try zapping them with a heat gun and try eyeballing it, but the success rate is not good. In professional setting it's not useful at all of course, you'll just buy the tools you need. Even in home setting if you are willing to buy any supplies getting a bottle of vapor phase reflow fluid would probably solve your issues. But if all you have is some paste, board and components, then it would be pretty neat if the board would solder itself.
If you're making the pcbs separate then connect to the "heater" with screwed-in wires rather than needing to solder onto it, it'll eliminate the chance of that solder melting too!
Or tabs you can fit a spade staycon on
he could also just use standoffs to connect the heater to the power supply
@@eliasrenner555 Yeah, really anything that doesn't require a soldered connection
Don't sleep on the possibility of an edge connector. No soldering on the hot board required, easy attach/detatch. Could even have a few extra pins on it to include the thermocouple and specify heater type and have the µC adjust its voltage/current parameters for each heater.
My first impression: Let the heater reflow it's own circuit
Exactly what I was thinking. Let's heat the control circuit so it doesn't work anymore!! Most semiconductors fail catastrophically at 150c. They stop working around 100c junction temp. Defo a stupid idea to have the control electronics on the hotplate !!
U should get rid of the holes. They are counterproductive since they lower thermal mass and heat is distributed unevenly to reflow PCB. Anyway good project as always !
Or minimize them
It seems to me that they'd also lead to much more convection, which is probably draining the heat away.
Why are there holes in the first place? Just curious
@@TheRainHarvester holes are for rapid cooling of the pcb after heating. check the soldering standard
@@micultimy91 instead, maybe create a lever that lifts the board up after heating: it could poke the board up from a few tiny holes from below (little metal toothpicks that push thru the heating element).
Set Zones In the next iteration. So you're controller can use only smaller sections depending on size of PCB to be reflowed. You might be able to use a PCB on edge as Underside Supports For your "heater" pcb. Past all that this is such a simple design I hope it works out.
A thick layer of anodizing is often used to create the dielectric layer of an aluminum PCB.... which is something I find to be pretty interesting. The copper traces can also be added to that dielectric layer via either electroplating or sputtering techniques to avoid having any polymer resins that would lower temperature ratings.
Great idea! A cool addition would be to split the board into heating zones. You could control which zones receive power using transistors or something like that. That way you don't have to heat the whole thing for a tiny PCB. It would both speed up the heating speed, limit the power usage and make the thing a bit safer.
I am super excited to see how the self-soldering PCB works out!
Thanks for including the part number of that thermocouple amplifier IC! Haven't used thermocouples in any of my projects yet, but thanks to you I now have a starting point for putting something together 🙂
Hey Carl, great video. This went exactly where I thought you were taking it in the last episode. I'm glad it worked, and agree with your planned upgrades. The warping could be mitigated with pilot holes for bolting it to a test bed (akin to a motherboard test bench) that way the middle can have some support while heating.
Keep up the great work, and I am glad you are doing well.
Great concept! This could be a great tindie product! Excited for version 2 too!
Me, who can't even solder two wires with a soldering iron: "Huh, interesting."
Edit: Carl, your videos are super cool. Could you make a video about basic electronic engineering or electronic engineering tips? Thanks for the heart btw.
Soldering is almost an art, and there are tricks to do it well. If you are going to start, I reccomend you dont use one of those just plug in ones. Get a variable temperature iron, you wont regret it. When I first tried, I was absolute crap. I didnt clean the boards and every single solder joint was a cold one, and I kept lifting the pads because the temperature was too high, but I couldent adjust it because I was using a cheap plug in one!
Thank you!
@@justin.campbell exactly ! temperature control is a biiig deal, also for me soldering flux changed absolutely everything, I went from struggling to solder some wires to soldering up a perfect PCB with just the right variable iron and a jar of soldering flux.
@@MrOmarabdulhadi yes, I agree. flux is really just absolute magic the first time you use it!
I'm not sure if anyone suggested it but add PCB legs. You can plot small notches for the legs to slide into with a small amount of room for expansion. I would put them in a cross configuration to support the center. This should help with the sagging. It will also get the heat off the surface you are working on and gives you a place to mount the electronics. Maybe a faceplate for temp control and screen. Then you can also plot a small hole in the center of the heating surface to put a thermocouple into the center from the bottom. Additionally, with the 3d concept, add a PCB arm on the top with a screw pivot and a point like a finger that aligns with the center. Add ample areas on the pivoting finger to blob on solder for weight then you don't have to tape the PCB down.
This is a really great idea! I look forward to part 2
Idea for version 2: Insulation. Add some refactory material and a housing around the PCB heater so heat radiating off it is reflected back like a toaster oven.
This is one of my top ten youtube channels, thank you for your time to make these videos!
Man, that’s such great use of existing technology in a novel way. Great job! 💜
I was thinking about turning a broken clothing iron into a PCB reflow device.
This video helps out a lot.
Excellent heating tool is the plasic pipes heat welder plumbers use.Desoldering led from backlight stripes,reballing bga chips and reflows small sizes pcb.And its price is very affordable.
Suggestion #1: Build an FR4 Frame around it to mount the components on and suspend the Al-core in the center. [X]
#2: the slots in the Al-core are not necessary, the heat conductivity of the aluminium is good enough to cool it from below, mount a fan underneigh.
#3: make your layout more flexible, to compensate for manufactoring errors or to allow different supply voltages you could allow to modify the resistance by separating/shorting tracks via (high temp) solder-jumpers
I'm looking forward to V2. I'm doing all my SMD soldering by hand, and would love something like this!
Hello Carl, Nice project. 1. My idea is to implement fan/s to cool the PCB board after reflowing. You can installed from below and connect to controller who can turn On the fan/s after finishing the reflowing. After the dropping the temperature to about 30°C it can stop automatically. Just nice future which I will like it to have. If the fan/s is installed from bellow is good to have more holes on the heater ( like many bigger vias). 2. I think that you holes are too wide and too long and heat maybe will not distribute evenly. Maybe is good to be 1-2 millimeters wide and about 10 millimeters long. 3. Other think is I like to have real time reflow profile on the computer screen or similar. I like to have loading/edit/run reflow profiles thru the network LAN or Wi-Fi. 4. I s good to have two surfaces for small PCB and big PCB boards. Lets say into the big surface you can separate one area about 4x4 inches for small PCBs. You can switch between areas small or full. This is not too much related to the saving of the electricity but to protect your finger from burning when handle the boards:). If you have cooling fan not burnet fingers:)). Let me know what do you think about it.
Bit of an off-topic comment, and I know this is months old now, but this reminds me of that PCB CNC machine. I don't remember the name of it or who makes it, and I'm not having luck finding that information currently. But it 'print' a circuit, using a proprietary paste, onto a PCB blank, typical fiberglass or flexible materials, and would bake the paste into traces; iirc, it could also drill for vias, but I would have to fact check this. I remember arguing that the machine was overpriced, the proprietary paste was most definitely overpriced, and that it would be fairly easy to DIY such a machine, assuming you could attain a similar paste, as the paste extruder was just a plunger mechanism, iirc, or possibly a blocked screw extruder. This reflow plate, and using a 3D printer as the CNC, and with a different tool being the paste extruder, proves that my concept of a DIY variant of this machine is, in fact, possible, and quite easy to accomplish; only problem, again, being the paste. The only difference, iirc, in both the original and DIY machines, would be that the pasted board needs to be facing the hot plate, with a holding mechanism that also forms a cavity, so that it creates what's essentially an oven, so that the paste properly bakes.
Of course, these machines would not be close to a true copper trace PCB, but it would be better than, say, a conductive 3D printer filament, as it's a more permanent, baked-on solution. I believe this PCB-making machine would also be able to reflow solder, as the baking temperature is much higher than solder reflow temperatures, making it a multi-use machine. Further this by also incorporating a pick and place machine, have a tool changer, and a separate non-heated bed for component layout, though this would be a hybrid manual-automated machine, as the machine wouldn't be able to select components without a predetermined layout or without machine learning and artificial intelligence to pick random components from random placements to be placed in correct orientations and board placements, or have laser cut part arrangement boards to make pick and placing simpler; alternatively, for home production machines, the possibility of a tape reel dispenser in the pick and place head. Essentially, one idea could turn into a home factory in one machine, after initial prototyping is done of course. An ambitious idea, yes, but it also seems entirely possible to do, as each separate component does already exist, it would just be combining them all into a single machine, as tool changing in this form can exist, and the 'dual bed' design would just be an extended axis, which does also exist; so what would stop this from not being feasible, other than the solder paste issue?
I've seen so many project products fail to come to fruition, or become a very niche product at most, because everything is hand assembled, and in the orders of hundreds of products, this takes a long time with a single person or small team. A combination machine like this, a desktop factory, would expedite this process, at least in bulk time as it can run continuously, and a print farm would further expedite the process. This would also be a direct link between rapid prototyping and production, something not often seen in the CNC world. It makes sense for a machine like this to exist for small-scale production, done out of someone's hobby room or small business.
If you make a self-reflowing PCB, you could have a tiny LED or something on the end that only lights up if the entire circuit has reflowed properly. Or maybe even a resistor connected to the thermal regulator so it knows to switch to the main microcontroller from now on. That way it automatically reflows the first time you turn it on and then it's done. You can pass power through the circuit when heating too, so maybe it's possible to do an "If isComplete applyCoolingAndEnd(); Else keepHeating(); waitUntilReflowComplete();" so that you can automatically stop it at just the right time before you get bridges
*The first line of that video has maybe described more of My days than I may admit*
You know what they say, you learn something new every day. I learned today that the guy from the cap and ball channel has a second channel and likes electronics.
Lol. I literally was thinking about this yesterday :D great video!
For rigidity I would add something mechanical as a cross brace, to give it either a U or a T shape cross section. thin aluminum in an L shape and rivets would be a very stiff and low effort solution, but would wick away some of the heat. If its too much heat you'd have to find something that could be attached (rivets or otherwise) with a low heat transfer coefficient.
It might be interesting to have several heating elements that could be controlled individually so that you could optimize a consistent temperature across the whole board.
Or heat only those regions defined by your smaller project reflow pcb, so saving energy
Self re-flowing PCBs: our robot masters of the future will thank you for your contribution :) Interesting stuff, take care.
🙏 🤖
This is really cool I was thinking of doing something similar to this with super caps from Kapton strips!
Cool project! As sugested I also believe the PCB slots should go away to get a more even temperature across the surface :)
Tip: don't suspend pcb heater in air, support it by glass wool pads it will provide insulation and support
Or other thermal insulation that can tolerate the heat and is handy. (I hate working with glass wool.)
This is assuming the system can match the thermal profile with the insulation. Too much insulation, and it won't cool fast enough. If it doesn't cool fast enough, then you have to add a cooling fan.
Use FR4. Realistically FR4 is designed to go to reflow temperatures !! Look for the higher temperature stuff.
Self reflowing PCB! That would be an amazing project
The folks at PCBWay much be so excited to get his orders.
you gave me a good idea bro.
you are a smart man. thank you.
I thought I was the only one using a 3D printer heatbed as preheater XD
If you make a self reflowing PCB and add some sort of vibrating actuator (that is glued on and controlled away from the self-reflowing section) you could have a self depopulating PCB. Might be useful for semi-secure devices.
if you mount the pcb heater directly you may need to give it some expansion room, it seems that is the source of your bending when it is heating up
You could divide the long track into maybe 4 tracks and make a pad pattern that lets you change the wiring (all in series/ all in parallel/ series-parallel combination) so you can tune the resistance and experiment better!
Consider also that aluminum and pcb have different expansion coefficients... Which might lead to bending. (I think that removing holes facilitates the rise in temperatuee as it looks less like a radiator)
Very Nice Project 👍
I liked it👌
i use an electrical induction stove and an steel plate and works marvelous it has temperature sensor and is quite accurate maybe u can make a similar concept but in small format
with such a slow profile, you could probably use just a proportional heater, maybe with a bit of differential control, but that should be fine, and reduce the whole PID tuning, and code size !
Nice, We expecting more like this...
9:06 What development environment is this?
I think the aluminium backing is acting as a Big heatsink, pumping and dissipating most of the heath on the back of the plate.
Maybe try removing the slots and put the pcb to be reflowed on the aluminium side? And add some insulation on the other side to focus the heath to the reflowed pcb?
the spec sheet said Vg threshold is 3-5 , so the mosfet will turn off at a max voltage of 5 V and a min voltage of 3V, so 5 V logics is unliekly to work , you could get better mosfet like IRF150 or IRF540N , because 2-4 Vg threshold wil work better with 5V logics.There are also mosfet with logic input like FQP30N06L or IRL serie. the could go as low as 1 - 2.5 , which could also work with 3.3V logics. Use the right compoenet could save so much trouble
Well, older rep rap 3d printers had pcb’s as bed heaters though they didn’t reach such temperatures. Good work!
Suggestion for V2. Serial or USB data-logging and control, so you can graph the temperature on the computer, and easily set thermal profiles.
Isn't the heatbed showed at the beginning of the video exactly the same what you had made..? It also has aluminium substrate
Wouldnt it heat more/better if you powered it from a constant current regulator instead of a constant voltage?
Wonderful idea.... looking forward to v2 ...
My first impression... Get some nichrome wire and lay it out in a nice spiral on a copper plate maybe 3mm thick... Use a layer of fiberglass and epoxy between the copper and nichrome...
You could put a few small holes in the pcb and design a small clip or use wire to hold the thermocouple onto the pcb. Then you dont have to use the tape.
Can u make a video or series on how to design pcbs
Check out fedevel academy
Probably it's a bad idea but i wonder if it's possible to make a PCB with a grid of very fine heat cells, like a TFT screen but cells produces heat in stead of light? I don't know if it's possbile or achievable but i think it would be great to aim directly to the soldering areas with a high accuracy. It would prevent components from damage and it makes the process more managable.
You could take a look at the E3D high temperature heatbeds, they go to 200°C with a max of 250°C. Maybe cheaper to just make control for that than trying to remake the huge PCB as well.
That can be done yes... but I still what to design my own! 😅
Try an alu (or copper if you can afford it) plate on top as a heat spreader. Should allow the heat to spread across the current gaps and provide more thermal mass. Just be sure to tweak your PID values after! :)
Dude that's brilliant idea
What microcontrollers you prefer and why
What are J1 and J2 and that plastic thing you removed once the soldering was done?
How would it behave if you put 2 panels face to face, and powered both (using the aluminum back as the heater surface)
Wouldn’t that be the same, resistance-wise, as cutting the board in half?
It would be interesting to integrate a thermocouple in the PCB or even use the PCB as one of the 2 metals. A possible option would be a type T (copper-constantan).
Like your bench & table set up! Great video! Have you ever tried to make a flow/reflow surface using quartz heaters, such as found in some toaster ovens? Thanks
Fantastic Idea!
I know nothing, but would it be useful to split into zones? More efficient, better control perhaps?
If you change from the default (for aluminum backed PCBs) ENIG to HASL (lead free), it's cheaper on PCBWay's website.
I'm wondering if it would be better to specify black soldermask, given this is essentially a heating element.
Hey, nice video. How did you managed the layer stack on Altium for the Al substrate?
Any tips would be appreciated, thanks
Temperature sensing suggestion- could you use heating element resistance as a temp sensor? Will probably need 4 terminal sensing, but should be possible.
Also - split the heater into sections with separate PWM. Will make it easier to compensate for etching variability across the heater.
I thought about this - monitoring the current and measure the resistance change with temperature.. In theory it should work but you have to consider other stuff like variations in track resistance.. But I think it still can be done if the current change is constant.. So I might consider this for v3 😜 once v2 works haha
@@CarlBugeja Might be possible to measure track resistance at room temperature per-board, and then assume a fixed temperature coefficient. Track width may vary, but aluminium expansion coefficient and copper tempco should be consistent across PCB batches. Best of luck for v2!
This sounds like a reasonable idea - cheap soldering irons do this.
Nice project Carl. Why don't you try it with silicone heating pAds which can work with ac mains and an AC SSR and go up to 230c
Maybe you can add type-c connector and usb pd chip for power supply. Like new small hot-plate.
If your firmware uses floating point math, you can regain a lot of memory by switching to integer math and avoiding the floating point library. It's a pain but it works, and once it works, it's done.
Could one take a normal PCB and lay a piece of metal on top of it?
what is the point of the air gaps?
if you want to head as efficient as possible you dont want air to cool the heater...
Hello did you be share the Gerber files,?
That aluminum pcb flexing made me think of a linear pcb actuator. Maybe something for you to try in the future?
Carl Excellent !!
Is this the same Altium that every time you try to use the help file, you either get information from an old version of the program which doesn't work like that anymore, or you get a message saying the help for that feature is 'coming soon'?
The same software that won't even let you do a simple print of the current view without compiling an output instruction, even if you only want a simple print?
You need to go up with the temperature (about 240⁰C) and try to fix the thermocouple on the pcb ( non plated vias?)
i think 240degc would be too much for the pcb to handle and it could damage it
@@CarlBugeja for lead free solder paste peak temperature is about 230-240, if you use sn/pb 200 is ok.
Very interesting, has huge potential for tinkering
Carl, could you share what product is the table mat you have (the blue one?) Thanks a lot and keep doing the great work!
Well done, nice video, thanks for sharing it with us :)
You can try using some peltier module .
Allium designer is fantastic and my preferred tool but industry standard not quite. I'm stuck with Pads standard plus it's alright I would kill for an Alltium licence...
Very great idea !, Nice video 👍👍
Glad you liked it :)
If you mount your self-reflowing PCB upside down you can make it drop the (heavier) components if the circumstances require it. No, I have no idea why the circumstances would require it. A stupid kind of fuse maybe. :D
You looks like the PCB version of "Alex" .. 😋
nice video, thanks for the info
Could patrons eventually have access to the finish schematic to make it themselves? Because, an typical hot plate for reflow is more expensive and much much bigger.
You might consider a layer of heat resistant insulation under the board to minimize heat loss.
Please no slits or holes. Very nice project, would be great as a PCB preheater. You could even divide the whole field into smaller selectable zones.
Any chance for making v2 as a buying product? I'll be glad to have one
if there's alot of interest i might do yes :)
@@CarlBugeja Great videos btw :)
due to known relation of copper resistance and temperature, you may not even need a thermal sensor
REFLOWCEPTION!!!! GL MATE :D
Damn, what a great idea 👍
What software you used to program?
Another fantastic idea! Good use of inner layers & (soon to be) ground "planes". Ever figure out how the resistance was so far off of your estimate? Was it the sharp u-turns?
thanks :) i'm not sure why it was off - the u-turns should have been calculated for
@@CarlBugeja I think you liked that X-ray machine and want to get to the bottom of this {}. Best of luck with your quarantine. Thank you for actually doing it. Oh, if you can get sides on your heater, it might help keep convection from "reaching" the target (perhaps slots in PCB for walls).
Didn't finish in time? Did someone give you a deadline or something?
Ohh ohh ohh, make a circuit board that is self soldering. Have coil layers under the components that you power up to melt the solder. I know it's useless for mass production shiz but it would be good for small runs and one offs for people who don't have a reflow oven. Oh wait, just finished the video, your already on it. Good job.
Hi! Very good idea!. What you need to do is a PCB with a printed pancake coil. Then, make a driver so that everything works as a heat inductor.
Place the PCB side with the printed coil facing down, and attach a thin stainless steel sheet connected to the ground on the top side. This will heat up and dissipate electromagnetic waves, thus preventing the components you are trying to weld from being affected.
Hola! Muy buena idea!. Lo que necesitas hacer es una PCB con una bobina tipo panqueque impresa. Luego, fabricarle un driver para que funcione todo como un inductor de calor.
Coloca la cara de la PCB con la bobina impresa hacia abajo, y anexa del lado de arriba una hoja de acero inoxidable delgada conectada a masa. Ésta se calentará y disipará las ondas electromagneticas, evitando asi que sean afectados los componentes que intentes soldar.