Experimenting with High Temperature PCBs!
Вставка
- Опубліковано 5 жов 2024
- Testing SH260 (Tg=250°C) and Polyimide (Tg=290°C) High Temperature PCBs!
Get a free trial of Altium Designer - carlbugeja.com...
PCB Manufacturer - www.pcbway.com
Help me make more projects by supporting me on Patreon - / carlbugeja
This project is Open-Source and licensed with Creative Commons Attribution Share Alike 4.0 International License
My Electronic Kits - flexar.io
Tools I use: (some of these are affiliate links)
Main Camera - amzn.to/30tRgY7
Second Camera - amzn.to/2ZLm2wy
Light Setup - amzn.to/3jtqckA
Heating Plate - amzn.to/36qYzD4
Ultrasonic Cleaner - amzn.to/3nfUGHY
Magnifying Lamp - amzn.to/3jrcaQv
Vernier Caliper - amzn.to/3eTju3I
Multi-meter - amzn.to/3hq6B2R
Digital Milligram Scale - amzn.to/30HIw1X
Digital Microscope - amzn.to/3hyzoFt
3D Printer - shop.prusa3d.co...
LCR Meter - cutt.ly/Wlmg0b0
Power Supply - amzn.to/3u4umF7
Social Media:
UA-cam Channel - / carlbugeja
Instagram - / bugeja.carl
Twitter - / bugejacarl
Facebook - / bugeja.carl
Linkedin - / carl-bugeja-0b922a135
Website - carlbugeja.com
Open Source Files: github.com/Car...
Music:
We March Together - Patrick Patrikios (UA-cam Audio library) - Наука та технологія
Who is reflowing for 3 hours? Just slap a controller onto it that cuts the heat after some time. Please don't give up!
How long do people reflow for?
i thought the same!
Longetivity tests.
If it works for 3 hours In the lab, it’ll work for 99.9% of home use cases.
It’s called over-engineering and end-user proofing!
Decoloration is the last of my worries!!
@@BleughBleugh "End user proofing" hehe nice
The ‘applied science’ UA-cam channel just did this video recently: “Underwater laser cutting and silver sintering to make ceramic circuit boards”
Since you are thinking of going ceramic, you should take a look at that video.
Breaking taps also did an excellent video on the same topic, well worth checking out.
Always nice to see how you learn from your "mistakes" and how you share it with the world! Keep u doing the wonderful work 🙂 Very interesting concepts!
the bulging has little to do with the adhesive, its just the thermal expansion of the material. It being 'clamped' in the metal frame. If you leave it floating in the frame you should have less to no bulging.
A frame that tensions the flex pcb like a trampoline should work well with the expansion.
@@FreeOfFantasy it's cool because you first get the PCB in one shape but over time it becomes a shapeshifter.
Kudos to pcbway. - Much respect for supporting your customers when they are trying something new.
Right?!
I'm sure its been suggested but you hang get an aluminum one without a solder mask (you could later apply one if needed High temp spraypaint). The only thing that seems to degrade is the mask. Also when using a heater it is useful to have an insulator on the bottom.
I guess the issue with no solder mask might be shorts via whatever PCB you put on top to reflow. How about a solid sheet of copper on the bottom at VCC, a solid sheet on the top at GND, no solder mask on the top, and use vias between them as the resistive paths to generate the heat? That would solve the shorting problem as the whole top is at GND, though it might be harder to calculate the resistance of a network of vias rather than a single trace, and of course it might be hard simply to get vias with a high enough resistance at any drill size. What you could do is simply have a short narrow trace on the bottom between each via and GND, which is actually generating the heat, but that heat is well transmitted through the via to the top.
Certainly adding an SMD temperature sensor to the middle of the PCB seems like it would be useful, and a time-out shutoff of 20 minutes or whatever should be more than enough anyway
Isn't there an FR4 spacer between the aluminium and the copper? Epoxy fiber glass doesn't really like high temperatures, especially over longer periods of time
Exactly. At the voltages this works at baking oven paint should work well enough.
Put the traces on the bottom of an aluminum pcb. The aluminum is such an excellent conductor of heat that it should be plenty hot on the untraced top
@@peterrhodes9121 At that point, what do you even need the PCB for?
You could try making pcb hand warmers as they need much lower temperatures.
I could definitely see something like a top of the hand warmer for PC gaming in winter time.
There's a growing market for lipo powered heated garments for horse riding, and, I imagine, cycling.
@@56Seeker Motorcycling as well, they already sell heated jackets, pants and gloves, but slip in solutions would be cool.
the heated handlebar on ski-doo are some sort of polyamide PCB.
Dude I had that problem at my job. We called it bubbling on PCBs. We put them in a oven overnight to evaporate and dry. We added a ceramic plate over the heat lamps that was a combination of ceramic and another side a different material.
Definitely try the ceramic, it’ll work better.
Lower temps would work well for phone and tablet repair to heat up glue before removing the screen
This!!!!!!!!
3:46 You can measure the average temperature of the entire cooper track by measuring the change of its resistance. There's a normalized method that is used to measure the resistance of transformer windings. But I don't know how good would this method be in this case, considering the low resistance of the track.
look at bens video on applied science about ceramic pcb's hes making them cheaply
Great work, keep it up! I believe this is still viable if you implement those time limits and add temperature feedback control. Generally, as a hobbyist, you shouldn't need the hot plate to stay hot for more than a few minutes at a time
I think this would be great to have around for 1 off reflows. I don't do enough work in this realm to justify a hotplate but this would be great even with the constraints. I'm currently re-flowing with a heat gun...
I suggest adding
a) some kind of thermal-isolating/IR-reflecting chamber to reduce ambient loss
b) temperature control to prevent overheating of the copper layer
Yay new Video! You made my day 😃
Hey man, for the delamination problem it might help not to make such long straight lines. Copper expands and this creates force then upwards away from the pcb. Outside its cooler and it still sticks so you create that bubble. Try more a zigzag shape then the expansion is distributed to both axis. :)
I wonder if they can tell you how cumulative the damage is? How many times can you heat it to 200 degrees for 10 minutes? If the answer is 3 or 4 times, then that’s obviously not good enough. But maybe it’ll last a *lot* longer in 10 minute bursts? Or maybe cycling will make it die faster due to expansion/contraction. Seems worth testing if you have spare boards that you can’t really use.
I love these videos because they are such strange and interesting ways to use PCBs!
That the rating is time related makes perfect sense, given the way those sorts of ratings are normally stated.
Most 3D printer beds like this use the metal surface as the bed. How hot does the stiffener get? How about insulation on the trace side, and then measure the temp of the stiffener?
Exactly. Much better to do it that way anyway because you want even heating in your components. Polymers usually have low thermal conductivity and I would think you have a higher likelihood of cold joints on a large component or unevenly populated board.
Thanks for the update, even though the news isnt what we wanted, it's awesome to see the progression.
Dont giveup, this was a good idea for a new product. (A small coffee cup version is handy too.)
I would like to see this with some insulation around it so that the whole pcb would be the same temperature. and some aluminium foil cover/lid so if that helps with convection.
how about flex pcb heating pads for diy beer wine brewing you could build on board temp control and monitoring, get real fancy and ramp up and down temperatures for optimum yeast activity and results for best taste flex pcbs are ideal as they can be wrapped arro8uund any size fermentation vessel behind some cheap insulation and just think of all free beer ;)
1. Looking at your steel frame, there is no way to expand under thermal expansion. 2. Your polymer will expand much more than the steel frame.
By not placing the 'spokes' opposite to each other (eg. 3*30 deg), and preferably not straight, but under an angle to the heated surface, you will reduce warping a bit, or at least give the steel plate some room to wiggle around and follow the expanding polymer to prevent delamination. Good luck with your project!
Great video :D Electronoobs recently did a diy reflow plate where he used pretty cheap "PTC Heating Plates" from China. Using a PCB to reflow PCBs would be cool but it is maybe just not the best way. Anyways awesome content 👍
Insulate most of the hot plate (ex: the back, and sides). This will reduce the difference in temperature between the inside and the surface at steady state. Then if you bring it up to temperature slowly it might be ok: I suspect it might not be enough to save the project, but it should help a little, and can can be easy (ex: just put in if a couple layers of an old towel). Maybe if you insulate the top between uses (ex: throw a rag on it), it might be enough to keep it basically at a 200 degree steady state all the way through.
Those boards are a work of art 🎨
This sounds like a solution looking for a problem
Here's our flexy guy again that bends the understanding of electronics
I hope u keep working on this and eventually find the right PCB to make the kit. As always great to see your learning journey!
make a hilbert curve one so that you can scale it and you're not wasting all the surface area
absolute mad work mate. rigid flex is wacky expensive i cant imagine how you've spent on testing
I probably won't build this but I love watching these video
Bro all your videos are like magic to me.
Seems like something that could be corrected in software. PID loop perhaps? Use your sacrificed board to tune it to the actual temperature, then see if the same profile works well enough on another board with an offset temperature value
What if you were to bolt two polyimide steel-backed PCBs facing each other? That way, the mechanical force will keep the sandwich together, thus the adhesive delamination will not even matter.
good to see you back
If you are careful ramping the temperature very slowly and spread it properly you can go up to about 450C with silver sputtered onto quartz glass. It is however a monumental pain in the neck. Pyrex would do for your purposes but the sputtering will still be a pain.
You could try widening the traces toward the center of the board to decrease resistance and heat a touch?
Nicee work man 👌
my solution was throwing away the horrible T-962A and do the job with an Aluminiumplate (overmilled cast aluminum) with ceramic resistors clamped onto the back. I have this 2 times ... one for preheating and one for the soldering process. An Arduino micro does both temperature controls
Maybe encapsulated between the metal plates(aluminum/copper) with pads to ground it to the flexible pcb with a fuse set just above the expected current. You could even have a smd fuse on the flexible pcb. You could use thermal grease or adhesive to help prevent deformation between the metal plates and heater.
I know its disapointing, but its really good info, and you may have just saved many people costly mistakes.
Carl Bugeja, you are truely an inspiration. Could you make a video of the path it has taken you to reach your courrent knowledge level?
What are your thoughts about Applied Science’s recent video about making your own ceramic PCBs?
Siemens and Bosch BSH dishwashers use Flexible pcb heating element. maybe you can take a look. its a heatpump system.
Probably would help to split heater onto 2-4 separated circuits, and power simultaneously only 1 circuit at a time. Using timer switch power from one circuit to another, by this you won't overheat traces, trace will heat up then shut down and while another trace powered this one will dissipate heat to PCB, and so on. With right picked time delay it should last 10-20 times longer, it will degrade eventually anyway.
Why not use an enig finish and leave off the soldermask? Maybe use a copper fill ground plane so the surface looks like solid gold except for the area between the track and the ground plane. The manufacturer might not like it because they don't like seeing huge areas covered with gold, but this should stress the PCB a lot less than having the soldermask insulator does...
There is ceramic substrate too, which doesn't has flex and is resistant to more than 300C, when the silver/copper is protected against oxidation. But a simple aluminum block plus a silicone heating mat insulated on the back should be cheaper.
We attempted to make heat beds for 3DP from FR4, Polyamide, and other materials using graphene in 2017. They didn't survive.
Didn't survive under what test conditions and with what failure mode?
@@SianaGearz They burned the hell up. Separated just like this issue he is having. Material couldn't take the increased change in temperature. Graphene heats quickly with less power than it takes to heat a circuit board track resistive heater. We used to have videos on Facebook of the testing but Creality wiped all of the work along with all of the mods to the CR-10.
Hello, hope you'll succeed in this project and finish it. Good luck friend.
Well, as another comment suggests, it would still be a good kit for small scale reflowing, or like a limited amount of time. Making reflow accessible at a low price is probably good anyways.
Also, they make good beds for most 3d printers. Just look at prusa.
I will probably do it 😅but a more generic version so that is could be used for multiple things (not just reflow) probably I would rate it at lower temperatures and then specify that at higher temperatures in has a limited number of times it can be used
Maybe make a chamber and allow the heater to completely saturate the air inside or use a vacuum chamber? Not sure how the solder and paste would respond to being in a vacuum though.
Connect these two aluminum heaters with screws, give a spacer from the capton tape. Not important what will happen inside anyone there was no short-circuit. Operate on metal side.
random idea- using PCB coils, an energy harvesting circuit with a magnet attached to some sort of button
may or may not have been watching andreas spiess's new video :D
can you try to thermal isolate the botton of the plate so that you can get these few more dregrees on top rather then loosing them?
Weird, Ben on Applied Science just made a video about ceramic PCBs (ref the most recent video on "Underwater laser cutting")
This was a random UA-cam recommendation, found it interesting, just trying to understand what's happening here. What's the PCB, is it the hot plate? Which means you created a solder melting product by ordering a PCB in particular pattern?
Yes. Fundamentally, lower temperature PCB heaters are well known, almost every 3D printer uses one rated to less than 120°C on the built plate, used at around 60-80°C for most printing materials, so you get adhesion during the print process, but afterwards, some shrinkage occurs and surface properties change with temperature, and the print is easier to remove. But he's trying to see how far he can push it with material options that the companies offer and design changes, to try to make a reflow plate, maybe it's better explained in a previous vid. He made flex-PCB magnetic actuators and motors before.
Can you please attempt a pcb induction coil version of this. Kinda like a stove
Good stuff.
Have you tried using an ac waveform to power it?
silicon heat pad + 2-3mm aluminum plate and you're done... Even if you going to find a PCB that is stable for 30-60m at reflow temperatures the heating cycle would kill it rather quickly unless you want to heat up and cool it down over 2-4 hours to avoid cracking or delamination.
You've tested the board, not the board material. Especially, the mask material became darker of temperature, not the board composite. Tg - is a glass transition temperature, not the working temperature.
It would be better to use the stiffener as the heating surface. The bubbling up might come from thermal expansion. The stiffener doesn't expand while the circuit board does. I would paint a metal plate with graphite paint as a resistive heater and only use the pcb for the circuit. PEEK melts at a high temperature (343 °C / 649.4 °F) and is used for high temp applications.
How do you insulate the heater from the metal plate? Steel is conductive enough that it's too low impedance at those 12V.
@@SianaGearz you mean in the case of a carbon based ink? I would use a layer of high heat powder/ceramic coating. You can get an electrostatic powder coating sprayer for $40-80 and once it is baked in an oven 180-200C for 10 min it is hard. Some coatings need longer or slightly more heat (300-350C or 500-600F) many standard ovens reach those when you use the self cleaning cycle.
Why do you need the solder mask on the heat portion? You are only dealing with less than 10V and it's DC for that matter. Unless you can't request flex PCBs without solder mask.
You will only have millivolts (at best) across any two points on the board being reflowed.
Breaking Taps used Kapton tape that had been irradiated with a laser to form a layer of graphene which can act as a low voltage DC heater that can reach ~200 °C. No idea if it'd be suitable here.
Would bare copper on PCB do as well with an insulation of parchment paper? Another idea: some method of stitching the layers together so they're less likely to separate?
would still love to see a ceramic version of this anyway :D :D lol
You could still build a PCB that has IR LEDs on it and do IR reflow
The thermal gradient between the surface and the Cu can be improved if you add thermal isolation in the areas where you dont need heat output. fiberglass? teflon?
Hi, maybe you can mini indiction heater for reflow heater (mini indiction heater > thickness metal plate) ?
Great video. 😁
now i only have one question if it can handle pressures of 6 tons, then one could use 2 plates to make some nice rosin
Have you looked into laser induced graphine on kapton tape? Breaking taps did a video on it a couple months ago.
Nice video :), thanks
why dont you try using one of those stirring/heating plates heating elements? they are quite cheap...
I would think quartz would provide a better source of heat (probably the ceramic your talking about )
I'm glad you enjoyed a successful failure, learning what went wrong so you can improve the design in the future.
Have you tried adding the metal stiffener to both sides? I know some of the silicone heating pads need to be sandwiched to get the heat away from the pad or they will get damaged.
if make distance between coil and iron plate, and used induction for heating.
For ceramic PCB maybe you can ask the UA-camr applied science! He made some by itself maybe he will be happy to help you
Why not just strap the nichrome wire on aluminum for better heat spreading? That's how I built my 3d printer heated bed with a ds18b20 stuck for temp reading.
A 3 to 4 mm Aluminum or Copper disk cut to diameter should act as a heat sink and might improve performance.
Would 2 layers copper top and bottom of the metal core help spread the temperature evenly?
What if the copper traces were exposed/not covered with solder resist? To allow better heat transfer since it seems the solder mask is acting as an insulator between the hot traces and PCB to be reflowed
Why not just do a ceramic heater? as un ceramic - cooper/nicrome - glass? it will b bulkier but isnt that expensive. You could even cheap out on the ceramic and use steel reinforcement and hav a thinner ceramic layer.
Why dont you use a 4 layer pcb and put the copper layers in between?
if youre just reflowing you dont need to leave these on constantly right?
Aren’t there heating elements that are relatively cheap and flat? I think this is a case of “why make it a PCB if something else will do it better”. Though of course you’d have to have done all that testing to realise that. The “high temperature ceramic heating plates” are relatively cheap, same for the aluminium shrouded ones. I’d also be rather interested in using a TEC with a heat-sink on the cold side. You’d use less power than a normal heating element, and the cold side would allow it to cool more rapidly if you wanted. Measuring the temperature might be difficult though, you may want to put a shim of aluminium between the hot side and the workpiece to distribute the heat more evenly, with the thermocouple inside that.
Could via (or electrified mechanical hole) could hold the trace? If you do the exact same pattern on the underside of the PCB and put vias on the traces, maybe it could prevent the PCB from delamination?
kyon Bhai kyon ???
use iron and PID control method...
I liked the video before even watching xD
why not use silicone heaters like in your 3d printer heat bed?
it is going up and become Bimetal, is it better to put 4x ceramic heaters /230V/24V and aluminium plate ??
How about using just a routed aluminum?
Why not follow the standard SMT Reflow Soldering Temperature Curve.
There wont be a need to set the temp to 200 *c all the time.
Yeah you could do that but there will be a limited number of reflow you could do - maybe reflow 3pcbs and then through it alway - like a " disposable hotplate " the price needs to be cheap though - I'm going to give this a shot with the HeatStick project I showed at the end
Have you tried taping aluminium tape to it?
I think you can recreate the sensor shifting thing used in iPhone. So can you do that
can we reverse this?
can it produce electricity from sun heat by focusing sun light with magnifying glass
Did you tried heating water with it? I have been thinking about a DIY stir plate with heat, as they cost crazy high to buy.
Dude, bulging and wraping and delamination is bad, but discoloration is fine. Think about it, it is the color pigments that degrade and nothing structural. I guess if you go with black which uses carbon as pigment then you should be good.
which program you are using, and what function it is that allow you generate the trace on the PCB? I would imagine manually trace them all would be a pain.
I wouldn't bother with the stiffener. Just pop on a silicon mat.
Don't give up, I'm sure you could have s cheap controlled that cuts out after a given time; noone is reflowing for hours.