looks like you got it ...but maby research how lab stuff operates the plungers on the needls ...you might be able to get the drops even smaller for smaller parts
I’ve also been working on integrating paste dispensing into openpnp - I headed down the path of creating a duplicate set of footprint definitions in openpnp for anything that needed to be pasted, and adding a ‘_P’ to the end of the footprint name. From there, you can create a macro that runs before a part is picked, and if it sees a ‘_P’ at the end of the footprint then it can apply paste. You can do this by either adding a pad in the openpnp footprint definition for every solderblob that you want, and can then set the position relative to the footprint as well as the size of the blob ( by say defining a circle with the circle diameter being the blob size ). You could also have a macro or external gcode files if you wanted to paste a line instead of a blob for example. All of this can be done without needing to change openpnp itself
This is huge for me, I work in a university fab-lab in the UK and we're trying to teach more of the SMD process to our students, but 90% of our boards are one-offs, extremely custom and used for one project, so it is almost never worth it to get a stencil. We're likely going to be getting a lumen in the next year or so, so I very much look forward to using this solution!
I'm cs lecturer and we had the same need. We got a voltera vone. Not cheap, and rubbish at creating pcbs. But the solder paste dispensing and built-in hotplate to reflow works great. Having it as part of the lumen would be very cool.
Just be aware that paste jetters can be... Fiddly. And not suitable for fine pitch or large BGAs. You'll be pulling your hair out over coplanarity issues.
Great to see paste dispensing coming on the Lumen! Here are a few tips I got setting up solder paste jetting on an Essemtec PnP: - stagger the paste dots on QFN, QFP, etc. This will lower the amount of solder bridge (short) on small pitch - paste temperature will impact its viscosity. Higher temperature = less viscous = more control over dot size & less oozing - find a way to clean automatically and do that regularly (every couple 100s dots), as paste on the side of the nozzle will divert the solder to the side - repeatability on dot volume and position is key to a reliable soldering (tombstone, bridging). This can be impacted by flex and friction of your dispenser, so use trust bearings and beefy assembly Although it's not the exact same process, this should also apply with your type of dispensing.
After years of lurking around this channel it's awesome to still see you being cheerful about the freaking amazing things you build! Keep up the great work!
DING! Well done - That's the shizzzz!!! Just keep in mind that ambient room temp can swing as much as 25deg C (cold / hot day) without thermal control of your production environment, so the viscosity can change quite a bit from day to day, or even during the same day. Maybe a temp sensor near the tube to determine the current amb temp to allow for auto adjustments of the dot sizes? Anyway, awesome work mate!
A few more years and you will just send files and lumen will take a plain board, create traces, apply solder place, place component , solder and test all by itself. :D This is super cool.
You and Opulo are absolutely amazing! From a Ret EE, >40 yrs designs and manufacturing experience, set up and ran multiple SMT lines in US and Asia. Congratulations! Now if you could just be a little more positive in your delivery... 😃
Absolutely love all the work you are putting into this system, and enjoying following along with you as you do it. Your videos are amazing. Very inspiring for my work too (Power Electronics and FW Engineer). Actively working to convince the boss to get a lumen so I can play with it!
Could you please make a video on how you design the lumenpnp with Freecad? I know that this isn't the typical type of video you make but that would be really useful because the lumenpnp is a kind of complex assembly and there aren't really a lot of popular examples of those developed in Freecad.
yeah that's something Lucian and I have been thinking about doing, maybe on a livestream we edit down into a tutorial? Lucian is WAYYYY better at FreeCAD than I am
@@stephen_hawesout of curiosity are you still using the latest stable or have you tried the recent 1.0 release candidates? I haven't had a need to test the RCs myself but allegedly it's a pretty major overhaul.
@@felixjohnson3874 the RC versions are an INCREDIBLE improvement. i used v1.0RC2 exclusively for the paste extruder, and holy smokes, night and day difference. I think the official 1.0 release is slated for the next couple weeks, i can't recommend upgrading enough once that happens! the FreeCAD team has been putting in some serious work to make these improvements.
@@stephen_hawes +1 For this. Especially when 1.0 comes out, there's so much potential that I feel I'm leaving on the table having learned on Stable. Especially since assemblies are add-ons.
I highly recommend @MangoJellySolutions' videos about using FreeCAD. They're very accessible and provide a good basis. Incidentally, his video about FreeCAD v1.0 features the LumenPnP on the thumbnail. www.youtube.com/@MangoJellySolutions
I loved this! I'm reminded a bit of some early MakerBot experiments when they were trying to frost cupcakes. The first version was something like this, but I think the second version used pneumatics. I wonder if you could just hook up the vacuum hose and push a little paste out instead of driving it with a motor. IIRC, the Lumen already pushes air to get parts off the nozzle.
2 месяца назад+4
There's a great manufacturer in Germany for conductive inks and conductive glues, it's name is Saralon (Saral Inks).
I've worked in PCB manufacturing...the thing is the solder liquidity is a big factor in dispensing ( flux separation in the syringe)...you will get nice flow then harden as you use up the solder which will course bad joints.... You can obtain a stencil from the PCB manufacturer easily.. The dispensing is too inconsistent...and you will be adjusting as you move down the syringe... stencil printing is two swipes..job done 👍
Do you think stencil application could be practically automated with a device that's in the same genre of construction as the Lumen? I assume a 2.5d motion system wouldn't be a particularly applicable motion system for it, but it'd be great if it could be built with the same motors, members, chips. In an automated scenario how would you dispense the solder onto the stencil? Does that step make syringes work out better?
But you still need to get the stencil… even if it’s cheap, it’s still an extra step. If you don’t have a stencil, you don’t have shipping delays or storage concerns if you order ahead of time. You don’t have to deal with the wrong stencils being sent, or them getting damaged either in transit or just during use.
Stencil is more than two swipes. I'm hoping this can be dialed in so I never have to stencil again, and so I can build robots to fully automate a small run.
@@23lkjdfjsdlfj No... !.. it's not more than 2 swipes??? i use stencils every day.....if so the solder is out of date and gone hard??? or not mixed correctly..!! you can obtain the stencil from the PCB manufacturer at the same time as they make the blank PCB's...
Amazing work Stephen, what a game changer. I would be really interested to see how small you can go. I hear in the comments talks about 0402, but if you could go smaller then this setup would eliminate the need for stencil. And given that I live in New Zealand my lead time is 10 to 14 days to get this back form overseas. I could, as you say, make something in 24. A life changer OMG
I cannot get the flashbacks to seeing people slather the board and components in solder paste and then seeing the small solder balls just float all over the board in the sea of flux _ææææ_ micro shorts all over the board
You can get rid of the air in the syringe by taping in the plunger and taping a strong string or wire to the side. It makes improv centrifuge. Nozzle pointing to your hand, plunger away from hand. Couple or decent spins (in a safe area) the solder paste will compact against the plunger. Remove taping, then squeeze the air out the nozzle.
This might help or make things worse with filling syringes with solder. To load syringes with thixotropic epoxy, I had luck filling a larger syringe with a long dispenser tip, then using it to fill my detail work syringe. But if I were to do it all over again, there might be a better way. Load just one side of the syringe so that air can flow out, put the plunger in slightly, add a check-valve to the tip and toss the whole thing in a vacuum. All the air will leave under vacuum, but the plunger won't move. When air is let back into the chamber, none can get in the syringe, so the plunger will move and pack everything nice and bubble free.
If you want to fill the syringe without getting air bubbles in there, you can fill directly from the 600g Semco cartridges through the front of the syringe with the syringe fully depressed. That way you avoid introducing air bubbles when you fill it. I think you'd need 1/4" NPT Male to Male Luer Lock (which would fit a standard needle and is the fitting on the front of the syringe you are currently using) and a (female to female) Luer Lock coupler. That way you don't need a huge 600g cartridge on the PnP.
I have been following you for years! This is great! You are going to turn pcb development into the next craze, like when 3D printing became mainstream! I have an idea to add, too. What about adding a CNC head (small) on an Idex carriage. The pick and place and paste despenser could sit tucked away on the side away from the dust as the cnc head mills the copper clad into a circuit. Dust collection (small) to keep it clean. Then with your reflow plate and this, you could go from a blank copper board to a working pcb. With a few macros, it could take the file dropped in and automate the whole process. 🤯🤯🤯
Excellent Work Stephen! To add a 3rd level of functionality to the machine, you could automate the PCB stepping down onto a hotplate when everything is pasted and placed so you can have everything soldered too! I'm on your wavelength in terms of integrating all-in-one unified systems. It's the best way! Personally I've spent a lot of time building a 3/4 Axis all metal CNC machine with integrated lasers, a 3D printing mode and I might also add a basic plasma head. I'm not good in front of a camera and 99% of my stuff/inventions are private atm. I am sure I will be ordering a Lumen some time in the near future 😉. Cheers. ~ Apple's can fall far from the tree.
Here's a rabbit hole I found while researching this: a diy "planetary centrifuge" can evenly mix your jars (and syringe) of solder paste and remove (or evenly distribute) air bubbles. (Just ...be sure to build it on the inside of a stockpot or something.) One harder: insert the plunger before filling, and fill through the luer via a larger syringe. Even harder than that: Introduce a little vacuum first? Also: Use the head camera to snap images of the pcb pads, dispense your paste, and then take another image. Compare the 2 using opencv to verify that paste is on the pad - as a sort-of AOI.
Congrats on figuring all this out. I'm impressed! I don't know if it's worth it, or if it would work well enough, but I wonder if you could adapt a "food saver" type kitchen vacuum sealing device or a hand pump to get all of the air out of the syringe. It is possible that your purging solution may be more practical.
Love your videos! I would suggest supporting the motor rear shaft if you're gonna hammer something onto the front shaft, those bearings are really not rated for that load!
usually the commercial systems are using air pressure. the advantage is that you can apply a negative pressure to "suck" the solder back when you are done with one spot.
But you can do the same with his mechanical system? I mean just rotate the gear in the other direction, syringe goes up and creates suction to pull back?
GC10 my go to paste, remember there are a couple of variants regarding ball sizes. GC10 isn’t bad viscosity versus heat from my experience, never getting too runny like some of the cheaper stuff.
Your videos make me feel way too good! Exciting progress! I wonder if there are some trivial tweaks to speed up the process. Like flexible dual nozzle tips or a nozzle tip rolling wheel with holes inside xD. Still hoping to have the need to get a LumenPNP at some point haha
Air buble act as spring in your Cartridge but , the Cartridge itself might play the same role : it still a thin plastic shell that deform under pressure then turn back into itself while pushing the paste.
Your work is like 3d printer at its early stages. To me you are basically doing the same job prusa did with it : making it way more realiable, cheap and easy to use with lot's of new features that make life easier and are soon to be the norm. Maker can make new things without having to be an expert in all of those domains !!
This is awesome. I hope this can be developed into a set and forget solution, like 3d printers have evolved from finicky to tool. I'd love to see a video of you showing off what the open source community has done with the beta test kits you're sending out.
As regards the small amount of air trapped acting as a 'virtual spring' I just added a minor 'reverse' action in my software that releases the pressure and stops the solder paste continuing to 'dribble' out. Once fine tuned, the syringe delivers correct sized droplets of paste with no unwanted 'dribbles' ....cool
I don’t think this would be easy to do but selective soldering would be a great idea if you could potentially get head from a machine although require modifications to the pick up place. It may be worth experimenting with this because making a selective old room machine could be very profitable for small scale.
You can use the camera to calibrate the amount that needs to be dispensed. You need to estimate the mass volume with a 2D camera which is the hard part.
I made a device that goes in a drill to de-gass syringes. It spins the syringe round, and all the air goes to the top and you can squeeze it out. You need to support the plunger of the syringe and not the syringe body itself though, or the weight of the material in the syringe pushes the piston out and it makes a right mess. Ive not done it with solder paste, ive only done it with jointing compound, but same thing should work. I did two at a time in a pillar drill, takes about 10 seconds and all the air is at the top.
To get the air out, you could possibly try centrifuging the syringe after loading it with paste. I'd orient the syringe with the plunger down so you end up with all the air at the nozzle end where it can easily escape when you prime it. (If you spin it with the nozzle down, you'll get an air bubble in front of the plunger.) One possible concern, however, is that centrifuging might also cause the solids to settle out of the liquid component. I have no idea how well the metal particles resist coming out of suspension. Perhaps it's not an issue at all, or if it is, you still might be able to find a sweet spot for the RPM that's high enough to separate the air, but low enough that it doesn't separate the solids.
How did I miss this? I’m game to get one of those Lumen things since I have a bunch of PCBs to make for my old military truck. I want to bring it out of the Stone Age with some modern tech for it.
Vibration. In the construction the when the cement is poured they use vibrationing tools to remove the bubbles. Something like ultrasonic toothbrush might help with the air in the syringe.
maybe to help with the air issue. a 'poor man's vacuum chamber' for de-gasing resin/silicone is to fill a syringe. Then stuff the tip into silicone/rubber. Then pulling back on the syringe hard and hold it for a moment, putting the inside of the syringe under a weak vacuum. This should collect all the air into one bubble you can push out.
See if it helps putting the paste dispenser in a vacuum chamber to drive the air out, like people do for silicon molding. Might not be the final solution, but might provide insights. Another solution might be trying some vibrator rig. Put the syringe with the opening up in a vibrator right, maybe together with light preheating, to make the bubbles wander up through the paste
+1 on this. Smart idea. With a worm drive, the syringe won't be able to "push back" against motor. Probably get a little more accurate/consistent dispensing.
I already love you for using FreeCAD. Thank you! Why do you have to have the pushing/extruding mechanism in the toolhead? With 3D filament there is the issue of the filament to eat up the tub from too much bending/friction but with solder paste I think you'll not have that limitation. Think of it like using a fire hydrant and a hose that drives the water. You'll also be able to design the pushing and pressure separate of the whole toolhead and that will remove some limitations. Maybe you'll be able to fill it with solder paste easier, much better pressure control, no weight limitation.
@@stephen_hawes that would be an amazing setup for one-offs and short runs... LumenPNP is such a cool project, great work on keeping it moving over the years, I know how hard that can be! ❤
There is a machine i saw online that pressurizes the plunger with air dispensing paste with a foot pedal, this mechanism is way more complicated but i would think it’s more accurate
I make way too few board to need something like this but I do prefer SMD over through hole for the easy of assembly. That means that I do all assembly by hand using a manual pick&place using one of those manual "scara" arm style units and that is very easy and fairly quick. But I do have (a lot) of problems applying solder paste, even with my pneumatic paste dispenser. It seems like the paste is sticking better to the luer lock tip than to my PCB (and yes, it's clean). I guess I at least need to get hold of some GC-10 paste and give that a try
You will want to check out pastes made for jetting, they work much better for this than stencil pastes. Harima/Loctite makes GC variants for jetting too.
Very cool stuff, I will keep an eye on this! For paste I've been using ChipQuik which seems to be easier to source. Where do you source the Loctite from? I'd love to give it a try.
There was a change-over in distribution a few years back that made sourcing difficult, but now that the dust has settled you can get it from the usual sources- DigiKey, TestEquity, etc.
Please try with a smaller nozzle to see how small of a footprint this can apply paste to? This would be very useful for small footprints that are near on impossible to put paste on by hand without a stencil. It would be so cool to see it with a heater integrated into the fixture so it can reflow in place!
4:00 Why don't you retract and have the nozzle suck up the paste rather than filling it yourself? any air left at the top will be the same no matter the refill as long as you use the same syringe and nozzle, makes it easier to compensate for the spring factor
If you can figure out a way to make it carve PCBs reliably without horrible etchants you are getting very close to complete home electronics production. Then of course you would have to take another step back and find a way to make your own PCB blanks from copper stock and force of will.
@stephen_hawes The year is 2037. The Lumen v18 is able to go and mine copper fully autonomously. The Lumen v19 release candidate changelog casually mentions adding full sentience as a feature.
I think you'll need an even higher gear ratio, especially to handle fine-pitch ICs rather than passive parts to overcome the back pressure from a finer-tipped lure lock.
SOLUTION: auger blades on the plunger. They bite into the solder paste so when you "retract" solder you have excellent "grip" on the paste and have very fine control no matter the viscosity (within reason). This solves the temperature / viscosity / reliability / consistency problem. Edit: the extra control you get will also help meet your fine pitch requirements.
If it were possible, it would be good to use a worm gear drive. It would take way less power and have better control. I don't know how well it would 3d print or if that would be possible because of the rubbing wear.
Ok, that's huge. Now next logical step is DIY heater that is connected with Lumen by motorized conveir belt. You place PCB, solder paste, pick&place, then move it to your oven and you got fully assembled PCB.
check out the new extruder!: www.opulo.io/products/paste-extruder
looks like you got it ...but maby research how lab stuff operates the plungers on the needls ...you might be able to get the drops even smaller for smaller parts
I’ve also been working on integrating paste dispensing into openpnp - I headed down the path of creating a duplicate set of footprint definitions in openpnp for anything that needed to be pasted, and adding a ‘_P’ to the end of the footprint name. From there, you can create a macro that runs before a part is picked, and if it sees a ‘_P’ at the end of the footprint then it can apply paste. You can do this by either adding a pad in the openpnp footprint definition for every solderblob that you want, and can then set the position relative to the footprint as well as the size of the blob ( by say defining a circle with the circle diameter being the blob size ). You could also have a macro or external gcode files if you wanted to paste a line instead of a blob for example.
All of this can be done without needing to change openpnp itself
This is huge for me, I work in a university fab-lab in the UK and we're trying to teach more of the SMD process to our students, but 90% of our boards are one-offs, extremely custom and used for one project, so it is almost never worth it to get a stencil. We're likely going to be getting a lumen in the next year or so, so I very much look forward to using this solution!
I'm cs lecturer and we had the same need. We got a voltera vone. Not cheap, and rubbish at creating pcbs. But the solder paste dispensing and built-in hotplate to reflow works great. Having it as part of the lumen would be very cool.
Just be aware that paste jetters can be... Fiddly. And not suitable for fine pitch or large BGAs. You'll be pulling your hair out over coplanarity issues.
@@kencameroncreatesI'm glad I'm not the only one that thought the voltera was total shit!
Great to see paste dispensing coming on the Lumen!
Here are a few tips I got setting up solder paste jetting on an Essemtec PnP:
- stagger the paste dots on QFN, QFP, etc. This will lower the amount of solder bridge (short) on small pitch
- paste temperature will impact its viscosity. Higher temperature = less viscous = more control over dot size & less oozing
- find a way to clean automatically and do that regularly (every couple 100s dots), as paste on the side of the nozzle will divert the solder to the side
- repeatability on dot volume and position is key to a reliable soldering (tombstone, bridging). This can be impacted by flex and friction of your dispenser, so use trust bearings and beefy assembly
Although it's not the exact same process, this should also apply with your type of dispensing.
awesome, thanks for all the tips!! really cool idea with staggering the dots, makes a ton of sense, ill definitely try that out!
That transition from the CAD model to the actual print was SICK!
After years of lurking around this channel it's awesome to still see you being cheerful about the freaking amazing things you build!
Keep up the great work!
My life goal is to be as excited about anything as Stephen is about everything.
DING! Well done - That's the shizzzz!!! Just keep in mind that ambient room temp can swing as much as 25deg C (cold / hot day) without thermal control of your production environment, so the viscosity can change quite a bit from day to day, or even during the same day. Maybe a temp sensor near the tube to determine the current amb temp to allow for auto adjustments of the dot sizes? Anyway, awesome work mate!
The speed of your progress of implementing the "small addition" of an entire usually manual and messy extra work step is just incredible!
These fun little projects are why I love this channel
A few more years and you will just send files and lumen will take a plain board, create traces, apply solder place, place component , solder and test all by itself. :D This is super cool.
You would probably have to put on the UV-sensitive forbidden fruit roll-up first, but that would be awesome.
You and Opulo are absolutely amazing! From a Ret EE, >40 yrs designs and manufacturing experience, set up and ran multiple SMT lines in US and Asia.
Congratulations! Now if you could just be a little more positive in your delivery... 😃
this is great, the pasting process is far and away the messiest and least deterministic part of this process now for me. love my lumen machine
Absolutely love all the work you are putting into this system, and enjoying following along with you as you do it. Your videos are amazing. Very inspiring for my work too (Power Electronics and FW Engineer).
Actively working to convince the boss to get a lumen so I can play with it!
Could you please make a video on how you design the lumenpnp with Freecad? I know that this isn't the typical type of video you make but that would be really useful because the lumenpnp is a kind of complex assembly and there aren't really a lot of popular examples of those developed in Freecad.
yeah that's something Lucian and I have been thinking about doing, maybe on a livestream we edit down into a tutorial? Lucian is WAYYYY better at FreeCAD than I am
@@stephen_hawesout of curiosity are you still using the latest stable or have you tried the recent 1.0 release candidates? I haven't had a need to test the RCs myself but allegedly it's a pretty major overhaul.
@@felixjohnson3874 the RC versions are an INCREDIBLE improvement. i used v1.0RC2 exclusively for the paste extruder, and holy smokes, night and day difference. I think the official 1.0 release is slated for the next couple weeks, i can't recommend upgrading enough once that happens! the FreeCAD team has been putting in some serious work to make these improvements.
@@stephen_hawes +1 For this. Especially when 1.0 comes out, there's so much potential that I feel I'm leaving on the table having learned on Stable. Especially since assemblies are add-ons.
I highly recommend @MangoJellySolutions' videos about using FreeCAD. They're very accessible and provide a good basis. Incidentally, his video about FreeCAD v1.0 features the LumenPnP on the thumbnail. www.youtube.com/@MangoJellySolutions
I loved this! I'm reminded a bit of some early MakerBot experiments when they were trying to frost cupcakes. The first version was something like this, but I think the second version used pneumatics. I wonder if you could just hook up the vacuum hose and push a little paste out instead of driving it with a motor. IIRC, the Lumen already pushes air to get parts off the nozzle.
There's a great manufacturer in Germany for conductive inks and conductive glues, it's name is Saralon (Saral Inks).
Cool! As for the paste oozing from the nozzle - you could think about some “nozzle wiper” as in some 3d printers now :)
OOOH that's a great idea, I should definitely do that! it was a pain sitting there with a paper towel waiting for it to stop
I've worked in PCB manufacturing...the thing is the solder liquidity is a big factor in dispensing ( flux separation in the syringe)...you will get nice flow then harden as you use up the solder which will course bad joints....
You can obtain a stencil from the PCB manufacturer easily..
The dispensing is too inconsistent...and you will be adjusting as you move down the syringe... stencil printing is two swipes..job done 👍
Do you think stencil application could be practically automated with a device that's in the same genre of construction as the Lumen? I assume a 2.5d motion system wouldn't be a particularly applicable motion system for it, but it'd be great if it could be built with the same motors, members, chips.
In an automated scenario how would you dispense the solder onto the stencil? Does that step make syringes work out better?
But you still need to get the stencil… even if it’s cheap, it’s still an extra step.
If you don’t have a stencil, you don’t have shipping delays or storage concerns if you order ahead of time. You don’t have to deal with the wrong stencils being sent, or them getting damaged either in transit or just during use.
Stencil is more than two swipes. I'm hoping this can be dialed in so I never have to stencil again, and so I can build robots to fully automate a small run.
@@23lkjdfjsdlfj No... !.. it's not more than 2 swipes??? i use stencils every day.....if so the solder is out of date and gone hard??? or not mixed correctly..!! you can obtain the stencil from the PCB manufacturer at the same time as they make the blank PCB's...
Paste for dispensing is different from the stuff in tubs for printing - I think it's less viscous - not sure if they make dispensing GC10
Amazing work Stephen, what a game changer. I would be really interested to see how small you can go. I hear in the comments talks about 0402, but if you could go smaller then this setup would eliminate the need for stencil. And given that I live in New Zealand my lead time is 10 to 14 days to get this back form overseas. I could, as you say, make something in 24. A life changer OMG
I can't wait to see the next version that's solder tube compatible. Would allow you to use those really fine tip needles for paste.
I cannot get the flashbacks to seeing people slather the board and components in solder paste and then seeing the small solder balls just float all over the board in the sea of flux _ææææ_ micro shorts all over the board
GC-10 is so good that I spread that shit on my toast for breakfast every day. nom nom nom
it is actually so delicious
Forbidden frosting 🍰
Looks like GC 10 is no longer available at DigiKey. Anyone have a good replacement for it?
@@HarjitSi try going to Henkel adhesives. It's the loctite mother company and who I go through. I did a quick search there and didn't see it either.
I don't know you, but it made me very happy to see you so excited and happy:)
You can get rid of the air in the syringe by taping in the plunger and taping a strong string or wire to the side. It makes improv centrifuge. Nozzle pointing to your hand, plunger away from hand. Couple or decent spins (in a safe area) the solder paste will compact against the plunger. Remove taping, then squeeze the air out the nozzle.
Hey Stephen, your videos are a great inspiration. You're doing a great job, very impressive. Keep it up.
This might help or make things worse with filling syringes with solder. To load syringes with thixotropic epoxy, I had luck filling a larger syringe with a long dispenser tip, then using it to fill my detail work syringe.
But if I were to do it all over again, there might be a better way. Load just one side of the syringe so that air can flow out, put the plunger in slightly, add a check-valve to the tip and toss the whole thing in a vacuum. All the air will leave under vacuum, but the plunger won't move. When air is let back into the chamber, none can get in the syringe, so the plunger will move and pack everything nice and bubble free.
That’s actually genius level thinking… I think he should try it.
If you want to fill the syringe without getting air bubbles in there, you can fill directly from the 600g Semco cartridges through the front of the syringe with the syringe fully depressed. That way you avoid introducing air bubbles when you fill it. I think you'd need 1/4" NPT Male to Male Luer Lock (which would fit a standard needle and is the fitting on the front of the syringe you are currently using) and a (female to female) Luer Lock coupler. That way you don't need a huge 600g cartridge on the PnP.
this project is a blessing to watch
Totally crazy how you showed off a solder paste printer around the same time Bad Obsession Motorsports unveils a hot solder 3D printer.
I have been following you for years! This is great! You are going to turn pcb development into the next craze, like when 3D printing became mainstream! I have an idea to add, too. What about adding a CNC head (small) on an Idex carriage. The pick and place and paste despenser could sit tucked away on the side away from the dust as the cnc head mills the copper clad into a circuit. Dust collection (small) to keep it clean. Then with your reflow plate and this, you could go from a blank copper board to a working pcb. With a few macros, it could take the file dropped in and automate the whole process. 🤯🤯🤯
This is soo cool! I can absolutely feel your excitement!!
Excellent Work Stephen! To add a 3rd level of functionality to the machine, you could automate the PCB stepping down onto a hotplate when everything is pasted and placed so you can have everything soldered too!
I'm on your wavelength in terms of integrating all-in-one unified systems. It's the best way! Personally I've spent a lot of time building a 3/4 Axis all metal CNC machine with integrated lasers, a 3D printing mode and I might also add a basic plasma head.
I'm not good in front of a camera and 99% of my stuff/inventions are private atm.
I am sure I will be ordering a Lumen some time in the near future 😉. Cheers.
~ Apple's can fall far from the tree.
Exciting!!! Especially if you all can get it working with 0402 eventually 🙂. Really hoping I can order a Lumen P&P next month 🍀
Here's a rabbit hole I found while researching this: a diy "planetary centrifuge" can evenly mix your jars (and syringe) of solder paste and remove (or evenly distribute) air bubbles. (Just ...be sure to build it on the inside of a stockpot or something.)
One harder: insert the plunger before filling, and fill through the luer via a larger syringe.
Even harder than that: Introduce a little vacuum first?
Also: Use the head camera to snap images of the pcb pads, dispense your paste, and then take another image. Compare the 2 using opencv to verify that paste is on the pad - as a sort-of AOI.
Congrats on figuring all this out. I'm impressed! I don't know if it's worth it, or if it would work well enough, but I wonder if you could adapt a "food saver" type kitchen vacuum sealing device or a hand pump to get all of the air out of the syringe. It is possible that your purging solution may be more practical.
Love your videos! I would suggest supporting the motor rear shaft if you're gonna hammer something onto the front shaft, those bearings are really not rated for that load!
That's SOMETHING, Stephen!
usually the commercial systems are using air pressure. the advantage is that you can apply a negative pressure to "suck" the solder back when you are done with one spot.
But you can do the same with his mechanical system? I mean just rotate the gear in the other direction, syringe goes up and creates suction to pull back?
@@vinzz2008 I don't think it's precise enough
Very nicely done. I'm not going to discuss what the best paste is, but this very nice add-on
GC10 my go to paste, remember there are a couple of variants regarding ball sizes.
GC10 isn’t bad viscosity versus heat from my experience, never getting too runny like some of the cheaper stuff.
what I can see this being useful for is component glue for double sided boards!
Your videos make me feel way too good!
Exciting progress! I wonder if there are some trivial tweaks to speed up the process. Like flexible dual nozzle tips or a nozzle tip rolling wheel with holes inside xD.
Still hoping to have the need to get a LumenPNP at some point haha
This is the future, man! I'm very excited. If this can do ICs in any capacity, I will shit myself.
Air buble act as spring in your Cartridge but , the Cartridge itself might play the same role : it still a thin plastic shell that deform under pressure then turn back into itself while pushing the paste.
Your work is like 3d printer at its early stages.
To me you are basically doing the same job prusa did with it : making it way more realiable, cheap and easy to use with lot's of new features that make life easier and are soon to be the norm. Maker can make new things without having to be an expert in all of those domains !!
love the idea of complete on your own timeline.
First hershies kisses was exactly what I expected :D awesome work!
🙌🙌🙌 finally
Sleeve on the motor cable harness was the best part 👌
This is awesome. I hope this can be developed into a set and forget solution, like 3d printers have evolved from finicky to tool. I'd love to see a video of you showing off what the open source community has done with the beta test kits you're sending out.
Hot plate or even hot-air from below for preheat, and a small laser for reflow. Saves having to bring the whole board up to transition temp.
As regards the small amount of air trapped acting as a 'virtual spring' I just added a minor 'reverse' action in my software that releases the pressure and stops the solder paste continuing to 'dribble' out. Once fine tuned, the syringe delivers correct sized droplets of paste with no unwanted 'dribbles' ....cool
I don’t think this would be easy to do but selective soldering would be a great idea if you could potentially get head from a machine although require modifications to the pick up place. It may be worth experimenting with this because making a selective old room machine could be very profitable for small scale.
You can use the camera to calibrate the amount that needs to be dispensed. You need to estimate the mass volume with a 2D camera which is the hard part.
Imagine having a hot air gun with an infra red camera to close the temperature control loop and reflow the board as well automatically
I made a device that goes in a drill to de-gass syringes. It spins the syringe round, and all the air goes to the top and you can squeeze it out. You need to support the plunger of the syringe and not the syringe body itself though, or the weight of the material in the syringe pushes the piston out and it makes a right mess. Ive not done it with solder paste, ive only done it with jointing compound, but same thing should work. I did two at a time in a pillar drill, takes about 10 seconds and all the air is at the top.
Nice job! Love the fact that you are using FreeCAD!
To get the air out, you could possibly try centrifuging the syringe after loading it with paste. I'd orient the syringe with the plunger down so you end up with all the air at the nozzle end where it can easily escape when you prime it. (If you spin it with the nozzle down, you'll get an air bubble in front of the plunger.)
One possible concern, however, is that centrifuging might also cause the solids to settle out of the liquid component. I have no idea how well the metal particles resist coming out of suspension. Perhaps it's not an issue at all, or if it is, you still might be able to find a sweet spot for the RPM that's high enough to separate the air, but low enough that it doesn't separate the solids.
How did I miss this? I’m game to get one of those Lumen things since I have a bunch of PCBs to make for my old military truck. I want to bring it out of the Stone Age with some modern tech for it.
i follow you from long time and a dispencer to make reballing easy would be the best thing ever
you can use a $25 desktop vinyl cutter to make adhesive stencils which are good to about 0603, they can be reused a few times with care
The idea is just brilliant, and I am really glad you made it work!
Did you consider adding retraction in the style of 3d-printers to minimise oozing?
a little holder with ultrasonic would be awesome to load the syringe without air gaps
Vibration. In the construction the when the cement is poured they use vibrationing tools to remove the bubbles. Something like ultrasonic toothbrush might help with the air in the syringe.
oh that's a killer idea, thanks for the rec!! ill definitely try that
maybe to help with the air issue. a 'poor man's vacuum chamber' for de-gasing resin/silicone is to fill a syringe. Then stuff the tip into silicone/rubber. Then pulling back on the syringe hard and hold it for a moment, putting the inside of the syringe under a weak vacuum. This should collect all the air into one bubble you can push out.
Pre-packed paste syringes will definitely improve things.
i really hope this will end up working on 0402 as well
been waiting for this one
See if it helps putting the paste dispenser in a vacuum chamber to drive the air out, like people do for silicon molding.
Might not be the final solution, but might provide insights.
Another solution might be trying some vibrator rig.
Put the syringe with the opening up in a vibrator right, maybe together with light preheating, to make the bubbles wander up through the paste
Id reccomend smaller motor with worm drive mechanism
+1 on this.
Smart idea. With a worm drive, the syringe won't be able to "push back" against motor. Probably get a little more accurate/consistent dispensing.
Kudos to you, not only for the design, but for being able to use FreeCAD!
That's really huge.
Make your self a little centrifuge for your syringes. A 3d print and a stepper motor done!
I already love you for using FreeCAD. Thank you!
Why do you have to have the pushing/extruding mechanism in the toolhead? With 3D filament there is the issue of the filament to eat up the tub from too much bending/friction but with solder paste I think you'll not have that limitation. Think of it like using a fire hydrant and a hose that drives the water. You'll also be able to design the pushing and pressure separate of the whole toolhead and that will remove some limitations. Maybe you'll be able to fill it with solder paste easier, much better pressure control, no weight limitation.
This is so dope man! Good work!
Really awesome Stephen! What about integrating a heat plate under the PCB mount so it reflows on the machine after placement? 🤔
thanks! that would be SOOO cool, all one step! i have a little hot plate i could try it with!
@@stephen_hawes that would be an amazing setup for one-offs and short runs... LumenPNP is such a cool project, great work on keeping it moving over the years, I know how hard that can be! ❤
Laser head so it can solder as well?
wouldn't the laser reflect in the solder after it cooled down just enough and possibly damage some other parts coating?
This is the upgrade I need!
On the old ink jets they had a little wast bucket with a flexable rubber flipper to clean/prep the head
Just a suggestion: you can add a laser head and see if it can melt the paste.
There is a machine i saw online that pressurizes the plunger with air dispensing paste with a foot pedal, this mechanism is way more complicated but i would think it’s more accurate
I make way too few board to need something like this but I do prefer SMD over through hole for the easy of assembly. That means that I do all assembly by hand using a manual pick&place using one of those manual "scara" arm style units and that is very easy and fairly quick. But I do have (a lot) of problems applying solder paste, even with my pneumatic paste dispenser. It seems like the paste is sticking better to the luer lock tip than to my PCB (and yes, it's clean). I guess I at least need to get hold of some GC-10 paste and give that a try
I will buy it if you make solder ball dispenser that would have a programmable template.
this is going to sound crazy but i bet you could replace the solder paste with icing and decorate a cake... yum
You will want to check out pastes made for jetting, they work much better for this than stencil pastes.
Harima/Loctite makes GC variants for jetting too.
f.e. www.harima.co.jp/en/products/electronics/multicore/solder_pastes/gc_50.html
Very cool stuff, I will keep an eye on this! For paste I've been using ChipQuik which seems to be easier to source. Where do you source the Loctite from? I'd love to give it a try.
There was a change-over in distribution a few years back that made sourcing difficult, but now that the dust has settled you can get it from the usual sources- DigiKey, TestEquity, etc.
@@StargirlFlowers what is the new part number? digikey says 4590-2064239-ND is obsolete.
Please try with a smaller nozzle to see how small of a footprint this can apply paste to? This would be very useful for small footprints that are near on impossible to put paste on by hand without a stencil.
It would be so cool to see it with a heater integrated into the fixture so it can reflow in place!
Maybe try with a servo drive instead of a stepper motor...
Can this be uses to automate reballing SOCs?
4:00 Why don't you retract and have the nozzle suck up the paste rather than filling it yourself? any air left at the top will be the same no matter the refill as long as you use the same syringe and nozzle, makes it easier to compensate for the spring factor
That's the same method 3D printers use.
I'm surprised he didn't think of that himself. Or maybe he did but decided against it for some reason?
If you can figure out a way to make it carve PCBs reliably without horrible etchants you are getting very close to complete home electronics production. Then of course you would have to take another step back and find a way to make your own PCB blanks from copper stock and force of will.
hang tight, i have a couple videos in the pipeline that may be going down that path...
@stephen_hawes The year is 2037. The Lumen v18 is able to go and mine copper fully autonomously. The Lumen v19 release candidate changelog casually mentions adding full sentience as a feature.
Wow this is brilliant!
I think you'll need an even higher gear ratio, especially to handle fine-pitch ICs rather than passive parts to overcome the back pressure from a finer-tipped lure lock.
3:30 …AND great taste!
SOLUTION: auger blades on the plunger. They bite into the solder paste so when you "retract" solder you have excellent "grip" on the paste and have very fine control no matter the viscosity (within reason). This solves the temperature / viscosity / reliability / consistency problem. Edit: the extra control you get will also help meet your fine pitch requirements.
cool af
Great work! I am working on a similar thing as well. Why doesn't the syringe lead screw rotate along with the gear?
How cool would it be to use liquid PVA glue to create dissolvable separation layer between supports and 3D printed part?
If it were possible, it would be good to use a worm gear drive. It would take way less power and have better control. I don't know how well it would 3d print or if that would be possible because of the rubbing wear.
.. also test how the bubbles in the solder paste behaves in an ultrasound cleaner.
The minimum amount of paste you need on a pad is "some"
Ok, that's huge. Now next logical step is DIY heater that is connected with Lumen by motorized conveir belt. You place PCB, solder paste, pick&place, then move it to your oven and you got fully assembled PCB.