I found it to be very reliable to use the laser to etch PCBs instead of trying to route them. I spray a coat of black enamel spray paint. Laser etch the circuit, then into copper etchant like the usual method. I use a cheap 200 dollar solid state laser.
Yeah I was wondering if some kind of laser or optical process would give better results, though I guess if you're drilling at all you'd have more tool offsets to worry about.
The whole point of doing DIY PCBs is that you're done in a couple of hours, rather than waiting a month from China. You can do iteration after iteration. Then once you get it right, you order from the cheap chinese companies.
@@1014p I would never send anything there to rando PCB companies if I was designing something for sale, but like it or not, nearly everything which you want assembled for cheap would have to be sent there anyway.
It doesn't take a month for JLPCB maybe 2 weeks tops. It literally took longer for me to receive PCBs from OSHpark located in the US than it did from from JLPCB. I will say that OSHPark's silkscreen is amazing and the cheap silkscreen option at JLPCB doesn't do a great job with fine detail graphics but it's not a big deal and works perfectly fine for footprints.
@@atomicskull6405 I ordered something from jlcpcb just the other day and lead time was 28 days. Probably has to do with location (I'm in the US). Even if it took 2 weeks, that's 2 weeks vs 2 hours.
What a prototyping powerhouse! If I'd just do pcb's I'd probably keep looking, but if I needed a whole suite, I'd go with this! This and a 3D printer could make anything.
That's a nice unit. A good review also. I use a cheap CNC for PCBs if I want one and also only single sided. I've learnt lots of techniques to make the cutting work well. I use Diptrace to design my boards. From there I use bitmaps for the track exports into Vectric 2D and then out to Mach3 for the cutting. I have made dozens of boards this way. Works a treat.
I am honestly impressed by this machine. I worked as a professional cnc CAM, CAD, 3D Modeling and 6 Axis CNC machinist for years but what I really want is a small tabletop machine that can do everything you've laid out in this video. I am 95% there to just ordering one of these.
Honestly the software missing is more a pro than a con for me, because it means you can use absolutely any program. Having a propietary software could be the start of limiting the machine to operate exclusively with their software so you end up forcefully relying on them rather than just searching for other program to run, like already happens with some 3D printers
I see your point, but there is value in having software that's bespoke to the machine from an end-user standpoint. It just *works* out of the box. Though I also feel it should allow G-code from any source provided it's set up for the machine.
Yeah, I've seen the downfall of many Kickstarter 3D printers by up-and-coming companies, being that they use proprietary software. They might look slick and professional on the outside but trade away a lot of practicality.
perfect machine is a dream. the reality is that PCB manufacturing is a constantly evolving process with a lot of people constantly tweaking it to overcome different shortcomings stemming from software, mechanics and chemistry. this is why you should really rely on a manufacturer for any serious work
Very nice. I would suggest having a look at FlatCAM for milling, especially 2 sided PCBs. Works really well and was able to get all the holes to line up etc on a 2 sided PCB as well. Best of all its free
@@MSP_TechLab yes it is , their motto is beta is always better 🤣🤣 The UI is pretty buggy to be honest, but the functionality works great for me, helps flip the gerber files around for milling the back side, even flip the holes around too so I can drill them last. Needs a bit of getting used to and not perfect, but works for me
I just bought $100 worth of supplies for thermal toner transfer and ferric chloride etching. I'm still using EasyEDA to design and print the patterns to paper, but now I can produce the final PCB's myself in about an hour with drilling and via's. I used to be a CNC programmer while working through collage and would love to have a nice mill, but I'll still be doing the toner transfer method for my PCB projects. Why - because once you get the hang to it - it works! I'd say it's better then using a CNC, no matter how much they try to perfect it.
Honestly watching this reminds me of how I used to proto boards and I don't miss it! Very grateful for overseas PCB houses being so cheap we don't have to deal with this anymore.
I've had great results from a generic "3040" ballscrew CNC that I've upgraded in three ways: 1) I replaced the odd spindle motor they provided with a high-quality 24000 RPM 3-phase spindle. 2) I swapped out the stock steppers and moved to closed-loop stepper/servos. 3) I added a probe input so I can capture a height map of the board before I ever cut it. Oh, and I chucked the printer-port stepper driver and went with a GRBLHAL networked controller that supports a ton of I/O and 5 axis. Anyway, the results are amazingly good, and definitely solid for creating your own boards, though obviously it's not going to add vias for you.
I've been photo-exposing and liquid-etching PCBs for 45 years. The detail, speed, and low cost obviously surpasses the messy and slow CNC junk. Also 30 years ago, I started using toner-transfer methods using my first affordable laser printer. Building a UV exposure box is cheap and simple. Building a heated etching tank is also not a big deal. Laser printers are now very cheap and the toner-transfer media is still available and easy and fast to use. That said, with the advent of fast and low-cost high-quality manufacturing (PCBWay, JLCPCB, and many others), I only use home-made methods when I'm in a huge hurry, which isn't often. Why anyone would ever want to go CNC is well beyond my comprehension.
If people are already using a CNC for other purposes it makes sense. Dealing with acid etching materials, and disposing of it isn't the nicest thing either.
I made a simple PCB for a passive crossover for a speaker im Building. Carved along the traces with a 90° v-bit and then added solder to the traces so they can handle more power.
I've been watching your videos for years but honesty my observation is that newer videos are not all that engaging than older videos. But all in all I appreciate your great work sir.🙏
That UV exposure solder mask can be used with an additional mask that you print on a transparency film and lay over the circuit so you can prevent the UV exposure to the areas you want exposed pads. The unexposed mask can then be removed with solvents (acetone I think?) and the pads exposed without the additional milling operation. The Carvera is expensive but is already a LOT cheaper than a comparable machine that has been on the market for a few years. I worked with the LPKF circuit board CNC for a while and those are a $30-60k machine. I went through the additional training classes they offered and still had issues with damage due to tool changer issues. It did produce nice boards when working right and it had decent software with good support so there is that.
Thank you Scott for taking the time and effort to look for these potential homemade solutions to making PCBs. I am hoping that as the technology develops something like the PCB maker you reviewed a couple of years back may improve and come down in price. $4700 (before tax and shipping) isn't exactly something a hobbyist can shell out. I'm hoping that it is just like 3D printers and the price will come down. I hope you keep looking for a good solution. I wonder if you might consider checking out a laser etched version like kwa mentioned in his comment below.
I built a rotary mill from scratch using Linux CNC on a PC I found on a rubbish dump(plus about $400 dollars in hardware). The reason I comment here about it, is that I use the Nefertiti bust, as you displayed, to test and tune the machine.
I had a question about the tool clamp size and type but found the answer on their site, so I'll leave the info here. Comes with 1/8" spindle collet, 4mm, 6mm and 1/4" also available separately. The wireless probe tail shaft can be changed to suite the different sizes.
My last design from JLC was almost perfect except that - yet again - some of my holes were a bit too small. They fit each pin just fine but the ESP32 dev board has slightly angled headers already installed. The holes are just small enough that it’s really hard to get them installed :) That’s one thing that I’d like about making them at home. I can fix mistakes more quickly.
I have a Carvera. Was backer number 6. Love it, haven't done much pcb milling but it works when Im testinh things and dont want to wait a few days for a pcb, Also great for aluminum milling.
I was getting reasonable results with a modified 3040 and FlatCAM (double sided). This was down to SOIC without any levelling which is all I needed at the time. Though cheap boards and cheap cutters I found could tear out the surface material. I should retry now with my even more modded machine and sw levelling to see how well I can get down to TQFP64 :) Nice video, so hard to try to cover so many aspects in such a short time!
Plated-thru holes, finish (HASL or ENIG), and solder-mask are must-haves for me. I can buy a LOT of top-quality PCBs from OSH-Park or PCBway for much less than the cost of a PCB milling machine.
Sounds like a good toll for rapid prototyping. If you have an finished Design, I would go with an professionel Producer like JLC. But during the Designphase there you have to: Try, fail and repeat until it works. Sunch an CNC sounds like an good option for multiple iterations.
Good for catching the little mistakes. We all make them, but having to modify and/or toss out a stack of PBCs because of a small mistake that you could have caught with a quick prototype, isn't good.
This sponsor or review video, is really nice. Its not perfect, but it had a lot of qualities that you really liked and could make good use of. Im excited to see the new projects you make with it! EDIT: It is a review.
Thanks to you, I have made a functional plasma ball based off a 555 timer and mosfet! It really works well and I have it powered up at the moment and it's barely even hot and I hope it doesn't blow up sometime soon. Thank you for all of your electronics basics and project videos! -Jon.
It's one thing to get good results from an example that has all the speeds and feeds set from you, it's another thing for the learning curve of figuring that all out for yourself. So I view your "I made this from scratch" a much more valid review.
use neutral oil for milling pcb, the result is perfect and more accurate :) i have a cheap cnc like 3018 and it works perfectly, and i love your channel !
I worked as a biomedical engineer back in the 80's and I'd setup a system that did something similar. And boy, it was a booger to use. The mill was ok in the x & y axis, but z depth was controlled by a solenoid and thumb screw. The PCB had to be perfectly flat, or else. And, as far as I knew, router bits that small weren't available. A conductive ink was applied to the thru holes, then copper was plated to the boards in a plating machine . (Keeping the chemistry right was also a pain.) Then photo resist was hot rolled on from a carrier film, followed by normal photo resist procedures. Took me many weeks to get the procedures down, and in the end, we never really used the system. Better than running a test scoring scanner when I first started working, though.
Definitely the features are paid for, they are certainly very useful and I would say that if it arrived better packaged it would be a good thing. Like any complex machine, there are always adjustments to be made during installation. Surely it is a nice top of the range, I dare say for semi-professional use, the hobbyist will hardly be willing to spend that amount to use it "three times a year" in all cases the range of accessories seems well developed even if the $ 299 for the PCB kit it seems to me a further increase in costs. In the images I saw that the virgin copper plates due to their elasticity risk rising, to solve the problem you could try with double adhesive tape between the pcb and the work surface. One thing to rationalize to reduce the movement of the spindle could be the zeroing of the height which could be done near the tool holder and not at the other end of the work plane, possibly if the zeroing did not concern only the Z axis it could be it is useful to place the tool holder to the left of the work surface. I see you soon!
If you 'just' want to make a PCB, try use a resin printer (without the VAT but with a photo resist coated pcb) as the printer provides the UV source, and the LCD can generate the required pattern. Expose the layout as layer 0 for the required time, develop and etch 'as usual'
Wow! That sounds great! I feel silly for having not thought of that…I keep looking every so often to see if there’s a cheap and easy way to do this. Transparencies and a UV table seem to require careful choices of materials and stuff that I’m afraid I will have to keep constantly experimenting with to get right.
Only downside is that one 100mm^2 piece of presensitized copper clad is probably going to cost more than 5 copies of a 100mm^2 from JLC or PCBWay, including shipping.
@@treelineresearch3387 You could be right - I don't know current prices, but I know this: - It is possible to get appliable photoresist, either in a spray bottle or as a film you apply. Great if you have a stack of bare Cu boards. - It is faster than waiting for a PCB manufacturer to produce and shipping time can be *long* especially if you're in a rush - You keep your design private
@@EgonSorensen Yeah I can see corporations/consultancies wanting to keep stuff in-house for quicker turnaround and higher security, but just not really a concern for me most of the time, especially given how much postprocessing a fab-at-home board already requires even with presensitized stock (bonding vias, solder mask, etc). Corporate also has different demands and much bigger budgets for materials, dedicated wet/dry lab space, and tooling, than I do so the equation is pretty different.
@@treelineresearch3387 I see two versions of a product: Prototype and Production. Imho a prototype rarely needs soldermask, but it can also be applied using the resin printer - some experimentation may be needed, esp. if working with 01005's ;ø) Via's can be easily made with rivets, if a double sided board. 2+ layered isn't a DIY project usually, unless you want to do your own multi-layered 2-layer 0.8mm's + more experimentation, gluing, etc. Prototypes are excellent for concepts, or *very* small batch jobs. For everything else, yes - it does not make sense to make pcb's at home any more. Etchants, developer and soldermask can/could be obtained for relatively low amounts of money - and stores for a while. Also doing it at home could save on import charges and fees. Having the post office/courier (read: required) doing customs clearance often adds a substantial amount of money onto the price of the PCB's - depending on where you live.
As a portable cnc and laser etching machine, it seems more than capable... For circuit boards, probably best at laser etching... The 3d capabilities alone make this a very desirable machine...
4:11 Oh my god, cases with slanted io Or usb ports gives me nightmates. I had a case that was like this and my 40euro usb headphones kept breaking, I have spent 120 euros on many headphones and even the 3.5mm jack ones have the jack have a loose wires bcs of gravity making it slant and my usb headphones had the DAC in the usb itself as a board, the usb kept ripping off the boards WITH THE SOLDER PADS so I couldn't repair it. Seeing you use a pc case with slanted I/O gave me nostalgia
Looks like a nice machine, but those scratches would have upset me greatly if I had purchased it. Particularly on the cover! Could you do testing on different PCB fabrication techniques, like professional from JLCPCB vs. CNC milled vs. DIY chemically etched?
I have a Workbee CNC and it was bought mainly for PCB prototypes. Just like you have said the software side is still very much work things out for your self and far from perfect. I can use it with decent results including double side and have made a few prototypes before sending away for pro boards to be made. This machine looks good and the tool change will save a lot of time.
For double sided, I imagine the “registration marks” (or zero X and Y starting positions) when flipping over must be near perfect. And the drilling of holes has to hit as close to dead center as possible.
@@stevebabiak6997 Depending on the complexity of the design yes it can be a problem to get it perfect. I use flatcam and it has a way to do it, but if you don't do it in the correct order everything will be wrong. It was a good 6 months of messing about before I had a good process to make boards.
Your videos are great as always and in fact I have to thank you for your videos . They really helped me in my school project that is about 20%from my total grade and without your videos I don't think I would be able to make it. It was a small robot that measures the effects of climate change 🙂
If all you need is a one-off or time sensitive board, pcb mills work great. Leveling is CRUCIAL since the cuts are do thin. This looks like a dream, but way outside my budget! In reality, you can get a decent milling machine for kess than $150. You can even add a laser to most of them for a cost proportional to the wattage of the laser.
It would be interesting to see all kinds of different pcb manufacturing technologies (classic, flexible, etc.) from you. Yes we can pay somebody to make a pcb, but wouldn't it be great to produce it themselves?
Jlpcb is pretty awesome. I had a project that needed some custom pwm controllers. So I ordered the minimum amount of boards so I have 4 of those boards in a drawer that will probably never be used again. But I always save stuff because you never know when something will break. My philosophy is to keep spares around if you can. I can’t count the number of times I’ve saved the day because I healed on to a extra component for 15 years and they were no longer available and would have required a lot of work to integrate a newer version of the item into the system and sometimes it is almost impossible to integrate the new part.
You remember 10 years ago I worked in an advertising company, with very good engraving machines for professional work. A 24x24 inch (60x60 cm) table costs $30,000 (with no additional accessories) and you think your hobby machines are expensive
I went to a vocational high school and took electronics. We learned how to make our own PCBs with those fun chemicals. Now, I remember my shop teacher talking to us about how hard it is to get the right chemicals for removing copper from a PCB and how easy it used to be. His exact words for the description of why we can't use better chemicals was: "You know what happened? 9/11 happened."
When clamping thin flexible stock like PCBs, don't have your clamps at an angle like you did in the video. That warps the stock! Instead, put some packing under the FAR end of the clamp, slightly thicker than the pcb. Now the clamp will only apply downward clamping force at its inside edge and your PCB will be pressed downward and held neatly without mashing and warping it.
Was a nice video. Could you go a bit more into detail about the fabrication of the solder mask, or create a separate video about it? You apply the mask, cure it and then remove the cured mask from the pads? So if the machine is well calibrated, I would not remove any copper and would not have to create a mask for the solder mask and etch it but how do I calibrate the machine for such low tolerances as I modify the pcb when applying the solder mask. Not sure if my comment makes sense to you ...
I know there's quite a few things a CNC mill can do that a 3D printer can't, but I think a 3D printer is just way better value for money. And I can make PCBs at home using press-n-peel toner transfer film, which just prints straight out of KiCad. In fact, I just finished a design that I'm going to print out and etch this weekend. If I really want a solder mask, I can use the old fashioned photographic transfer method with some dry film solder mask. But it usually only takes a few seconds to run the solder over all the traces to help stop them corroding, and I've never had any issues soldering SMD parts to home made boards, so I don't usually bother.
Watch wegstr cnc. If you just take the time and have more attention to detail you can get really good looking pcbs. For example, using the centering holes for double sided, using cheese cloth for the solder mask application, using a tinning liquid to tin the copper pads, and if you're feeling really adventurous using multiple solder masks (white, then green) and a laser to etch away the green for circuit legends. I think this hardware is more than capable of making very nice pcbs at home.
11:05 it should be pretty easy to do double sided designs without all that rivet nonsense. I do similar things on a daily basis milling aluminum on a big 3mx2m CNC router. It is a matter of having sacrificial material and cut the negative of the outside perimeter to make a hole to stick the pcb with the other side up. That way you do have a reference to align the PCB for the other side. To makes things way easier, put one of the PCB corners as the 0,0 then put the zero manually so the contour falls inside the sacrificial material. Make the hole undersized on purpose then increase it's size little by little trying the PCB until it fits. Ideally you'd have a table with suction, just so it is a drop and done affair, but you can always cut double sided tape to size for that purpose.
It seems like a lot of companies seem to shoot for the $4000.00 price point when designing these kinds of things (Glowforge). But most "hobbyists" don't really have this kind of "disposable" income to spend on these things.
I have a many years-old machine that I purchased new and is very similar except for the enclosure, dust collection, and ATC. This machine looks like it could be a version 2 of that particular machine, if only the manufacturer of my machine did not go out of business after a rather successful crowdfunding and decent marketing/ sales to schools. My point is that at the current price point it is not likely that this is going to be something you see in every workshop and makerspace. For less money I purchased another, and far more capable CNC, at least in terms of work are (over 6x in x-y and roughly 10x in z) and spindle (1.5kW), although again without enclosure, included dust collection, or ATC. Let's hope for people who purchase this that the company does not have unrealistic sales expectations that sink the company rapidly and leave the purchasers without support, as has happened time and time again with 3D printing and small CNC routers/ milling machines. This final issue is why I now feel that for the hobbyist or very small business market it is far better to get one of the mass-produced import machines that use off-the-shelf components and standard extrusions. They require more tinkering, but they also are extremely simple to repair or upgrade, because they use off-the-shelf components and standard extrusions.
Thank you, I have two questions first when driiling the PCB,There is a metal base under PCB,will not broken as will touch the metal PCB,Second qustion are there protection from laser.Thanks
I've been developing CAM software designed primarily for CNC art/signmaking/engravings, and it works fine for PCB milling if you have a good file format that can be loaded in (i.e. images, vectors). It's called PixelCNC. It's still in beta and right now I'm working on a built-in tutorial system. If you have any suggestions for tutorials or need help with anything shoot an email to us.
You should try FlatCAM to process the gerber files into GCODE. The program is a bit unstable, but it has a lot of tools that allow you to get your job properly done. I recommend watching a few tutorials first, because the program may not be that intuitive at first use.
JLCPCB wait time for me in Charleston South Carolina in the United States is now over three weeks and I still haven't received the package using the ground shipping. What they don't tell you is that if you want it super fast you're going to pay 30 or $40 more for the shipping. So it's more expensive than they lead you to believe.
Looks like a decent enough mini mill; and having an ATC spindle helps set it apart from the typical 6040 style routers. However, for PCBs, I think I’m gonna stick with outfits like JLC, PCBWay, etc…
Fully agree, why spend thousands of dollars and hours of time only to get a much worse product than what can be bought for very little money from a skilled company.
To prepare GRBL to GCODE PCBs I use Flatcam Beta 8.994, it is the absolute best software for that purposes and there are tons of videos on YT about how to use it. (it's a bit tricky to find the beta but not impossible)
Just to set the record straight, there are a few more downsides to using board shops (and this is not directed at any shop in particular). First though let me point out that they (in my experience) deliver great work and high quality pcbs (and associated services), credit where credit is due. But (for the small maker with some labor of love designs that you try to market to like-minded people) : - Shipping cost, import fees, taxes ... All of these are excessive where I live. I have had times (depends on getting lucky with customs) where the PCB's cost me 2 dollars and shipping + fees + taxes cost me 120 Euros. And there have been times where the PCB's cost me 40 dollars and the shipping + fees + taxes cost me 30 Euros. Aside from the obvious 'wtf is with these add-on costs' one has to realize that it is a crapshoot if you actually need to calculate this cost into your end product. - Protecting your intellectual property, and this is a general concern not specific to any board shop. It might not be a fun thing to consider but if you make a good design that garners some degree of interest or appreciation it will be stolen. Sending out your design to an external party is one of many ways in which this will happen. By no means the only one, but one none the less. - You pointed this one out yourself, speed, again most board shops do amazing work and fast but shipping will slow down things considerably. If you are making fast iterations on a new design this can be a problem. A system like the Carvera, the Wegstr cnc or other desktop PCB solution is going to help out with that a lot. On the up side, if you need quantity and quality, you will eventually have to use a board shop of some sort.
Honestly, the biggest issue with PCB's is via's. Using conductive paint and a chemical process is one option. But then again, if you're going chemicals anyways, you might as well use a SLA printer for the UV & solder mask, tinning via chemicals, etc... If you think about it, it almost sounds easier to just buy them for $3 from jlc.
I love these machines, but for pcb I think there is still a lot to improve. And I think there is technology available for that, it just takes a company to take the matter more seriously. With a laser, you can do something too, but it's not 100%. I keep using breadbords and when I need something better, I order from JLCPCB.
Not to be too salty, but wanted to mention that I supported the kickstarter and have seen the fulfillment of kickstarter orders pushed back at least 3 times. Original delivery date was, iirc, August 2022 (currently end of Jan 23 and they still have something like 400 machines left to ship)
Having worked with copperCAM before i can for sure say it has alot of problems. The only other issue is with the method of applying the mask.Professional pcb manufacturers use a silk screen for the mask and component designators. If you also use screen printing you will get very professional looking pcbs.Other than that its a solid machine but its priced for a small buisness rather than a hobbyist like most of us are.
I think knowing what limitations and benefits professional machines have really puts things into perspective. For reference. My single operation, mass production milling machine with a 20 cm by 10 cm (4x9 inch) working surface Costed $30,000 USD Weighs 408 kg 900 lbs (without the coolant and oil) Only cuts one profile or slot at a time Is much faster, and takes much less attention, but takes a *college degree* to set up reliably, all tools must be custom designed and made for the each cut and must be programed in *assembly commands* Has no sensors at all. Can be set up and checked in *once per week* when running. Costs $.10 per hour to run, and about $500 to $2,000 to set up. *PER feature* Is not CNC obviously, although is computer controlled, it is not numeric. all tools can be resharpened. Another professional model I have used that is a mill-drill center using G-Code, made by Fanuc, with about the same table size. Has 8 tool slots, Has no ability to sense the table Takes 3 phase 410 volt. *is significantly faster moving* Weighs in at 650 kg, 1450 pounds (without coolant) Does *NOT* come with software. Requires professional attention about once per hour. Costs about $80,000 in bare bones, used models can cost up to $200,000 with all the options and low hour run times. Costs about $100 to $200 per hour to run. Take your trade offs, don't complain, unless the cost and performance do not add up.
What is so special about that machine that it requires a COLLEGE DEGREE (instead of just some training/certificate) to set up, and why does it cost up to $2k per feature? Unless you are talking about the CAD/CAM part of it as well, including the actual PCB design, then i can somewhat understand it. And why does that other one cost $100/hour to run just for milling/drilling? Lots of wear parts?
I use bantam tools cnc for milling pcb's and am able to do double sided PCBs with around the same quality as this. The only downside is they don't have a way to create a height map. Other than that it is very easy to use.
Love the automatic tool changer. This is something I'd love to have, but I can't quite justify yet. I'm still trying to justify my laser cutter/engraver. :)
I found it to be very reliable to use the laser to etch PCBs instead of trying to route them. I spray a coat of black enamel spray paint. Laser etch the circuit, then into copper etchant like the usual method. I use a cheap 200 dollar solid state laser.
Good idea :-)
Yeah I was wondering if some kind of laser or optical process would give better results, though I guess if you're drilling at all you'd have more tool offsets to worry about.
As an eastern european "cheap" and "200 dollars" don't mix too well for me 😅
@@DiaconescuAlexandru2024 It's cheap compared to the multiple thousands of dollars that many lasers cost.
@@skylerlehmkuhl135 oh
The whole point of doing DIY PCBs is that you're done in a couple of hours, rather than waiting a month from China. You can do iteration after iteration. Then once you get it right, you order from the cheap chinese companies.
That is correct :-)
Sure if your cool with giving away your work. Surely your aware China takes copies of everything sent to built there for technology.
@@1014p I would never send anything there to rando PCB companies if I was designing something for sale, but like it or not, nearly everything which you want assembled for cheap would have to be sent there anyway.
It doesn't take a month for JLPCB maybe 2 weeks tops. It literally took longer for me to receive PCBs from OSHpark located in the US than it did from from JLPCB. I will say that OSHPark's silkscreen is amazing and the cheap silkscreen option at JLPCB doesn't do a great job with fine detail graphics but it's not a big deal and works perfectly fine for footprints.
@@atomicskull6405 I ordered something from jlcpcb just the other day and lead time was 28 days. Probably has to do with location (I'm in the US).
Even if it took 2 weeks, that's 2 weeks vs 2 hours.
What a prototyping powerhouse! If I'd just do pcb's I'd probably keep looking, but if I needed a whole suite, I'd go with this! This and a 3D printer could make anything.
That's a nice unit. A good review also. I use a cheap CNC for PCBs if I want one and also only single sided. I've learnt lots of techniques to make the cutting work well. I use Diptrace to design my boards. From there I use bitmaps for the track exports into Vectric 2D and then out to Mach3 for the cutting. I have made dozens of boards this way. Works a treat.
7days ago?
@@mojiburrahman3956 WHATTT
@@mojiburrahman3956 patreon
@@mojiburrahman3956 Black magic
@@mojiburrahman3956 Yes Patreon 🙂
I am honestly impressed by this machine. I worked as a professional cnc CAM, CAD, 3D Modeling and 6 Axis CNC machinist for years but what I really want is a small tabletop machine that can do everything you've laid out in this video. I am 95% there to just ordering one of these.
Honestly the software missing is more a pro than a con for me, because it means you can use absolutely any program.
Having a propietary software could be the start of limiting the machine to operate exclusively with their software so you end up forcefully relying on them rather than just searching for other program to run, like already happens with some 3D printers
So long as they let you use any program and those options are available. Otherwise, you have a very expensive brick.
*cough* ChiTuBox *cough*
Yes, and then tomorrow when they stop making it, lose interest, or go under you own a $4000 brick.
I def agree on this.
I see your point, but there is value in having software that's bespoke to the machine from an end-user standpoint. It just *works* out of the box. Though I also feel it should allow G-code from any source provided it's set up for the machine.
Yeah, I've seen the downfall of many Kickstarter 3D printers by up-and-coming companies, being that they use proprietary software. They might look slick and professional on the outside but trade away a lot of practicality.
You certainly cannot beat professional manufacturing with any diy/homemade machines/pcb's. That's my point,and I do not wanna argue with anybody.
perfect machine is a dream.
the reality is that PCB manufacturing is a constantly evolving process with a lot of people constantly tweaking it to overcome different shortcomings stemming from software, mechanics and chemistry. this is why you should really rely on a manufacturer for any serious work
Very nice. I would suggest having a look at FlatCAM for milling, especially 2 sided PCBs. Works really well and was able to get all the holes to line up etc on a 2 sided PCB as well. Best of all its free
However, it is still beta and has some glitches from time to time. But it seems there is no anything better in open-source.
@@MSP_TechLab yes it is , their motto is beta is always better 🤣🤣 The UI is pretty buggy to be honest, but the functionality works great for me, helps flip the gerber files around for milling the back side, even flip the holes around too so I can drill them last. Needs a bit of getting used to and not perfect, but works for me
I just bought $100 worth of supplies for thermal toner transfer and ferric chloride etching. I'm still using EasyEDA to design and print the patterns to paper, but now I can produce the final PCB's myself in about an hour with drilling and via's. I used to be a CNC programmer while working through collage and would love to have a nice mill, but I'll still be doing the toner transfer method for my PCB projects. Why - because once you get the hang to it - it works! I'd say it's better then using a CNC, no matter how much they try to perfect it.
Use a 3d resin printer with UV sensitive pcb material.
I enjoyed the fact that this particular video was sponsored by JLCPCB.
ha, ha
Honestly watching this reminds me of how I used to proto boards and I don't miss it! Very grateful for overseas PCB houses being so cheap we don't have to deal with this anymore.
I've had great results from a generic "3040" ballscrew CNC that I've upgraded in three ways: 1) I replaced the odd spindle motor they provided with a high-quality 24000 RPM 3-phase spindle. 2) I swapped out the stock steppers and moved to closed-loop stepper/servos. 3) I added a probe input so I can capture a height map of the board before I ever cut it. Oh, and I chucked the printer-port stepper driver and went with a GRBLHAL networked controller that supports a ton of I/O and 5 axis. Anyway, the results are amazingly good, and definitely solid for creating your own boards, though obviously it's not going to add vias for you.
I've been photo-exposing and liquid-etching PCBs for 45 years. The detail, speed, and low cost obviously surpasses the messy and slow CNC junk. Also 30 years ago, I started using toner-transfer methods using my first affordable laser printer. Building a UV exposure box is cheap and simple. Building a heated etching tank is also not a big deal. Laser printers are now very cheap and the toner-transfer media is still available and easy and fast to use. That said, with the advent of fast and low-cost high-quality manufacturing (PCBWay, JLCPCB, and many others), I only use home-made methods when I'm in a huge hurry, which isn't often. Why anyone would ever want to go CNC is well beyond my comprehension.
If people are already using a CNC for other purposes it makes sense.
Dealing with acid etching materials, and disposing of it isn't the nicest thing either.
I made a simple PCB for a passive crossover for a speaker im Building. Carved along the traces with a 90° v-bit and then added solder to the traces so they can handle more power.
Like I always say, DIY is always fun and the best in some occasions!
Haha yes. Definitely useable in certain applications.
@@greatscottlab Yes!
This looks like a fantastic option for prototyping, combined with sending out for final product.
I think so too ;-) Still some software limitations but that will hopefully get fixed soon :-)
I've been watching your videos for years but honesty my observation is that newer videos are not all that engaging than older videos. But all in all I appreciate your great work sir.🙏
That UV exposure solder mask can be used with an additional mask that you print on a transparency film and lay over the circuit so you can prevent the UV exposure to the areas you want exposed pads. The unexposed mask can then be removed with solvents (acetone I think?) and the pads exposed without the additional milling operation.
The Carvera is expensive but is already a LOT cheaper than a comparable machine that has been on the market for a few years. I worked with the LPKF circuit board CNC for a while and those are a $30-60k machine. I went through the additional training classes they offered and still had issues with damage due to tool changer issues. It did produce nice boards when working right and it had decent software with good support so there is that.
Thank you Scott for taking the time and effort to look for these potential homemade solutions to making PCBs. I am hoping that as the technology develops something like the PCB maker you reviewed a couple of years back may improve and come down in price. $4700 (before tax and shipping) isn't exactly something a hobbyist can shell out. I'm hoping that it is just like 3D printers and the price will come down. I hope you keep looking for a good solution. I wonder if you might consider checking out a laser etched version like kwa mentioned in his comment below.
I built a rotary mill from scratch using Linux CNC on a PC I found on a rubbish dump(plus about $400 dollars in hardware). The reason I comment here about it, is that I use the Nefertiti bust, as you displayed, to test and tune the machine.
I had a question about the tool clamp size and type but found the answer on their site, so I'll leave the info here.
Comes with 1/8" spindle collet, 4mm, 6mm and 1/4" also available separately. The wireless probe tail shaft can be changed to suite the different sizes.
I recently got my first boards from JLC pcb and they worked great despite my questionable design!
Great to hear!
My last design from JLC was almost perfect except that - yet again - some of my holes were a bit too small. They fit each pin just fine but the ESP32 dev board has slightly angled headers already installed. The holes are just small enough that it’s really hard to get them installed :)
That’s one thing that I’d like about making them at home. I can fix mistakes more quickly.
8:50 This must be the perfect TH-cam video, you used all three of your sponsors together!
I have a Carvera. Was backer number 6. Love it, haven't done much pcb milling but it works when Im testinh things and dont want to wait a few days for a pcb, Also great for aluminum milling.
I was getting reasonable results with a modified 3040 and FlatCAM (double sided). This was down to SOIC without any levelling which is all I needed at the time. Though cheap boards and cheap cutters I found could tear out the surface material. I should retry now with my even more modded machine and sw levelling to see how well I can get down to TQFP64 :) Nice video, so hard to try to cover so many aspects in such a short time!
Plated-thru holes, finish (HASL or ENIG), and solder-mask are must-haves for me. I can buy a LOT of top-quality PCBs from OSH-Park or PCBway for much less than the cost of a PCB milling machine.
Sounds like a good toll for rapid prototyping.
If you have an finished Design, I would go with an professionel Producer like JLC.
But during the Designphase there you have to: Try, fail and repeat until it works. Sunch an CNC sounds like an good option for multiple iterations.
Good for catching the little mistakes. We all make them, but having to modify and/or toss out a stack of PBCs because of a small mistake that you could have caught with a quick prototype, isn't good.
The more I check it out the better it seems! Amazeballs xD Brilliant piece of tech. Need to have this. Make absolutely anything
This sponsor or review video, is really nice. Its not perfect, but it had a lot of qualities that you really liked and could make good use of. Im excited to see the new projects you make with it!
EDIT: It is a review.
Thanks to you, I have made a functional plasma ball based off a 555 timer and mosfet! It really works well and I have it powered up at the moment and it's barely even hot and I hope it doesn't blow up sometime soon.
Thank you for all of your electronics basics and project videos!
-Jon.
Fantastic!
Thanks
It's one thing to get good results from an example that has all the speeds and feeds set from you, it's another thing for the learning curve of figuring that all out for yourself. So I view your "I made this from scratch" a much more valid review.
A good alternative to coppercam is flatcam. I use it exclusively now when making prototype boards.
I like the new intro :) Thanks for showing us the machine.
For soldermask you can use silkscreen screen. And for better solderability i use cold tinning ^^. And for the vias there are rivets :)
use neutral oil for milling pcb, the result is perfect and more accurate :) i have a cheap cnc like 3018 and it works perfectly, and i love your channel !
I worked as a biomedical engineer back in the 80's and I'd setup a system that did something similar. And boy, it was a booger to use. The mill was ok in the x & y axis, but z depth was controlled by a solenoid and thumb screw. The PCB had to be perfectly flat, or else. And, as far as I knew, router bits that small weren't available. A conductive ink was applied to the thru holes, then copper was plated to the boards in a plating machine . (Keeping the chemistry right was also a pain.) Then photo resist was hot rolled on from a carrier film, followed by normal photo resist procedures.
Took me many weeks to get the procedures down, and in the end, we never really used the system.
Better than running a test scoring scanner when I first started working, though.
I’m a machinist and I would really love to have a mill like this and the laser with it is really an awesome addition.
Definitely the features are paid for, they are certainly very useful and I would say that if it arrived better packaged it would be a good thing.
Like any complex machine, there are always adjustments to be made during installation.
Surely it is a nice top of the range, I dare say for semi-professional use, the hobbyist will hardly be willing to spend that amount to use it "three times a year" in all cases the range of accessories seems well developed even if the $ 299 for the PCB kit it seems to me a further increase in costs.
In the images I saw that the virgin copper plates due to their elasticity risk rising, to solve the problem you could try with double adhesive tape between the pcb and the work surface.
One thing to rationalize to reduce the movement of the spindle could be the zeroing of the height which could be done near the tool holder and not at the other end of the work plane, possibly if the zeroing did not concern only the Z axis it could be it is useful to place the tool holder to the left of the work surface.
I see you soon!
Thanks for the feedback :-)
thank you from Morocco i learned lot of things from you and electro boom channels
Thanks for watching :-)
If you 'just' want to make a PCB, try use a resin printer (without the VAT but with a photo resist coated pcb) as the printer provides the UV source, and the LCD can generate the required pattern. Expose the layout as layer 0 for the required time, develop and etch 'as usual'
Wow! That sounds great! I feel silly for having not thought of that…I keep looking every so often to see if there’s a cheap and easy way to do this. Transparencies and a UV table seem to require careful choices of materials and stuff that I’m afraid I will have to keep constantly experimenting with to get right.
Only downside is that one 100mm^2 piece of presensitized copper clad is probably going to cost more than 5 copies of a 100mm^2 from JLC or PCBWay, including shipping.
@@treelineresearch3387 You could be right - I don't know current prices, but I know this:
- It is possible to get appliable photoresist, either in a spray bottle or as a film you apply. Great if you have a stack of bare Cu boards.
- It is faster than waiting for a PCB manufacturer to produce and shipping time can be *long* especially if you're in a rush
- You keep your design private
@@EgonSorensen Yeah I can see corporations/consultancies wanting to keep stuff in-house for quicker turnaround and higher security, but just not really a concern for me most of the time, especially given how much postprocessing a fab-at-home board already requires even with presensitized stock (bonding vias, solder mask, etc). Corporate also has different demands and much bigger budgets for materials, dedicated wet/dry lab space, and tooling, than I do so the equation is pretty different.
@@treelineresearch3387 I see two versions of a product: Prototype and Production.
Imho a prototype rarely needs soldermask, but it can also be applied using the resin printer - some experimentation may be needed, esp. if working with 01005's ;ø)
Via's can be easily made with rivets, if a double sided board. 2+ layered isn't a DIY project usually, unless you want to do your own multi-layered 2-layer 0.8mm's + more experimentation, gluing, etc.
Prototypes are excellent for concepts, or *very* small batch jobs.
For everything else, yes - it does not make sense to make pcb's at home any more.
Etchants, developer and soldermask can/could be obtained for relatively low amounts of money - and stores for a while.
Also doing it at home could save on import charges and fees. Having the post office/courier (read: required) doing customs clearance often adds a substantial amount of money onto the price of the PCB's - depending on where you live.
For a company its not to expensive. But for home use I don’t think the market will be that large. Really cool machine anyway.
Thanks for the feedback. Good assessment. I think so too ;-)
Looks very nice, I hope you get the issues ironed out. Thanks for sharing.
As a portable cnc and laser etching machine, it seems more than capable... For circuit boards, probably best at laser etching... The 3d capabilities alone make this a very desirable machine...
4:11
Oh my god, cases with slanted io Or usb ports gives me nightmates. I had a case that was like this and my 40euro usb headphones kept breaking, I have spent 120 euros on many headphones and even the 3.5mm jack ones have the jack have a loose wires bcs of gravity making it slant and my usb headphones had the DAC in the usb itself as a board, the usb kept ripping off the boards WITH THE SOLDER PADS so I couldn't repair it. Seeing you use a pc case with slanted I/O gave me nostalgia
Looks like a nice machine, but those scratches would have upset me greatly if I had purchased it. Particularly on the cover!
Could you do testing on different PCB fabrication techniques, like professional from JLCPCB vs. CNC milled vs. DIY chemically etched?
I can put it on my to do list
your video is so good and educational..love from bangladesh..
Thank you so much 😀
I have a Workbee CNC and it was bought mainly for PCB prototypes. Just like you have said the software side is still very much work things out for your self and far from perfect. I can use it with decent results including double side and have made a few prototypes before sending away for pro boards to be made. This machine looks good and the tool change will save a lot of time.
For double sided, I imagine the “registration marks” (or zero X and Y starting positions) when flipping over must be near perfect. And the drilling of holes has to hit as close to dead center as possible.
@@stevebabiak6997 Depending on the complexity of the design yes it can be a problem to get it perfect. I use flatcam and it has a way to do it, but if you don't do it in the correct order everything will be wrong. It was a good 6 months of messing about before I had a good process to make boards.
Your videos are great as always and in fact I have to thank you for your videos . They really helped me in my school project that is about 20%from my total grade and without your videos I don't think I would be able to make it. It was a small robot that measures the effects of climate change 🙂
That's great!
If all you need is a one-off or time sensitive board, pcb mills work great. Leveling is CRUCIAL since the cuts are do thin. This looks like a dream, but way outside my budget! In reality, you can get a decent milling machine for kess than $150. You can even add a laser to most of them for a cost proportional to the wattage of the laser.
Nice! Your gonna be unstoppable!
It would be interesting to see all kinds of different pcb manufacturing technologies (classic, flexible, etc.) from you. Yes we can pay somebody to make a pcb, but wouldn't it be great to produce it themselves?
I will see what I can do :-)
@@greatscottlab :^)
But wouldn't that be disrespecting the all-mighty JLCPCB?
@@brandontechnerd Of course! You would be straying from the PCBway into the savannah where a wild Aisler could bite you in the ankle!
A 4th axis and auto tool changer are super awesome.
Jlpcb is pretty awesome. I had a project that needed some custom pwm controllers. So I ordered the minimum amount of boards so I have 4 of those boards in a drawer that will probably never be used again. But I always save stuff because you never know when something will break. My philosophy is to keep spares around if you can. I can’t count the number of times I’ve saved the day because I healed on to a extra component for 15 years and they were no longer available and would have required a lot of work to integrate a newer version of the item into the system and sometimes it is almost impossible to integrate the new part.
hohohoho omg nice! Might be the best desktop machine besides AdditiveProductionMachines. Very impressive. So complete
You remember 10 years ago I worked in an advertising company, with very good engraving machines for professional work. A 24x24 inch (60x60 cm) table costs $30,000 (with no additional accessories) and you think your hobby machines are expensive
വെയ്റ്റിംഗ് ആയിരുന്നു നിങ്ങളുടെ വീഡിയോയ്ക്കു വേണ്ടി 👍
I use FlatCam with my cheap CNC and it's pretty awesome
would love a shot of it, but far above my humble price point
I went to a vocational high school and took electronics. We learned how to make our own PCBs with those fun chemicals. Now, I remember my shop teacher talking to us about how hard it is to get the right chemicals for removing copper from a PCB and how easy it used to be. His exact words for the description of why we can't use better chemicals was: "You know what happened? 9/11 happened."
💀💀💀
When clamping thin flexible stock like PCBs, don't have your clamps at an angle like you did in the video. That warps the stock!
Instead, put some packing under the FAR end of the clamp, slightly thicker than the pcb. Now the clamp will only apply downward clamping force at its inside edge and your PCB will be pressed downward and held neatly without mashing and warping it.
consider using flatcam for gerbers. it's not really advancing anywhere now, but current state is very capable for making pcbs
How would you compare the experience of milling a PCB to etching? I found the photoresist process to be painless and reliable in the past.
Was a nice video. Could you go a bit more into detail about the fabrication of the solder mask, or create a separate video about it? You apply the mask, cure it and then remove the cured mask from the pads? So if the machine is well calibrated, I would not remove any copper and would not have to create a mask for the solder mask and etch it but how do I calibrate the machine for such low tolerances as I modify the pcb when applying the solder mask. Not sure if my comment makes sense to you ...
I know there's quite a few things a CNC mill can do that a 3D printer can't, but I think a 3D printer is just way better value for money. And I can make PCBs at home using press-n-peel toner transfer film, which just prints straight out of KiCad.
In fact, I just finished a design that I'm going to print out and etch this weekend. If I really want a solder mask, I can use the old fashioned photographic transfer method with some dry film solder mask. But it usually only takes a few seconds to run the solder over all the traces to help stop them corroding, and I've never had any issues soldering SMD parts to home made boards, so I don't usually bother.
4000 is insane, considering the results...it is still wiser to continue using JLCPCB , i can wait
Only if you ignore the other functions
Watch wegstr cnc. If you just take the time and have more attention to detail you can get really good looking pcbs. For example, using the centering holes for double sided, using cheese cloth for the solder mask application, using a tinning liquid to tin the copper pads, and if you're feeling really adventurous using multiple solder masks (white, then green) and a laser to etch away the green for circuit legends. I think this hardware is more than capable of making very nice pcbs at home.
11:05 it should be pretty easy to do double sided designs without all that rivet nonsense. I do similar things on a daily basis milling aluminum on a big 3mx2m CNC router. It is a matter of having sacrificial material and cut the negative of the outside perimeter to make a hole to stick the pcb with the other side up. That way you do have a reference to align the PCB for the other side. To makes things way easier, put one of the PCB corners as the 0,0 then put the zero manually so the contour falls inside the sacrificial material. Make the hole undersized on purpose then increase it's size little by little trying the PCB until it fits. Ideally you'd have a table with suction, just so it is a drop and done affair, but you can always cut double sided tape to size for that purpose.
It seems like a lot of companies seem to shoot for the $4000.00 price point when designing these kinds of things (Glowforge). But most "hobbyists" don't really have this kind of "disposable" income to spend on these things.
I have a many years-old machine that I purchased new and is very similar except for the enclosure, dust collection, and ATC. This machine looks like it could be a version 2 of that particular machine, if only the manufacturer of my machine did not go out of business after a rather successful crowdfunding and decent marketing/ sales to schools. My point is that at the current price point it is not likely that this is going to be something you see in every workshop and makerspace. For less money I purchased another, and far more capable CNC, at least in terms of work are (over 6x in x-y and roughly 10x in z) and spindle (1.5kW), although again without enclosure, included dust collection, or ATC. Let's hope for people who purchase this that the company does not have unrealistic sales expectations that sink the company rapidly and leave the purchasers without support, as has happened time and time again with 3D printing and small CNC routers/ milling machines. This final issue is why I now feel that for the hobbyist or very small business market it is far better to get one of the mass-produced import machines that use off-the-shelf components and standard extrusions. They require more tinkering, but they also are extremely simple to repair or upgrade, because they use off-the-shelf components and standard extrusions.
Awesome and thoughtful review
That looks incredible! I wish I had the space for one and the time and imagination to come up with projects for it.
Thank you, I have two questions first when driiling the PCB,There is a metal base under PCB,will not broken as will touch the metal PCB,Second qustion are there protection from laser.Thanks
wow nice video thanks Scott your the best
In terms of PCB milling I'd still in favor of WEGSTR. But this machine is also awesome.
I've been developing CAM software designed primarily for CNC art/signmaking/engravings, and it works fine for PCB milling if you have a good file format that can be loaded in (i.e. images, vectors). It's called PixelCNC. It's still in beta and right now I'm working on a built-in tutorial system. If you have any suggestions for tutorials or need help with anything shoot an email to us.
That machine looks awesome! I’ll bet they resolve those issues with firmware or software updates.
You should try FlatCAM to process the gerber files into GCODE. The program is a bit unstable, but it has a lot of tools that allow you to get your job properly done. I recommend watching a few tutorials first, because the program may not be that intuitive at first use.
JLCPCB wait time for me in Charleston South Carolina in the United States is now over three weeks and I still haven't received the package using the ground shipping. What they don't tell you is that if you want it super fast you're going to pay 30 or $40 more for the shipping. So it's more expensive than they lead you to believe.
Is it possible to control that CNC with Candle? You could try using FlatCAM + Candle for the PCB stuff. It actually works really well.
Looks like a decent enough mini mill; and having an ATC spindle helps set it apart from the typical 6040 style routers. However, for PCBs, I think I’m gonna stick with outfits like JLC, PCBWay, etc…
It'd be great for proving prototypes before sending off for them though.
Fully agree, why spend thousands of dollars and hours of time only to get a much worse product than what can be bought for very little money from a skilled company.
To prepare GRBL to GCODE PCBs I use Flatcam Beta 8.994, it is the absolute best software for that purposes and there are tons of videos on YT about how to use it. (it's a bit tricky to find the beta but not impossible)
the only thing letting your homemade PCBs down is the soldermask application. The sponge roller gives it a weird texture that makes it look cheap
Very good 👍
Thank you 👍
Just to set the record straight, there are a few more downsides to using board shops (and this is not directed at any shop in particular). First though let me point out that they (in my experience) deliver great work and high quality pcbs (and associated services), credit where credit is due.
But (for the small maker with some labor of love designs that you try to market to like-minded people) :
- Shipping cost, import fees, taxes ... All of these are excessive where I live. I have had times (depends on getting lucky with customs) where the PCB's cost me 2 dollars and shipping + fees + taxes cost me 120 Euros. And there have been times where the PCB's cost me 40 dollars and the shipping + fees + taxes cost me 30 Euros. Aside from the obvious 'wtf is with these add-on costs' one has to realize that it is a crapshoot if you actually need to calculate this cost into your end product.
- Protecting your intellectual property, and this is a general concern not specific to any board shop. It might not be a fun thing to consider but if you make a good design that garners some degree of interest or appreciation it will be stolen. Sending out your design to an external party is one of many ways in which this will happen. By no means the only one, but one none the less.
- You pointed this one out yourself, speed, again most board shops do amazing work and fast but shipping will slow down things considerably. If you are making fast iterations on a new design this can be a problem. A system like the Carvera, the Wegstr cnc or other desktop PCB solution is going to help out with that a lot.
On the up side, if you need quantity and quality, you will eventually have to use a board shop of some sort.
Wow, $4000 is a lot of money but this machine still looks like it's a bargain considering all the features you get.
Ok, thanks for the feedback. Interesting opinions in the comment section :-)
Honestly, the biggest issue with PCB's is via's. Using conductive paint and a chemical process is one option. But then again, if you're going chemicals anyways, you might as well use a SLA printer for the UV & solder mask, tinning via chemicals, etc... If you think about it, it almost sounds easier to just buy them for $3 from jlc.
I love these machines, but for pcb I think there is still a lot to improve. And I think there is technology available for that, it just takes a company to take the matter more seriously.
With a laser, you can do something too, but it's not 100%.
I keep using breadbords and when I need something better, I order from JLCPCB.
I just appreciate the hard work you put in for us. Good video concept 👌. I love it.
Thank you so much 😀
It's possible to get a much better finish with the soldermask using a different technique to apply the soldermask paint.
Not to be too salty, but wanted to mention that I supported the kickstarter and have seen the fulfillment of kickstarter orders pushed back at least 3 times. Original delivery date was, iirc, August 2022 (currently end of Jan 23 and they still have something like 400 machines left to ship)
Having worked with copperCAM before i can for sure say it has alot of problems. The only other issue is with the method of applying the mask.Professional pcb manufacturers use a silk screen for the mask and component designators. If you also use screen printing you will get very professional looking pcbs.Other than that its a solid machine but its priced for a small buisness rather than a hobbyist like most of us are.
Do a firmware dump, and then write your own software. Ezzy Peezzy.
Great review !
;)
Since it comes with a laser: Is the lid blocking the laser light or is it basically like those open frame laser cutters when it comes to laser safety?
Awesome intro my dude :)
I think knowing what limitations and benefits professional machines have really puts things into perspective.
For reference.
My single operation, mass production milling machine with a 20 cm by 10 cm (4x9 inch) working surface
Costed $30,000 USD
Weighs 408 kg 900 lbs (without the coolant and oil)
Only cuts one profile or slot at a time
Is much faster, and takes much less attention,
but
takes a *college degree* to set up reliably, all tools must be custom designed and made for the each cut and must be programed in *assembly commands*
Has no sensors at all.
Can be set up and checked in *once per week* when running.
Costs $.10 per hour to run, and about $500 to $2,000 to set up. *PER feature*
Is not CNC obviously, although is computer controlled, it is not numeric.
all tools can be resharpened.
Another professional model I have used that is a mill-drill center using G-Code, made by Fanuc, with about the same table size.
Has 8 tool slots,
Has no ability to sense the table
Takes 3 phase 410 volt.
*is significantly faster moving*
Weighs in at 650 kg, 1450 pounds (without coolant)
Does *NOT* come with software.
Requires professional attention about once per hour.
Costs about $80,000 in bare bones, used models can cost up to $200,000 with all the options and low hour run times.
Costs about $100 to $200 per hour to run.
Take your trade offs, don't complain, unless the cost and performance do not add up.
What is so special about that machine that it requires a COLLEGE DEGREE (instead of just some training/certificate) to set up, and why does it cost up to $2k per feature? Unless you are talking about the CAD/CAM part of it as well, including the actual PCB design, then i can somewhat understand it. And why does that other one cost $100/hour to run just for milling/drilling? Lots of wear parts?
Have you seen carl bugeja's self soldering PCBs that he just posted this week?
I did. Awesome stuff.
Thank you so much for this video!
You are welcome :-)
I use bantam tools cnc for milling pcb's and am able to do double sided PCBs with around the same quality as this. The only downside is they don't have a way to create a height map. Other than that it is very easy to use.
Is height map the same as auto levelling?
Love the automatic tool changer.
This is something I'd love to have, but I can't quite justify yet. I'm still trying to justify my laser cutter/engraver. :)