What’s interesting about this platform is it should be relatively easy to add additional tool heads to make an IDEX even beyond. No additional belts are needed, just extra drivers and a longer x rail to provide parking space for unused tools.
@@BeefIngot it would add less mass than a conventional IDEX, which needs to move the mass of an additional (full size) stepper in y. In any case, if you print multiple parts (in parallel or mirror mode) a lot, a slower-moving IDEX will be much faster than a faster accelerating tool changer. So it's not really one or the other. Both have their place.
@@BeefIngot not everyone is looking for the same printer. IDEX is very useful for print farms, for example. I have an IDEX printer. I don't use it for everything, but when I need multiple copies of the same thing it's amazing, even though it has a much slower motion system than my fast printers. For dual-material prints, it's also very reliable in a way that tool changers have yet to prove they can achieve. There is a market for IDEX, companies like BCN3D, Raise3D, Craftbot, and at the lower-end Flashforge and Snapmaker, produce IDEX printers because there is a demand for them.
And few minutes after he mentions they did not wanted him to do anything crazy before it started printing - I had immediate flashback to that drill scene :D
Definitely not the sort of thing the average user should be expected to do, so it's nice to have them get it out to technically minded people who are capable of helping them identify and solve these issues for them prior to real production.
Tom is great. We tested print for a few days before shipping. We believe the issue was caused by too tight of tolerance on the ring and impact during shipping that locked it up. Thankfully, Tom was able to fix it. We will address this in the production going forward.
I worked at a company back in the early 90's that had a pen plotter that used this tech, and the plotter was from the late 60's to early 70's era. It was about the size of a twin-size bed, and about 5 feet tall - but was outrageously fast.
@@BeefIngot High speed plotting of maps. I worked for a digital cartography company back then. We were easily pumping out 30-50 plots a day and even then, that was "slow" time for the company. That, and you could slow it down for use with actual wet ink and vellum plots, very large sizes almost 5'x7' - but with super precision. Multi-color (it had four pen slots.) If you'd like to see what it looked like, look up the Xynetics 1000. The plotting head was magnetized to the underside of the top, and needed a little 1/3hp air compressor to push it down enough to where it could move freely. Where the paper was, it needed a giant blower motor to suck the paper flat on the bed and keep it in place. What really broke my heart was when we moved offices, the new place didn't have facilities to bring it to (we moved to a 2nd story of an office building instead of a warehouse type place) and we had to get a roll plotter to replace it. We literally dismantled it and sold it for scrap. I think the boss said it was $100k+ when they bought it in the early 1970's.
I would assume you're referring to xynetics--i was field service on those in the day (80’s)--in addition to flat bed plotters, there was a version called “wirematic” that made aircraft wiring harnesses…everything old is new aga…
Looks like RUclips is having a lazy day with processing this one up to 4k60... Guess it'll get there eventually! Just to clarify, I was assured the printer we're looking at in this video is identical to the production units - but as with anything, there will certainly still be changes to it down the road, whether those be for cost-saving or for improving certain aspects of the machine. The software (both the printer's firmware and the slicer) are still obviously a bit rough around the edges, but are much easier to update than hardware - with so few machines in the wild right now, I'm not considering this too much of an issue yet, but I do hope that the printer's tuning will improve over time or with community involvement. Let me know what comparisons, benchmarks or experiments you'd like me to do with the Magneto!
I know some further software updates are coming, should be a new round for the testers to validate this week. Hardware seems pretty solid at this point, it's mostly as you said the software. Luckily, it's klipper, so 98% of that is just dropping in a new config file!
This is bad. I thought this was a pre production unit. I mean, the safety, the bended nozzle. Those are huge problems and you said it yourself, this can't be sold to EU. I expected more from them. As a tech demo it is good but I'm not buying a tech demo.
By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the motor from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't. Surprisingly accurate description of how the closed-loop system works 👍 On the topic of grounding, the entire frame, bed and moving axes are meticulously grounded, so there should be absolutely no risk of the frame itself ever becoming live.
@@MadeWithLayers 😅👍 Since you know "a thing or two" about 3d printers, maybe you should switch into development. Using linear motors for x and y is somehow obvious, still no one has done it. I bet you also have some ideas on what can be improved in persue of the optimal/best fdm printer there can be.
@@MadeWithLayers But Clipper does still don't know where the printhead is. Time to fusion Marlin/Clipper/GRBL with it's source from EMC2/LinuxCNC and go realy full closed loop positioning system including the trajectory control system. This would be the real way to eliminate the ringing artefacts
*This whole section seems to have been copied almost word-for-word from an ancient training film about MISSILE GUIDANCE, which I have seen years ago, here on RUclips:* _The guidance subsystem uses deviations to generate corrective commands to drive the motor from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't._ @@MadeWithLayers
Despite all the quirks - this is the first printer in at least 5 years I would be exited to own. The sturdiness, the platform, the mechanism are an awesome starting point of something that will last for a very long time.
Exactly! I am tired of sketchy-mechanics nothing-new kind of commercial products we have seen in the past 5ish years. Now, this IS THE innovation! Almost perfect. No excuse for the lack of Z endstops and of the lack of mains insulation. For the bed part, I think that at this level, the bed *thermal* insulation should be standard as well. One little nitpick with the mechanics is that the axial bearings should ideally hold the Z axis screws directly, not through the flexible coupling. However, I see that they used an almost rigid coupling, and the bed system is almost a static load, so I can forgive them for that. Overall, as someone who sees the innovation and the improvement opportunities and also likes to fabricate things: I'd buy it!
Hats off to Peopoly for creating this awesome first of its kind printer. Agreed it needs some tuning and some adjustments for the safety concerns but yea, I can see this being a serious contender for best of the best 3D printers! A lot here for Peopoly to be proud of and cant wait to see this on the open market! Even the price while expensive is really not when you consider what went into building it! It for sure in its current state would not pass US safety laws and it really needs a crash mitigation for that hot end. I really cringed when I saw how hard that hit the build plate.
This is something that has finally trickled its way down from Comercial CNC machines. No more belt noise at high speed. I loove this machine and it is on my Must have list!
its sure promising, but at this stage you and anyone else buying one, would be buying it out of an irrational lust for technological advancement and where they believe the quality will be/get to, not for any practical outcome, as any prints ive seen are pretty terrible, like really terrible. Not the benchie, obviously that was pretty good in some areas, so heres hoping its just tuning required.
@@Industry-insider expensive ones. linear motors have been around for years. Anyone presenting this as a peopoly invention is misinformed. All they have done is manufacture them at a scaled back power and a cost the 3d printer market will bare.
I actually got one of the pre-sale units ordered, and that enabled me to get updates when they made changes or if something happened like the delays. The reason I point this out is they seem extremely in tune with their customers and the sample units they sent out (like the one shown here) have been improved multiple times since they were built and shipped out to get some feedback from makers/content creators that deal with lots and lots of 3D printers and other hardware, after all - you guys will be tearing them apart and really finding out what is fantastic, what is good but needs improved, and what is so bad it would prevent sales, as well as have the knowledge on how one might go about fixing the issues. As of lately we got another 'month delay' so I won't be getting mine anytime soon, but it's nice the videos can start coming out to discuss.
Do you know if any of the problems he showed in this video have been ironed out I got a grant to get one for my business but after this video, I’m worried About the quality, a bunch of artifacting the bed ramming itself into the head, exposed mains voltage
I much prefer having a printer that is a little delayed than one that is totally unfinished. IMO, the Magnto hardware is in a good spot, and software is easily upgradable even once the units are already out in the wild.
that is a serious oversight of them, klipper does not allow moving the bed without knowing home first, unlesss you manually enter a Unsafe move macro so either it was a you issue, or they did something that messed this up, and allows unsafe moves... usually the nozzle probe would be used as a Z endstop, and you can only move the bed once the nozzle has verified Z max or min
Moving is indeed only allowed once the printer is homed, but it currently does not prevent you from moving to unsafe coordinates. What happened is that the printer was freshly homed, and instead of clicking "Z +25mm", I clicked "Z -25mm" as my exhausted brain thought I needed to move the bed "down", and the printer obliged. Only that this moved the nozzle further into the bed... Typically this is prevented with software endstops, but it seems like that feature is not currently enabled, and the loadcell as a Z endstop is likely also disabled outside of the actual homing and probing cycles to prevent false triggers e.g. during filament retracts.
@@MadeWithLayersit needs a failsafe to prevent nozzle/bed damage The load cell is there, so this is a coding issue, not a need for extra parts Ideally I'd like to see a position encoder on the vertical axis. Relying on the leadscrews is ok but not as good as actually measuring and linear encoders are fairly cheap
then it sounds like they did indeed misconfigure the printer in klipper, as in klipper you can set the endstop position, and the max z that is allowed to move, incase of a nozzle probe it should both be the same, which would prevent moving past the nozzle probed location@@MadeWithLayers they seriously need to get their configs straight before sending out printers like this, they also need to configure the klipperscreen correctly, as it would have allowed to swap the direction buttons, on my trident i swapped them so that the down button moves the bed down and the up button moves it up
For inkjet printers, it's about saving cost by using dirt-cheap, barely powerful enough brushed motors (that have no positional awareness otherwise). Here, the linear BLDC motor would actually be able to position itself without extra help, but would have immensely coarse resolution. So while the resulting closed-loop system and its error-correcting abilities are a nice side-effect, the encoder really is an integral part to make this system work in the first place.
@@MadeWithLayers This is exactly the reason I am excited. I wonder why 3d printer designers focus on steppers or rotary servos. The cheaper system in Inkjet printers is not affected from belt stretch, motor in-linearity etc. all this is compensated for by the control algorithm while also improving speed, precision and fidelity. TBH I wanted to make a POC printer with linear encoders for some time now. Usually these are simple optical encoders. These look like capacitive or magnetic encoders, interesting.
@@flipschwipp6572Inkjet printers just move the head at a constant speed, which is easy with a DC motor and linear encoder to count movements. A 3D printer needs to be able to accelerate and decelerate and control position to within sub-milimeters, something that would be near impossible without some sort of 'step' drive.
@@frankhovis DC or BLDC servo drives are perfectly fit for positioning and acceleration. They motor may need to be little bigger than in a usual inkjet to satisfy acceleration needs but thats it.
@@MadeWithLayersInkjet printers require a magnitude or two better resolution than a 3D printer, most of them are faster too. However, the task is easier as they have to do just constant speed and fixed trajectory movements. Besides that, if you were to closed-loop control a stepper, you might as well leave out the steps and just use a BLDC servo which is essentially almost the same control model just the number of phases are different.
This is the kind of machine that needs to be reviewed by a man like you! Very interesting review. Thorough as always! I think this (linear stepper motors) will set the new standard for 3D printers in the future. Very surprised to see the very promising long melt chamber. This machine is a very good answer to and even a step forward from the Bambu machines IMO. Just needs a little ironing still... But very well done by Peopoly .. again proving that they still are innovators! One question remains though: Are you sure that the coil shouldn't be placed between the bed and the Z-corner support instead of underneath. Then a crash like the one you had would not result in damage so easily.
Thanks! The bed springs are intended to be installed as I showed, as using them in an Ender-style arrangement would make for a quite wobbly and shaky bed assembly, I think. wiki.peopoly.net/en/magneto/magneto-x/setup-guide
I really hope linear motors start being more accessible. I could totally imagine a future where we just forget about belts completely. Id absolutely take all of the tradeoffs to get that no worry consistent accuracy and low maintenance.
@@Domophilliuswell sort of, except it doesn't have stepper motors and these will probably get way less hot, and no belts to slip it still is better than a bedslinger
Peopoly did say that they designed these ones from the ground up and were open to releasing them as their own thing. If they did, I'd be so up for upgrading my ratrig with these.
@@Domophillius Sure but its never about the name but the function. This isnt offset on the x axis because the y motor is pretty much spread across all of the x axis, and obviously the y axis is stationary. Combine that with closed loop control, a wider area to spread and dissipate heat and strength and It sounds like many plusses to me.
@@BeefIngot exactly the point i said to someone else on another channel especially since the mags wont get anywhere as hot i'm sure it'll be more efficient either way and maintain more "torque" or pulling/stall force rather the hotter the neo mags get the less force that remains and the more current you need
i think that some of quality issues might be caused by some sort of hotend damage that happened when you ran it into the plate, i saw similar issues when my nozzle was damaged (tried cleaning leftover filament with snippers and cut some of metal) and got uneven
To be honest this whole part after the bed hit is so unprofessional on his side that is hard to watch. Clearly the hotend got bend and he still continues to print with it. Whole print is already compromised and the fact that he is still talking about the print quality after that is honestly so unprofessional. He should actually stop after the incident, replace the hotend and make sure that the frame is not compromised in any way before continuing the review.
Glad that it sounds like the majority of the issues are things that can be improved on in a software update rather than a fundamental issue with hardware. Definitely would be good to have some insulation updates, I'll add that to the to do list for when mine arrives.
@@patching- Yep! Most of the complaints about the print quality are just down to basic end-user calibration. Factory calibration is a good starting point but it should always be completed by the end-user after the printer has shipped to them, using their own filament. No printer is going to remain unchanged after shipping and assembly, nor is any printer or filament going to be absolutely identical to the factory ideal. Even the local temperature and humidity and going to change your results. That's something I'd like to see in a long-term review though, this printer looks very promising as it is.
@@ahaveland No! In a perfect world, profanities would not exist, and we’d speak the same language. But humans are strange in a way that we usually default to our mother tongue for expletives.
@@m_IDEX I speak 4 languages, two natively. I can't imagine a worse world than having one language and no profanities, and being able to swear in many more than 4 languages is a great stress relief, but each to their own. Diversity has riches beyond compare.
Thank you so very much for showing the issue with ability to jam the bed into the print nozzle. So many people expect these machines to be plug and play without the chance of damaging the print, machine or themselves. We need more education in this hobby that we are operators of a power tool and WE are responsible for the movements of that tool. Yes, the manufacturer should install safety mechanisms, but we also need to learn the limitations of the printer and understand how to keep calm when it starts operating outside of expected parameters. You stayed calm and solved the problem without further damage. Thank you for sharing that experience. Great video and very informative. On machines that I operated that had linear rails, they were really good at very fast movement over long distances and short stopping distances meaning minimal overshoot, but those machines were never for small detailed movement. It was designed more for that speed than accuracy. I wonder if this is the best technology for 3D printing accuracy as you showed. I would love a followup video if you ever manage to clean up the quality. I would be interesting to see if its possible and the steps you took to get there.
I had been pondering on building a printer with linear motors for about a month before I saw the Magneto. Glad to see someone did it! They are used often in circuit card manufacturing and watching how fast and accurate the are is pretty inspiring.
Four linear rails... mounted straight to what appear to be anodized aluminum extrusions that are part of an assembly. Those guide blocks undoubtedly have a ton of clearance to make them work without binding. The 3d printing community still hasn't figured out how to use linear rails. A 15 ton vertical machining center can sling around a multi-ton chunk of cast iron on two rails with a single ballscrew. The important part is rigid mounting of the rails to flat and parallel surfaces, which permits using guide blocks with preload (or at least zero clearance). If you print out some calibration parts on this and stick them on a CMM, or even do a cursory inspection of them with manual metrology tools, I bet you'd be shocked at how imprecise it really is.
As a preorderer, I wish field testers would give dates of testing and what revisions of hardware and software they’re testing. Testing results seem inconsistent, and it’s tough to know if the reviewer’s process, the hardware and software itself, delays in testing or video production or something else is bringing different issues to light or folks are reporting problems already addressed. Good preview nonetheless, however.
I have Yamaha pick and place machine at work and because of this video i have finally realized how Y axis works for two heads on same axis. I did try to find belts or something else and didnt paying any attention to magnets inside linear axis
Ohhh! I see. Now linear motors make more sense (with two moving heads on the same linear rail). I was struggling to find any positives to this drive system.
Good video Thomas - Thank You. But then the "How can this be improved" part of my brain stood up and started shouting at me. Put the magnets in the linear rail truck and a whole series of coils into the rail and then a lot of wire movement will be saved. This sort of propulsion system should work very well in the 3D PRINTER market but when it comes to driving a 12mm cutter through a piece of steel then it probably will not have the sheer power to be able to take big cuts at higher speeds to do it. For the most part, belt linear drives simply get overwhelmed which is why screws are the go-to for high energy machines. BUT for low energy linear movement applications like 3D printing and laser machines I can see this style of machine motion going far. And the G0 type commands should be able to really shoot the truck to the end point at a very rapid rate. Seeing where and how this technology goes in the next few years should be interesting.
I want to try one with coils in the rails too, with crossed rail gantry so there are no moving motor wires at all. It would get wasteful with energy at large size since you're magnetizing a lot of rail area that isn't doing anything, but it would be super fast moving around permanent magnets rather than the motor stator with its wires and drag chain. The only part I don't know how to do is the linear encoders.
it seems like an impressive gamma prototype. Lots of little refinements that need to be dialed in to make it user friendly and safe. The linear motors seem great, although they seem a bit massive (and heavy) for the small forces they need to generate. It is great to see some real innovation happening in the 3d space.
That's a very very coarse linear motor. And I think the patterns you see are a direct result of that. This could be in theory compensated thru a calibrated motor profile, but the closed loop controllers need to be able to do that. That would mean they need a lookup table for the entire travel length, those controllers don't look like they would have enough RAM for that.
Next thing is, that they get feed from a stepper interface which is suboptimal for CL servos. Next position and an acceleration curve would be much much better. Klipper should be able to give that.
That's a quite impressive machine ! Sure, there are (always) things to improve, more tuning to get predictable and reliable print quality, yet it's awesome to finally see something new and interesting ! Looking at the next iterations of this product line will be extremely instructive.
That does look like a very nice system to move the print head. It sure gives a prototype feel in many ways, I can imagine this getting a lot lighter in the future.
There is a good chance that you kinked the heatbreak with the crash and adding extra resistance to the filament path. That messes with the extrusion quality. I have had this problem myself
I'm so happy that YOU are the one presenting this. Thanks, Tom! I've been wondering about this since the announcement! Hope you're well and can't wait for what's next!
Not sure if it's something weird on my side, but the best quality I can watch this in is 360p, which is definitely not what I expect from Made with Layers :) Am I just too early (video uploaded 4 minutes ago). *EDIT* All good now, lovely 4k
How in the hell did you know about the gantry fix. I've been around 3d printing for about as long as the 1st Dremel 3d printer has been out and I didn't ever see that. Thank you for giving me insight to that issue.
re: the step size vs magnet size - it uses the ratio of the values in different places to accurately locate exactly where it is - 2 sensors separated by half a magnet size would give sin/cos output, so you can easily tell what fraction of a magnet step sideways you are, and you can use more than 2 measurement points to cover for variation on magnet strength
The funky pattern on the outside is definitely due to the linearity of the motors. The issue with linear motors is that the accuracy is dependant upon the calibration of the feedback loop *and* magnetic field. And yes, they can be calibrated. But because each axis relies on a big stack of magnets, the magnetic field pattern won't be totally uniform all across the range of motion. You can actually see the position of the magnets is super uneven when you peel back the cover strip. This means that to achieve good linearity across the entire range of motion, you have to calibrate it across the entire range of motion. This is an insane amount of data, and probably not something they have done. Or possibly they have done this at a coarse resolution, interpolating between points. Leading to this slight loss of linearity. Also, I can't really tell what the feedback mechanism is on this printer. It should be a high accuracy optical encoder, but I fear that it's a magnetic strip tape encoder on the bottom of each axis rail. At least that's what it looks like from the images I have seen. So not only is that subject to magnetic interference, the linearity is lower to begin with.
I like it, mostly because it seems like what I call honest product; it appears that effort was put into making a product that is technically really good and only unimportant corners were cut.
Im not terribly knowledgeable about the minutiae of 3d printers, but in regards to the flow issues, is it not possible that the bending that occurred prior to your custom print damaged the hotends inner tube causing less consistent extrusion?
That's most likely why he had the issues on the last model he printed. Me thinks it would have helped for this bloke to read the instruction manual and become a little familiar with the machine before producing a review video. Honestly he did a huge disservice to both Peopoly and his own brand.
The release of this video will spur Peopoly to do further touch-ups and polishing to the full release machines. A cover for the terminal connections on the back of the primary power input cord socket will be implemented quickly I'm sure as well as upgraded spacers on the x-axis motors and etc. Finding these things is the reason for sending out test models for beta testing.
I always wanted to have linear encoder on 3d printer. I think they allow for precise printing in the long term. If they can be calibrated properly, bents, bumps or shakes will have a lot less impact than on other printer. And they allow to use linear motors that have others adventages. I can't wait to see perfectionment of this technology, so we can have truly reliable 3D printer for everyone.
Could some of those extrusion issues have been caused by the bed crash and the bent hotend? Have you tried another benchy after the bed crash to see if they are the same?
That chunky build is something I can get behind, same with the lower mechanical complexity due to the linear motors. Especially for models with this kind of print volume this seems like a must-have in the foreseeable future.
An impressive first impression of an absolutely unique new 3d printer. I hope the project progresses properly and the price is kept within limits. The videos published of this device look very promising so far.
Technically Essentium HSE series offered linear motor 3D printers for the industrial market for a long while, but it's lovely to see the technology trickle down to a more affordable machine.
It is great to see this technology working in consumer 3d printing! Hopefully this is the start of of this becoming mainstream! 👍 Also, I love the straight to the point review.
Awesome to see PID feedback in consumer 3D printers more and more, this linear setup is pretty sweet! Even the price isn't that bad. I really wish stepper motors hadn't stuck around nearly as long as they have, yeah they work... But imagine how many headaches everyone would have avoided if they just simply had some kind of feedback instead of just counting steps.
on the wifi issuie you saw, if it's using an ESP32 for wifi (which it seemed to be doing) it might only be able to connect to 2.4ghz wifi. I know my mesh network forces wifi to 5ghz... /shrug
the print quality need more deep tuning more than standard open loop system... because closed loop system have their own accleration.. like if you try to hold the movement of x or y axis to make little late, or stop totally.. the extruder still extruding the fillament that machine is going to give.... open loop have absolute precise movement at given coordinate... otherwise it will miss step, but closed loop have their own small little late and little faster (to catch up) the given coordinate.. so the extruder will over extrude (late) and under extrude (when catch up) on some area... on small print like benchy it's not noticeable, but larger print it will have this issue.. i hope soon the closed loop system also communicate to extruder to make accurate flow when late or moving faster, it real time data from closed loop system, not like flow compesation that use prediction based the data coordination...
This is actually a really good point On my (currently only) LM mill, the RPM can change upwards of 5% to adjust for this I've only ever seen it as jitters because of how fast it does it but it does
@@angrydragonslayer for milling? cnc? for cnc closed loop it's good because no material added (fillamnent) adn cnc can use passes as many as needed to remove the part, but for 3d print.. extruder flow should follow the movement of the motor (that closed loop have some late or faster to catchup the designated position by mcu) in high tech cnc, they may also have this calculation.... but i doubt in today 3d printer, because closed loop still rare for 3d print... maybe because linear motor like this and trend of using closed loop, maybe new program could sync the motor movement with extruder... so not only mcu give "command" to motor but motor give command to MCU to compesate the extruder to follow the motor accleration..
@@adimegan4677 yes but the inconsistency still messes a bit with surface finish The adjustment is mostly meant for people pushing Ra with diamond honing inserts that need that perfect tool pressure (which admittedly was me about 8 months ago) and it works wonders at that
@angrydragonslayer ah that's news for me... so closed loop also give some issues to some cnc... although closed loop have been used by CNC fo a long time ago.... Maybe some high end cnc that use servo uses some information of servo to motion control(mcu) to adjust appropriately to make good finish... Still long way to go for closed loop system for 3d printer....
@@adimegan4677 they do but it's utterly negligible for most usecases because of how stable cnc machines are built. You won't really have a problem due to it before 4 μ" and 6-10 μ" is what is generally meant by "mirror finish" I was just barely able to get under the 2 μ" tolerance with the help of that adjustment though.
If that is PETG you should have tuned it first. PETG is notoriously hard to print with default settings. It sounded like the nozzle was dragging on the print 21:51 (either that or it's the magnets). The best way to prevent nozzle drag with PETG is to choose continuous infills where lines seldom cross each other - best infill for PETG is gyroid by far. This is because PETG tends to ooz and goes down heavy at the beginning of an extrusion and light toward the end. In situations where you cross lines, the retraction and extrusion have to be perfectly tuned otherwise you get nozzle dragging and accumulation on the nozzle. The ringing is likely from printing PETG too fast. I would have probably done a PLA example if I were going to showcase a printer without tuning it for the material. That was really bold. It took me a determined week and a lot of research to get my PRUSA mk4 making good quality PETG prints: Temperature tower, speed adjustments, infill changes, observations while printing, cleaning the nozzle, changing to textured print bed, adding a silcone sock on the extruder. PETG takes work on any printer to successfully print large surface items like that (without corners coming up and/or layer shifts) and much more to have something with no artifacts.
Looks like the kind of tech that would be good on high end (no expense spared) pro systems because of all the support tech needed to make it reliable and safe and produce high quality prints, but may take a long time to become practically useful on 'hobby'-cost level printers. Your honest video demonstrates that nicely. Thanks.
I hope you are proving feedback to Peoply, the motor shim and they need to stop the Z when the load sensor is triggered - should not have been the print end.
Reminds me of some of some of the semiconductor equipment EFEMs ive seen with far more complex handler x-axis drives. Its insane the speed and acceleration achieved.
So, the springs for the bed go between the bed and the mounts to the z axis juat in case the bed and extruder crash into each other. It cushions the blow so the hotend and bed take minimal damage.
I'm all for it, less motors and belts is a good idea. Looks amazing for a gen 1. This will become the new standard for design companies will try to replicate.
I think that oscillations is going to need no a little bit but a LOT of tuning. You have there a magnets don't have any kind of dampening like a belt does, and also the closed loop is going to introduce some vibrations and etc etc etc.
I cant get over the fact the magnets are just hanging out there, unrestrained, surely this will mean the field is changing with vibration, so that any tuning will only be temporary? Maybe im overthinking that, but i would have thought setting them into something, or potting them in some kind of high temp silicone would be preferred?
@@jeremyglover5541 No, the magnets are not hanging freely, those are in the head and in the raio, like a train. Are kind of exposed? Yes, but not hanging free
Maybe it's because I have a toddler running around now, but those 230V wiring had me sweating... Even distributing demo devices within the EU already requires CE compliance. There is an exception for importing products in development to contractors and employees and since you fixed that motor design it might even apply here :D
You are exactly right, it could throw an error code to klipper same as it does when it experiences a step missed error and shut down klipper. The feature request has been passed onto the dev team.
This is a really cool printer! I'd love to see a video where you make this printer safer (the easily exposed leads on the bottom, something to better protect the bottom of the bed). Just found your channel a couple days ago and I LOVE your content. Very informative
Reminds me of the pick&place machines at work, same linear motors, they are way bigger than this printer but are quiter than a regular stepper motor, essentialy all you hear is the bushing/bearing sliding.
The feedback sensor for the linear motor might have some deviation. You will see some repeating pattern. And some low frequency deviation. The repeating pattern comes from the interpolation is not 100% accurate. Low frequency deviation ist from that the scale is not 100% accurate. Those errors cannot be mitigated by input shaping. It would need some error correction look up table. You can prove that sensor deviation by printing one test sample slowly and one test sample fast. The artifacts in the print surface should stay the same.
@MadeWithLayers re: janky GCFI hookup. Yes! But it's not an electrical / safety code violation, sadly. First, the engineers who designed and built this were EEs and MEs, but not electricians. 3D printers fall under EU Directive 2006/42/EC which would normally require this, however an exemption is made for "machinery specially designed and constructed for research purposes for temporary use in laboratories" thus a 'notified body' like the UE doesn't validate the design. That was 18 years ago -- 3D printers are no longer being used just by engineers who are prototyping for products that will eventually be used by the public -- these are products for public consumption in their own right now, and this is the sort of problems that come up when legislation lags behind social and technological progress. Sauces: 5:31 Power cable on the table - probably a type E or F (european), 'AT' power supply cable 9:31 exterior plate visible, with 'AT' style plug. 10:12 interior plate visible, with 'quick disconnect' plug. RCD is required for any 230v residential plug at the outlet in Europe. Outside of Europe RCD protection at the outlet is not always guaranteed. When the standard was popularized in the 80s, it was with this assumption in mind. You can't buy power cables with integrated RCD due to this shortcoming, so best practices is to ONLY use this type of hookup inside a sealed unit (ie you need screws to get into it once assembled and powered), or at least to have an access cover with the exterior plate having a 'high voltage' warning noting that the equipment should be unplugged before servicing. However, due to the 'experimental' nature of 3D printers, even these standards can (and are) ignored. The truth is, while this is a design process failure by the engineers, it's also a failure of government regulation, and your comments that this represents an advancement in what is now considered fairly mature technology stand in stark contrast to this legislative reality. As a content creator who focuses on 3D printing, I feel you have a journalistic duty to provide this context to your viewers, not just the technical tear down (which was solid).
hey, really cool video! I just had a question. about 10:03 you mentioned "active PHC" (I think) and I have never heard of that and didn't really find anything in my brief search. What is that? Thanks in advance!
PFC (Power Factor Correction) - for switched-mode power supplies, it makes current as sinusoidal as posible, or at least to draw current not only at the top of the voltage signal, but also symetrical on both sides of the wave.
Super-interesting video. Also I just clicked through to the website and was surprised it was "only" $2000. But I guess that was just a preorder price and will increase in future.
What’s interesting about this platform is it should be relatively easy to add additional tool heads to make an IDEX even beyond. No additional belts are needed, just extra drivers and a longer x rail to provide parking space for unused tools.
Tons of extra moving mass though. If anything Id say this was more ideal for a tool changer where there would be less wores to worry about routing.
@@BeefIngot yeah , it sounds great as a toolchanger
@@BeefIngot it would add less mass than a conventional IDEX, which needs to move the mass of an additional (full size) stepper in y.
In any case, if you print multiple parts (in parallel or mirror mode) a lot, a slower-moving IDEX will be much faster than a faster accelerating tool changer. So it's not really one or the other. Both have their place.
@@JonS I count the ol double corexy as also being conventional.
As for dual printing, I guess, but how often do people actually do that.
@@BeefIngot not everyone is looking for the same printer. IDEX is very useful for print farms, for example. I have an IDEX printer. I don't use it for everything, but when I need multiple copies of the same thing it's amazing, even though it has a much slower motion system than my fast printers. For dual-material prints, it's also very reliable in a way that tool changers have yet to prove they can achieve.
There is a market for IDEX, companies like BCN3D, Raise3D, Craftbot, and at the lower-end Flashforge and Snapmaker, produce IDEX printers because there is a demand for them.
Tom casually fixing a motor design issue, no big deal.
I hope he invoiced them for engineering design support.
And few minutes after he mentions they did not wanted him to do anything crazy before it started printing - I had immediate flashback to that drill scene :D
Definitely not the sort of thing the average user should be expected to do, so it's nice to have them get it out to technically minded people who are capable of helping them identify and solve these issues for them prior to real production.
Tom is great. We tested print for a few days before shipping. We believe the issue was caused by too tight of tolerance on the ring and impact during shipping that locked it up. Thankfully, Tom was able to fix it. We will address this in the production going forward.
legend
I worked at a company back in the early 90's that had a pen plotter that used this tech, and the plotter was from the late 60's to early 70's era. It was about the size of a twin-size bed, and about 5 feet tall - but was outrageously fast.
What in gods name would you need a pen plotter like that for?
@@BeefIngot Signs and CAD drawings, usually.
@@BeefIngot High speed plotting of maps. I worked for a digital cartography company back then. We were easily pumping out 30-50 plots a day and even then, that was "slow" time for the company. That, and you could slow it down for use with actual wet ink and vellum plots, very large sizes almost 5'x7' - but with super precision. Multi-color (it had four pen slots.)
If you'd like to see what it looked like, look up the Xynetics 1000. The plotting head was magnetized to the underside of the top, and needed a little 1/3hp air compressor to push it down enough to where it could move freely. Where the paper was, it needed a giant blower motor to suck the paper flat on the bed and keep it in place.
What really broke my heart was when we moved offices, the new place didn't have facilities to bring it to (we moved to a 2nd story of an office building instead of a warehouse type place) and we had to get a roll plotter to replace it. We literally dismantled it and sold it for scrap. I think the boss said it was $100k+ when they bought it in the early 1970's.
@@BeefIngot Large plans, there are plans for urban development that fill the floor of a large room.
I would assume you're referring to xynetics--i was field service on those in the day (80’s)--in addition to flat bed plotters, there was a version called “wirematic” that made aircraft wiring harnesses…everything old is new aga…
I'm loving the "let's assemble and discover this printer together" vibe with this video. Super cool tech!
Awesome! Thank you!
"Peopoly were very keen on me not trying any weirdness" *starts drilling holes into the stepper motor assembly* x)
As he explained, it couldn't calibrate the bed gantry without doing so, So I mean, can you blame him?
Looks like RUclips is having a lazy day with processing this one up to 4k60... Guess it'll get there eventually!
Just to clarify, I was assured the printer we're looking at in this video is identical to the production units - but as with anything, there will certainly still be changes to it down the road, whether those be for cost-saving or for improving certain aspects of the machine.
The software (both the printer's firmware and the slicer) are still obviously a bit rough around the edges, but are much easier to update than hardware - with so few machines in the wild right now, I'm not considering this too much of an issue yet, but I do hope that the printer's tuning will improve over time or with community involvement.
Let me know what comparisons, benchmarks or experiments you'd like me to do with the Magneto!
I know some further software updates are coming, should be a new round for the testers to validate this week. Hardware seems pretty solid at this point, it's mostly as you said the software. Luckily, it's klipper, so 98% of that is just dropping in a new config file!
Peopoly printer is very promising ❤
RUclips is being extra lazy, it's still only showing 360p for me.
I really want to know how accurate the closed loop absolute encoder system is, or if it even is absolute. Like, does it not need to home?
This is bad. I thought this was a pre production unit. I mean, the safety, the bended nozzle. Those are huge problems and you said it yourself, this can't be sold to EU. I expected more from them. As a tech demo it is good but I'm not buying a tech demo.
Best line in the video: “Sort of squeaking like a very quiet duck” 😂
Yes but ducks don't squeak, they quack.
That probably explains the feather that you found inside the case 😅
@@TheWeepingCorpsedepends on how you squeeze them...
You'd be surprised to learn how well-articulated the average duck is 😉
I have heard ducks quack, and I have heard them hiss, I have never heard one squeak though….
8:27 "it knows where it is, because it knows where it isn't"
Definitely impressive!
Could you see if the print bed frame was connected to earth?
Thought the same thing 😂
By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the motor from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.
Surprisingly accurate description of how the closed-loop system works 👍
On the topic of grounding, the entire frame, bed and moving axes are meticulously grounded, so there should be absolutely no risk of the frame itself ever becoming live.
@@MadeWithLayers 😅👍
Since you know "a thing or two" about 3d printers, maybe you should switch into development.
Using linear motors for x and y is somehow obvious, still no one has done it. I bet you also have some ideas on what can be improved in persue of the optimal/best fdm printer there can be.
@@MadeWithLayers But Clipper does still don't know where the printhead is.
Time to fusion Marlin/Clipper/GRBL with it's source from EMC2/LinuxCNC and go realy full closed loop positioning system including the trajectory control system.
This would be the real way to eliminate the ringing artefacts
*This whole section seems to have been copied almost word-for-word from an ancient training film about MISSILE GUIDANCE, which I have seen years ago, here on RUclips:*
_The guidance subsystem uses deviations to generate corrective commands to drive the motor from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't._
@@MadeWithLayers
Despite all the quirks - this is the first printer in at least 5 years I would be exited to own. The sturdiness, the platform, the mechanism are an awesome starting point of something
that will last for a very long time.
I'm no pro, but this also has me interested. If my wife wouldn't kill me for spending that much, I'd be in.
Exactly! I am tired of sketchy-mechanics nothing-new kind of commercial products we have seen in the past 5ish years.
Now, this IS THE innovation!
Almost perfect. No excuse for the lack of Z endstops and of the lack of mains insulation. For the bed part, I think that at this level, the bed *thermal* insulation should be standard as well.
One little nitpick with the mechanics is that the axial bearings should ideally hold the Z axis screws directly, not through the flexible coupling. However, I see that they used an almost rigid coupling, and the bed system is almost a static load, so I can forgive them for that.
Overall, as someone who sees the innovation and the improvement opportunities and also likes to fabricate things: I'd buy it!
Hats off to Peopoly for creating this awesome first of its kind printer. Agreed it needs some tuning and some adjustments for the safety concerns but yea, I can see this being a serious contender for best of the best 3D printers! A lot here for Peopoly to be proud of and cant wait to see this on the open market! Even the price while expensive is really not when you consider what went into building it! It for sure in its current state would not pass US safety laws and it really needs a crash mitigation for that hot end. I really cringed when I saw how hard that hit the build plate.
its going to give rat rig v core 4 a run for the money in the states i'd guess. the 400mm^3 is the same $ from what i've seen
This is something that has finally trickled its way down from Comercial CNC machines. No more belt noise at high speed. I loove this machine and it is on my Must have list!
its sure promising, but at this stage you and anyone else buying one, would be buying it out of an irrational lust for technological advancement and where they believe the quality will be/get to, not for any practical outcome, as any prints ive seen are pretty terrible, like really terrible. Not the benchie, obviously that was pretty good in some areas, so heres hoping its just tuning required.
What CNC machines? I didn’t know this was actually being used already
@@Industry-insider expensive ones. linear motors have been around for years. Anyone presenting this as a peopoly invention is misinformed. All they have done is manufacture them at a scaled back power and a cost the 3d printer market will bare.
Honestly though, ill take ballscrews on axis before this anyday.
@@jeremyglover5541 let me introduce you to ASML lithography machines, this has been around for years
I actually got one of the pre-sale units ordered, and that enabled me to get updates when they made changes or if something happened like the delays. The reason I point this out is they seem extremely in tune with their customers and the sample units they sent out (like the one shown here) have been improved multiple times since they were built and shipped out to get some feedback from makers/content creators that deal with lots and lots of 3D printers and other hardware, after all - you guys will be tearing them apart and really finding out what is fantastic, what is good but needs improved, and what is so bad it would prevent sales, as well as have the knowledge on how one might go about fixing the issues.
As of lately we got another 'month delay' so I won't be getting mine anytime soon, but it's nice the videos can start coming out to discuss.
This is a good approach, though 100% I think they should have safety features in from the outset regardless
Do you know if any of the problems he showed in this video have been ironed out I got a grant to get one for my business but after this video, I’m worried About the quality, a bunch of artifacting the bed ramming itself into the head, exposed mains voltage
Thank you for your support. We will add an insulation solution to all printers. An update blog is coming this week.
I much prefer having a printer that is a little delayed than one that is totally unfinished. IMO, the Magnto hardware is in a good spot, and software is easily upgradable even once the units are already out in the wild.
Sounds like what Microsoft did with Windows 7 release where they had the customer figure out what did not work!!
that is a serious oversight of them, klipper does not allow moving the bed without knowing home first, unlesss you manually enter a Unsafe move macro
so either it was a you issue, or they did something that messed this up, and allows unsafe moves...
usually the nozzle probe would be used as a Z endstop, and you can only move the bed once the nozzle has verified Z max or min
Moving is indeed only allowed once the printer is homed, but it currently does not prevent you from moving to unsafe coordinates. What happened is that the printer was freshly homed, and instead of clicking "Z +25mm", I clicked "Z -25mm" as my exhausted brain thought I needed to move the bed "down", and the printer obliged. Only that this moved the nozzle further into the bed...
Typically this is prevented with software endstops, but it seems like that feature is not currently enabled, and the loadcell as a Z endstop is likely also disabled outside of the actual homing and probing cycles to prevent false triggers e.g. during filament retracts.
@@MadeWithLayersit needs a failsafe to prevent nozzle/bed damage
The load cell is there, so this is a coding issue, not a need for extra parts
Ideally I'd like to see a position encoder on the vertical axis. Relying on the leadscrews is ok but not as good as actually measuring and linear encoders are fairly cheap
then it sounds like they did indeed misconfigure the printer in klipper, as in klipper you can set the endstop position, and the max z that is allowed to move, incase of a nozzle probe it should both be the same, which would prevent moving past the nozzle probed location@@MadeWithLayers
they seriously need to get their configs straight before sending out printers like this, they also need to configure the klipperscreen correctly, as it would have allowed to swap the direction buttons, on my trident i swapped them so that the down button moves the bed down and the up button moves it up
This is the most ballsy printer design that we have seen in a while.
Essentium has been making these linear motor printers for years. They are for industrial use. Still the technology has been proven by essentium.
I'm very impressed with the robust tool head. A crash like that would have killed many other heads.
Great video. Keep up the good work.
That doesn't mean this one is not killed. :D
It still printed after the crash though.@@crnalajna
Closed loop linear encoder control is used in every inkjet printer. Happy to see one in a 3d printer.
For inkjet printers, it's about saving cost by using dirt-cheap, barely powerful enough brushed motors (that have no positional awareness otherwise). Here, the linear BLDC motor would actually be able to position itself without extra help, but would have immensely coarse resolution. So while the resulting closed-loop system and its error-correcting abilities are a nice side-effect, the encoder really is an integral part to make this system work in the first place.
@@MadeWithLayers This is exactly the reason I am excited. I wonder why 3d printer designers focus on steppers or rotary servos. The cheaper system in Inkjet printers is not affected from belt stretch, motor in-linearity etc. all this is compensated for by the control algorithm while also improving speed, precision and fidelity. TBH I wanted to make a POC printer with linear encoders for some time now. Usually these are simple optical encoders. These look like capacitive or magnetic encoders, interesting.
@@flipschwipp6572Inkjet printers just move the head at a constant speed, which is easy with a DC motor and linear encoder to count movements. A 3D printer needs to be able to accelerate and decelerate and control position to within sub-milimeters, something that would be near impossible without some sort of 'step' drive.
@@frankhovis DC or BLDC servo drives are perfectly fit for positioning and acceleration. They motor may need to be little bigger than in a usual inkjet to satisfy acceleration needs but thats it.
@@MadeWithLayersInkjet printers require a magnitude or two better resolution than a 3D printer, most of them are faster too. However, the task is easier as they have to do just constant speed and fixed trajectory movements.
Besides that, if you were to closed-loop control a stepper, you might as well leave out the steps and just use a BLDC servo which is essentially almost the same control model just the number of phases are different.
This is the kind of machine that needs to be reviewed by a man like you! Very interesting review. Thorough as always! I think this (linear stepper motors) will set the new standard for 3D printers in the future. Very surprised to see the very promising long melt chamber. This machine is a very good answer to and even a step forward from the Bambu machines IMO. Just needs a little ironing still... But very well done by Peopoly .. again proving that they still are innovators!
One question remains though: Are you sure that the coil shouldn't be placed between the bed and the Z-corner support instead of underneath. Then a crash like the one you had would not result in damage so easily.
Thanks! The bed springs are intended to be installed as I showed, as using them in an Ender-style arrangement would make for a quite wobbly and shaky bed assembly, I think.
wiki.peopoly.net/en/magneto/magneto-x/setup-guide
I really hope linear motors start being more accessible. I could totally imagine a future where we just forget about belts completely. Id absolutely take all of the tradeoffs to get that no worry consistent accuracy and low maintenance.
Wouldn't that mean that we wouldn't have CoreXY and back to Cartesian?
@@Domophilliuswell sort of, except it doesn't have stepper motors and these will probably get way less hot, and no belts to slip
it still is better than a bedslinger
Peopoly did say that they designed these ones from the ground up and were open to releasing them as their own thing. If they did, I'd be so up for upgrading my ratrig with these.
@@Domophillius Sure but its never about the name but the function. This isnt offset on the x axis because the y motor is pretty much spread across all of the x axis, and obviously the y axis is stationary. Combine that with closed loop control, a wider area to spread and dissipate heat and strength and It sounds like many plusses to me.
@@BeefIngot exactly the point i said to someone else on another channel
especially since the mags wont get anywhere as hot i'm sure it'll be more efficient either way and maintain more "torque" or pulling/stall force rather
the hotter the neo mags get the less force that remains and the more current you need
i think that some of quality issues might be caused by some sort of hotend damage that happened when you ran it into the plate, i saw similar issues when my nozzle was damaged (tried cleaning leftover filament with snippers and cut some of metal) and got uneven
To be honest this whole part after the bed hit is so unprofessional on his side that is hard to watch. Clearly the hotend got bend and he still continues to print with it. Whole print is already compromised and the fact that he is still talking about the print quality after that is honestly so unprofessional. He should actually stop after the incident, replace the hotend and make sure that the frame is not compromised in any way before continuing the review.
That hotend reminds me something 😊 hehe I love the linear motors here. Very nicely built printer
Hey Si... hmm Vez, long time no see, sorry did not log in with the @toma.3d or @toma.cnc1 for quite some time.
I was thinking the same thing. Although, under the sock, the ceramic heaters are quite different from Goliath.
Yeah, I had the same thought when I saw that
A long rapido? Yeah. Nothing like a goliath 🤔
@xsolar a mixt of both maybe :) anyhow, if it was inspired by goliath that is flattering :)
360p gang! This looks really exciting, will be great to watch it evolve more
It's still 360p 5 minutes later xD
yeah, maybe tomorrow
haha glad its not just me
Lol, chill people. Tom has some of the cleanest looking videos on YT. The resolution issue is on YT, and it'll be full resolution soon.
Locked at 720p
Glad that it sounds like the majority of the issues are things that can be improved on in a software update rather than a fundamental issue with hardware. Definitely would be good to have some insulation updates, I'll add that to the to do list for when mine arrives.
its klipper you don't need an update just to tune it
@@patching- Yep! Most of the complaints about the print quality are just down to basic end-user calibration. Factory calibration is a good starting point but it should always be completed by the end-user after the printer has shipped to them, using their own filament. No printer is going to remain unchanged after shipping and assembly, nor is any printer or filament going to be absolutely identical to the factory ideal. Even the local temperature and humidity and going to change your results.
That's something I'd like to see in a long-term review though, this printer looks very promising as it is.
Linear motors are so cool! But, during the nozzle crash, I did not expect that word in English 😃
Well, shit happens 🤷♂️
In that situation, any and all profanities in every language would be perfectly justifiable!
@@ahaveland No! In a perfect world, profanities would not exist, and we’d speak the same language. But humans are strange in a way that we usually default to our mother tongue for expletives.
@@m_IDEX I speak 4 languages, two natively. I can't imagine a worse world than having one language and no profanities, and being able to swear in many more than 4 languages is a great stress relief, but each to their own.
Diversity has riches beyond compare.
I'm surprised the z max in the klipper config was not set up properly
Thank you so very much for showing the issue with ability to jam the bed into the print nozzle. So many people expect these machines to be plug and play without the chance of damaging the print, machine or themselves. We need more education in this hobby that we are operators of a power tool and WE are responsible for the movements of that tool. Yes, the manufacturer should install safety mechanisms, but we also need to learn the limitations of the printer and understand how to keep calm when it starts operating outside of expected parameters. You stayed calm and solved the problem without further damage. Thank you for sharing that experience.
Great video and very informative. On machines that I operated that had linear rails, they were really good at very fast movement over long distances and short stopping distances meaning minimal overshoot, but those machines were never for small detailed movement. It was designed more for that speed than accuracy. I wonder if this is the best technology for 3D printing accuracy as you showed. I would love a followup video if you ever manage to clean up the quality. I would be interesting to see if its possible and the steps you took to get there.
I had been pondering on building a printer with linear motors for about a month before I saw the Magneto. Glad to see someone did it! They are used often in circuit card manufacturing and watching how fast and accurate the are is pretty inspiring.
That stepper shim modification you did impressed the hell out of me. Great job!
Next time when talking about noise and sounds - let us just hear them.
Four linear rails... mounted straight to what appear to be anodized aluminum extrusions that are part of an assembly. Those guide blocks undoubtedly have a ton of clearance to make them work without binding. The 3d printing community still hasn't figured out how to use linear rails. A 15 ton vertical machining center can sling around a multi-ton chunk of cast iron on two rails with a single ballscrew. The important part is rigid mounting of the rails to flat and parallel surfaces, which permits using guide blocks with preload (or at least zero clearance). If you print out some calibration parts on this and stick them on a CMM, or even do a cursory inspection of them with manual metrology tools, I bet you'd be shocked at how imprecise it really is.
I absolutely loved seeing you be surprised at what you were seeing and discovering the machine live. It was refreshing.
looks really cool i hope the linear motor tech can eventually make its way down to other types of machines too like hobby CNCs and laser cutters.
Thanks for sharing! I have preordered this printer and can confirm that Peopoly has addressed the electrical issues that you showed here.
10:00 seems odd those power supply fans don't line up with the panel holes? $2000?? ouch
As a preorderer, I wish field testers would give dates of testing and what revisions of hardware and software they’re testing. Testing results seem inconsistent, and it’s tough to know if the reviewer’s process, the hardware and software itself, delays in testing or video production or something else is bringing different issues to light or folks are reporting problems already addressed. Good preview nonetheless, however.
This is literally THE PRINTER I wanted and gave up on finding, when I first got into 3D printing. This could get me back into FDM!
I have Yamaha pick and place machine at work and because of this video i have finally realized how Y axis works for two heads on same axis. I did try to find belts or something else and didnt paying any attention to magnets inside linear axis
Ohhh! I see. Now linear motors make more sense (with two moving heads on the same linear rail). I was struggling to find any positives to this drive system.
the first useful sponsor of a vid i've come across, super helpful
Good video Thomas - Thank You.
But then the "How can this be improved" part of my brain stood up and started shouting at me.
Put the magnets in the linear rail truck and a whole series of coils into the rail and then a lot of wire movement will be saved.
This sort of propulsion system should work very well in the 3D PRINTER market but when it comes to driving a 12mm cutter through a piece of steel then it probably will not have the sheer power to be able to take big cuts at higher speeds to do it.
For the most part, belt linear drives simply get overwhelmed which is why screws are the go-to for high energy machines.
BUT for low energy linear movement applications like 3D printing and laser machines I can see this style of machine motion going far.
And the G0 type commands should be able to really shoot the truck to the end point at a very rapid rate.
Seeing where and how this technology goes in the next few years should be interesting.
I want to try one with coils in the rails too, with crossed rail gantry so there are no moving motor wires at all. It would get wasteful with energy at large size since you're magnetizing a lot of rail area that isn't doing anything, but it would be super fast moving around permanent magnets rather than the motor stator with its wires and drag chain. The only part I don't know how to do is the linear encoders.
The X axis still has to move, doesn’t it? So it would just move a handful of smaller wires from head assembly to X gantry
it seems like an impressive gamma prototype. Lots of little refinements that need to be dialed in to make it user friendly and safe. The linear motors seem great, although they seem a bit massive (and heavy) for the small forces they need to generate. It is great to see some real innovation happening in the 3d space.
That's a very very coarse linear motor.
And I think the patterns you see are a direct result of that.
This could be in theory compensated thru a calibrated motor profile, but the closed loop controllers need to be able to do that.
That would mean they need a lookup table for the entire travel length, those controllers don't look like they would have enough RAM for that.
Next thing is, that they get feed from a stepper interface which is suboptimal for CL servos.
Next position and an acceleration curve would be much much better.
Klipper should be able to give that.
That's a quite impressive machine ! Sure, there are (always) things to improve, more tuning to get predictable and reliable print quality, yet it's awesome to finally see something new and interesting ! Looking at the next iterations of this product line will be extremely instructive.
Thank you for overdubing the mistake rather than just putting some text on screen!
That does look like a very nice system to move the print head. It sure gives a prototype feel in many ways, I can imagine this getting a lot lighter in the future.
There is a good chance that you kinked the heatbreak with the crash and adding extra resistance to the filament path. That messes with the extrusion quality. I have had this problem myself
I'm so happy that YOU are the one presenting this. Thanks, Tom! I've been wondering about this since the announcement! Hope you're well and can't wait for what's next!
360p 😢, hope you can get it worked out, interesting that they seemingly overlooked the stepper cross loading.
Your contents are always valuable and amazing.
What's up with the resolution?
RUclips encoding can be a fickle mistress. Give it a few hours and it'll be available at a better resolution.
Not sure if it's something weird on my side, but the best quality I can watch this in is 360p, which is definitely not what I expect from Made with Layers :) Am I just too early (video uploaded 4 minutes ago). *EDIT* All good now, lovely 4k
It's still the same on my end almost 20 minutes later
Thank you Thomas 🤩 ... it feels like the old day TOM with new tech nerdiness. *Way to go*
How in the hell did you know about the gantry fix. I've been around 3d printing for about as long as the 1st Dremel 3d printer has been out and I didn't ever see that. Thank you for giving me insight to that issue.
re: the step size vs magnet size - it uses the ratio of the values in different places to accurately locate exactly where it is - 2 sensors separated by half a magnet size would give sin/cos output, so you can easily tell what fraction of a magnet step sideways you are, and you can use more than 2 measurement points to cover for variation on magnet strength
@19:50, Use 2.4 Ghz wifi. That should work, 5Ghz will not work.
The funky pattern on the outside is definitely due to the linearity of the motors. The issue with linear motors is that the accuracy is dependant upon the calibration of the feedback loop *and* magnetic field. And yes, they can be calibrated. But because each axis relies on a big stack of magnets, the magnetic field pattern won't be totally uniform all across the range of motion. You can actually see the position of the magnets is super uneven when you peel back the cover strip. This means that to achieve good linearity across the entire range of motion, you have to calibrate it across the entire range of motion. This is an insane amount of data, and probably not something they have done. Or possibly they have done this at a coarse resolution, interpolating between points. Leading to this slight loss of linearity.
Also, I can't really tell what the feedback mechanism is on this printer. It should be a high accuracy optical encoder, but I fear that it's a magnetic strip tape encoder on the bottom of each axis rail. At least that's what it looks like from the images I have seen. So not only is that subject to magnetic interference, the linearity is lower to begin with.
This was such a good watch! I love the style!
I like it, mostly because it seems like what I call honest product; it appears that effort was put into making a product that is technically really good and only unimportant corners were cut.
„Slick-three-arr”: so cute, cracks me up every time ❤
Im not terribly knowledgeable about the minutiae of 3d printers, but in regards to the flow issues, is it not possible that the bending that occurred prior to your custom print damaged the hotends inner tube causing less consistent extrusion?
That's most likely why he had the issues on the last model he printed. Me thinks it would have helped for this bloke to read the instruction manual and become a little familiar with the machine before producing a review video. Honestly he did a huge disservice to both Peopoly and his own brand.
hope Peopoly listen to this vid and learn,your the man
Thanks for posting about it after its fully sold out, I watched first day you uploaded -.-
The release of this video will spur Peopoly to do further touch-ups and polishing to the full release machines. A cover for the terminal connections on the back of the primary power input cord socket will be implemented quickly I'm sure as well as upgraded spacers on the x-axis motors and etc. Finding these things is the reason for sending out test models for beta testing.
I always wanted to have linear encoder on 3d printer. I think they allow for precise printing in the long term. If they can be calibrated properly, bents, bumps or shakes will have a lot less impact than on other printer. And they allow to use linear motors that have others adventages.
I can't wait to see perfectionment of this technology, so we can have truly reliable 3D printer for everyone.
Could some of those extrusion issues have been caused by the bed crash and the bent hotend? Have you tried another benchy after the bed crash to see if they are the same?
That chunky build is something I can get behind, same with the lower mechanical complexity due to the linear motors. Especially for models with this kind of print volume this seems like a must-have in the foreseeable future.
@3:10 I'd say it's not 2 heating zones, the ribbed aluminum is likely to cool it quickly so it doesn't burn in side/get jammed after retraction.
An impressive first impression of an absolutely unique new 3d printer.
I hope the project progresses properly and the price is kept within limits.
The videos published of this device look very promising so far.
Technically Essentium HSE series offered linear motor 3D printers for the industrial market for a long while, but it's lovely to see the technology trickle down to a more affordable machine.
It is great to see this technology working in consumer 3d printing! Hopefully this is the start of of this becoming mainstream! 👍
Also, I love the straight to the point review.
Every manufacturer should include a sound system with a beat while bed levelling like @16:20
Awesome to see PID feedback in consumer 3D printers more and more, this linear setup is pretty sweet! Even the price isn't that bad.
I really wish stepper motors hadn't stuck around nearly as long as they have, yeah they work... But imagine how many headaches everyone would have avoided if they just simply had some kind of feedback instead of just counting steps.
I missed what caused the bed crash? Was it firmware?
on the wifi issuie you saw, if it's using an ESP32 for wifi (which it seemed to be doing) it might only be able to connect to 2.4ghz wifi. I know my mesh network forces wifi to 5ghz... /shrug
This is something I hope we see more of in the future!
the print quality need more deep tuning more than standard open loop system... because closed loop system have their own accleration..
like if you try to hold the movement of x or y axis to make little late, or stop totally.. the extruder still extruding the fillament that machine is going to give....
open loop have absolute precise movement at given coordinate... otherwise it will miss step, but closed loop have their own small little late and little faster (to catch up) the given coordinate.. so the extruder will over extrude (late) and under extrude (when catch up) on some area...
on small print like benchy it's not noticeable, but larger print it will have this issue..
i hope soon the closed loop system also communicate to extruder to make accurate flow when late or moving faster, it real time data from closed loop system, not like flow compesation that use prediction based the data coordination...
This is actually a really good point
On my (currently only) LM mill, the RPM can change upwards of 5% to adjust for this
I've only ever seen it as jitters because of how fast it does it but it does
@@angrydragonslayer for milling? cnc?
for cnc closed loop it's good because no material added (fillamnent) adn cnc can use passes as many as needed to remove the part, but for 3d print.. extruder flow should follow the movement of the motor (that closed loop have some late or faster to catchup the designated position by mcu) in high tech cnc, they may also have this calculation.... but i doubt in today 3d printer, because closed loop still rare for 3d print... maybe because linear motor like this and trend of using closed loop, maybe new program could sync the motor movement with extruder... so not only mcu give "command" to motor but motor give command to MCU to compesate the extruder to follow the motor accleration..
@@adimegan4677 yes but the inconsistency still messes a bit with surface finish
The adjustment is mostly meant for people pushing Ra with diamond honing inserts that need that perfect tool pressure (which admittedly was me about 8 months ago) and it works wonders at that
@angrydragonslayer ah that's news for me... so closed loop also give some issues to some cnc... although closed loop have been used by CNC fo a long time ago....
Maybe some high end cnc that use servo uses some information of servo to motion control(mcu) to adjust appropriately to make good finish...
Still long way to go for closed loop system for 3d printer....
@@adimegan4677 they do but it's utterly negligible for most usecases because of how stable cnc machines are built. You won't really have a problem due to it before 4 μ" and 6-10 μ" is what is generally meant by "mirror finish"
I was just barely able to get under the 2 μ" tolerance with the help of that adjustment though.
"squeaking like a quiet duck" Not words I ever thought I would hear in a 4d printer review.
If that is PETG you should have tuned it first. PETG is notoriously hard to print with default settings. It sounded like the nozzle was dragging on the print 21:51 (either that or it's the magnets). The best way to prevent nozzle drag with PETG is to choose continuous infills where lines seldom cross each other - best infill for PETG is gyroid by far. This is because PETG tends to ooz and goes down heavy at the beginning of an extrusion and light toward the end. In situations where you cross lines, the retraction and extrusion have to be perfectly tuned otherwise you get nozzle dragging and accumulation on the nozzle. The ringing is likely from printing PETG too fast. I would have probably done a PLA example if I were going to showcase a printer without tuning it for the material. That was really bold. It took me a determined week and a lot of research to get my PRUSA mk4 making good quality PETG prints: Temperature tower, speed adjustments, infill changes, observations while printing, cleaning the nozzle, changing to textured print bed, adding a silcone sock on the extruder. PETG takes work on any printer to successfully print large surface items like that (without corners coming up and/or layer shifts) and much more to have something with no artifacts.
Looks like the kind of tech that would be good on high end (no expense spared) pro systems because of all the support tech needed to make it reliable and safe and produce high quality prints, but may take a long time to become practically useful on 'hobby'-cost level printers. Your honest video demonstrates that nicely. Thanks.
I hope you are proving feedback to Peoply, the motor shim and they need to stop the Z when the load sensor is triggered - should not have been the print end.
Reminds me of some of some of the semiconductor equipment EFEMs ive seen with far more complex handler x-axis drives. Its insane the speed and acceleration achieved.
So, the springs for the bed go between the bed and the mounts to the z axis juat in case the bed and extruder crash into each other. It cushions the blow so the hotend and bed take minimal damage.
Thank you Thomas, nice review, great problem finding,great solutions and no drama.
I'm all for it, less motors and belts is a good idea. Looks amazing for a gen 1. This will become the new standard for design companies will try to replicate.
I think that oscillations is going to need no a little bit but a LOT of tuning. You have there a magnets don't have any kind of dampening like a belt does, and also the closed loop is going to introduce some vibrations and etc etc etc.
I hope they can fix it. That looks just like 2mm spaced belt VFA that most current stepper/belt printers currently suffer from at certain speeds.
Keep in mind he rammed the hotend into the build plate before running a single print. That will absolutely throw off any real printing results
@@xsolar cant imagine it introducing any artefact like that though ...
I cant get over the fact the magnets are just hanging out there, unrestrained, surely this will mean the field is changing with vibration, so that any tuning will only be temporary? Maybe im overthinking that, but i would have thought setting them into something, or potting them in some kind of high temp silicone would be preferred?
@@jeremyglover5541 No, the magnets are not hanging freely, those are in the head and in the raio, like a train. Are kind of exposed? Yes, but not hanging free
Maybe it's because I have a toddler running around now, but those 230V wiring had me sweating... Even distributing demo devices within the EU already requires CE compliance.
There is an exception for importing products in development to contractors and employees and since you fixed that motor design it might even apply here :D
Does it mean, that this kind of 3d printer can be MUCH larger than belt ones?
You know, it makes me wonder - there's a *load cell* in the print head, why couldn't they have it stop the crushing?
You are exactly right, it could throw an error code to klipper same as it does when it experiences a step missed error and shut down klipper. The feature request has been passed onto the dev team.
This is a really cool printer! I'd love to see a video where you make this printer safer (the easily exposed leads on the bottom, something to better protect the bottom of the bed).
Just found your channel a couple days ago and I LOVE your content. Very informative
Squeaking like a duck that must have been where the feather came from 🤔
I see HUGE potential with this. There's a couple things they need to fix, but overall? It's a REALLY good start.
Great review Thomas. Always look forward to your take on new technology.
Reminds me of the pick&place machines at work, same linear motors, they are way bigger than this printer but are quiter than a regular stepper motor, essentialy all you hear is the bushing/bearing sliding.
Linear motor actuators are the future of 3D printing. I always wondered why they weren't used.
The feedback sensor for the linear motor might have some deviation. You will see some repeating pattern. And some low frequency deviation. The repeating pattern comes from the interpolation is not 100% accurate. Low frequency deviation ist from that the scale is not 100% accurate. Those errors cannot be mitigated by input shaping. It would need some error correction look up table.
You can prove that sensor deviation by printing one test sample slowly and one test sample fast. The artifacts in the print surface should stay the same.
awesome printer, can't wait to see this software polished. Going to produce some insane results in a few years
I was wondering WHEN are they going to start using them ! FINALLY ! ! ! YES
@MadeWithLayers re: janky GCFI hookup. Yes! But it's not an electrical / safety code violation, sadly.
First, the engineers who designed and built this were EEs and MEs, but not electricians. 3D printers fall under EU Directive 2006/42/EC which would normally require this, however an exemption is made for "machinery specially designed and constructed for research purposes for temporary use in laboratories" thus a 'notified body' like the UE doesn't validate the design. That was 18 years ago -- 3D printers are no longer being used just by engineers who are prototyping for products that will eventually be used by the public -- these are products for public consumption in their own right now, and this is the sort of problems that come up when legislation lags behind social and technological progress.
Sauces:
5:31 Power cable on the table - probably a type E or F (european), 'AT' power supply cable
9:31 exterior plate visible, with 'AT' style plug.
10:12 interior plate visible, with 'quick disconnect' plug.
RCD is required for any 230v residential plug at the outlet in Europe. Outside of Europe RCD protection at the outlet is not always guaranteed. When the standard was popularized in the 80s, it was with this assumption in mind. You can't buy power cables with integrated RCD due to this shortcoming, so best practices is to ONLY use this type of hookup inside a sealed unit (ie you need screws to get into it once assembled and powered), or at least to have an access cover with the exterior plate having a 'high voltage' warning noting that the equipment should be unplugged before servicing. However, due to the 'experimental' nature of 3D printers, even these standards can (and are) ignored.
The truth is, while this is a design process failure by the engineers, it's also a failure of government regulation, and your comments that this represents an advancement in what is now considered fairly mature technology stand in stark contrast to this legislative reality. As a content creator who focuses on 3D printing, I feel you have a journalistic duty to provide this context to your viewers, not just the technical tear down (which was solid).
Indeed a wonderful Tech 3D Printer. As it is their first FDM 3D Printer with some amazing motion system. You have some great humor!
hey, really cool video! I just had a question. about 10:03 you mentioned "active PHC" (I think) and I have never heard of that and didn't really find anything in my brief search. What is that? Thanks in advance!
PFC (Power Factor Correction) - for switched-mode power supplies, it makes current as sinusoidal as posible, or at least to draw current not only at the top of the voltage signal, but also symetrical on both sides of the wave.
Super-interesting video. Also I just clicked through to the website and was surprised it was "only" $2000. But I guess that was just a preorder price and will increase in future.
It's not like video games where the preorder is a reserve price. Usually in this industry paying the pre-order will get you a machine.
@@BrianVanderbuschI'm not saying people who paid the preorder price won't get a machine; I'm saying that I expect the price to rise subsequently.
The pre-order price was 1499. The final product price is 1999.
@@parallelsandtangents938Interesting, thanks. If it stays at that price and they fix the safety issues etc I'll be tempted
8:25 -"The printer knows where it is, because it knows where it isn't...." :D
So glad to see you this exited again! 💪 Keep it up! 🎉