Features for a "Perfect" entry level printer?

I'm just curious as to what people's idea would be of a "perfect" entry level printer in terms of cost and features. Entry level here meaning people with no experience, no history with 3d printing that don't really know if they want to fully dive into the hobby, or kids wanting to enter the hobby but have no prior experience.

Price is a big one I'm keeping in mind, even a couple of hundred bucks for the cheapest chinese kits is a lot of money to some people. Especially if they don't know whether they will keep going with the hobby, or don't have their own source of income and rely on their parents.

Feature wise I feel bed leveling is a must, it saves a ton of hassle and headache for someone who isn't experienced with troubleshooting. Thing's like LCDs and SD card readers (from experience) aren't crucial on a first printer. It's a feature people only miss once they've tried it in my opinion. One controversial point I'd like to bring up is that I don't think entry level (again, look at my definition of entry level above) printers should have a heated bed. The extra danger from getting burned and the power/wiring associated with them doesn't seem appropriate for the target demographic. On the other hand, it could be argued that a printer with a heated bed, wifi and all the bells and whistles is the better way to go for entry level, but that bumps the cost up to make it more easy to use.

I like the concept behind the 101HERO, kodama obsidian and the STARTT. The 101HERO made a bunch of mistakes (that i'm hoping to avoid), the obsidian has gone up in price since the kickstarter and I think the STARTT will be going up in price as well soon since they can't really make much profit selling it that cheap and that's not sustainable. I really like what these printers are trying to do by making things cheap and accessible and want to try my hand at it too.

Quote
Trakyan
I'm just curious as to what people's idea would be of a "perfect" entry level printer in terms of cost and features. Entry level here meaning people with no experience, no history with 3d printing that don't really know if they want to fully dive into the hobby, or kids wanting to enter the hobby but have no prior experience.

Price is a big one I'm keeping in mind, even a couple of hundred bucks for the cheapest chinese kits is a lot of money to some people. Especially if they don't know whether they will keep going with the hobby, or don't have their own source of income and rely on their parents.

Feature wise I feel bed leveling is a must, it saves a ton of hassle and headache for someone who isn't experienced with troubleshooting. Thing's like LCDs and SD card readers (from experience) aren't crucial on a first printer. It's a feature people only miss once they've tried it in my opinion. One controversial point I'd like to bring up is that I don't think entry level (again, look at my definition of entry level above) printers should have a heated bed. The extra danger from getting burned and the power/wiring associated with them doesn't seem appropriate for the target demographic. On the other hand, it could be argued that a printer with a heated bed, wifi and all the bells and whistles is the better way to go for entry level, but that bumps the cost up to make it more easy to use.

I like the concept behind the 101HERO, kodama obsidian and the STARTT. The 101HERO made a bunch of mistakes (that i'm hoping to avoid), the obsidian has gone up in price since the kickstarter and I think the STARTT will be going up in price as well soon since they can't really make much profit selling it that cheap and that's not sustainable. I really like what these printers are trying to do by making things cheap and accessible and want to try my hand at it too.

It depends what you want from a printer. do you want reliable printing, or a machine to tinker with?

If you want reliability, it doesn't exist below £500. Stop dreaming about cheap chinese kits and spend the money on a prusa, lulzbot, ultimaker etc.

option 2: I like to tinker.

Get a machine that is easy to take apart, and is adaptable. Anything extrusion based is usually easy to change up, make your own parts etc. The D-bot is good for this. This option is certainly available below £500 initially, but you will make up the cost over time. All chinese kits need upgrading without fail.
Get a heated bed. They aren't dangerous - Your hotend will go hotter and is much more likely to burn you or cause a fire. If you have any reservations about safety -- GOOGLE it. The same problems that could cause a heated bed to go wrong, are the same as anything else on the printer. (bad wiring). Heat beds are pretty much essential for todays printers, why go backwards?

Is bed levelling a must? As a new user, it's actually something that needs yet more calibrating/research to get it working. Stick to a perfectly flat surface and levelling is easy. Cast tooling plate or milled tooling plate is not that expensive.

As for those brands, they made mistakes because they used cheap substitutes that a hobbyist would never use. They made their decisions to make them in bulk, that a hobbyist would not care about, ie, £2 for a toy motor or £4 for a real nema17.

Edited 2 time(s). Last edit at 11/20/2017 11:32AM by Origamib.

I wasn't really asking for printer recommendations on what to buy, and there are chinese kits that work pretty reliably out of the box, monoprice, wanhao and so on. I was looking for answers to this question not from a hobbyist, one off perspective, but as if you were to manufacture a printer, that means making those decisions (in which case the toy motors cost cents and nemas still cost upwards of 4 dollars, plus the cost they add to shipping). The choices a mass produced printer needs to make are vastly different to those a custom hobby machine does.

As for the heated beds, they draw a lot more power than hotends, so they are a far bigger risk. Their mosfets and connectors are also often under rated and they are much easier to accidentally touch than a hotend (bigger, more exposed, likely something youll need to touch to remove a print).

As for the leveling point, that really depends. With cheaper sensors which vary more from spec, or mounts that arent super percise in their positioning, there is a lot to calibrate with nozzle offsets and so on. But systems that use the nozzle as a probe (ala lulzbot, the layer one atom delta, ultimaker) don't require the user to calibrate anything, so no learning curve there. Sensors like on the prusa are another story, you need to adjust xy offsets and z height since the probe is seperate to the nozzle, definately can complicate things for new users.

"Thing's like LCDs and SD card readers (from experience) aren't crucial on a first printer."
I disagree with that, think it is well worth having an controller and printing off SD it avoids those pc mishaps when serial comms fails into a ten hour print , and for what they cost these days no excuse not having one

"The extra danger from getting burned and the power/wiring associated with them doesn't seem appropriate for the target demographic."
just dont read/hear off folks getting burned by poking around the main board while printing, with a good design the board want be on show ,so that stupid people can poke at !

heat beds a must else you limiting the user experience with filament choice than there those cold morning coming! gonna leave a newbie filled with frustration with first layer failing! lifting and curling "first timers should avoid using PID and bit bang the bed."

swap out mechanical endstop for a proxi sensor on Z axis and IR for the XY, do away with axis banging into the frame

"Feature wise I feel bed leveling is a must,"
Na a well made machine with a flat surface want need that feature and a 3 or 4 point thumb screws will do just fine

Price is a big one I'm keeping in mind, even a couple of hundred bucks for the cheapest chinese kits is a lot of money to some people"
aint that why reprap exist , cant afford it in one go then save and build over a couple months, amazing how much we read in that time.

Strong & Stable(where have i heard that before), Stays Level, few frills & clutter(except the huge clamp), modular, hope to refine further(remove clamp, move lead to same side as motor), reduce part count, i could save a few quid & make it better in the process. The search for Perfection continues, but can't be far away.

Edited 4 time(s). Last edit at 11/22/2017 08:35AM by MechaBits.

Quote
Trakyan
I wasn't really asking for printer recommendations on what to buy, and there are chinese kits that work pretty reliably out of the box, monoprice, wanhao and so on. I was looking for answers to this question not from a hobbyist, one off perspective, but as if you were to manufacture a printer, that means making those decisions (in which case the toy motors cost cents and nemas still cost upwards of 4 dollars, plus the cost they add to shipping). The choices a mass produced printer needs to make are vastly different to those a custom hobby machine does.

As for the heated beds, they draw a lot more power than hotends, so they are a far bigger risk. Their mosfets and connectors are also often under rated and they are much easier to accidentally touch than a hotend (bigger, more exposed, likely something youll need to touch to remove a print).

As for the levelling point, that really depends. With cheaper sensors which vary more from spec, or mounts that arent super percise in their positioning, there is a lot to calibrate with nozzle offsets and so on. But systems that use the nozzle as a probe (ala lulzbot, the layer one atom delta, ultimaker) don't require the user to calibrate anything, so no learning curve there. Sensors like on the prusa are another story, you need to adjust xy offsets and z height since the probe is seperate to the nozzle, definately can complicate things for new users.

I know you're not asking for a printer -- but the 101hero is not a reliable machine and its prints are rubbish as a result. This is all directly linked to cheap parts. Good printers are hard to get at a good price... With 2D printers, they are sold at a loss and the manufacturer makes the money back via overpriced ink. In the 3D world we don't like chipped filament cartridges, so there is a limit to how cheap a manufacturer can go. To be honest I think printer manufacturers are starting to realise this, and many are producing their own filaments that they highly recommend for their own printers. And now we get back to the original question, are your customers after reliability or a machine to tinker with? If the latter, a cheap kit is fine. If not, and you have a legitimate end use for your parts then you want reliability. I talk about it like a line, but its really a scale. Manufacturers that have access to injection moulding techniques can usually cut the cost down and keep reliability, but then they have to have an expectation to sell x number of printers. This is bigwig territory -- The Kodak, dremel and polaroid printers use a high number of injection moulded parts.

Your other points --- Your heated bed is not safe because of mosfets and connectors... then get bigger mosfets and connectors? If you are buying in bulk the difference is negligible and is not the breaking point of a printer. I really do think the heated bed is a must if you want reliability. Bed adhesion is an extremely important factor for reliability.

My point about sensors is that you still have to invest in hardware, they arn't free. At their most basic they are a cheap micro switch (that doesn't work very well), but at the upper end they cost £10-15. Why not buy a flat bed to begin with? The combined cost of pcb+glass+sensor will likely be very similar to heater + cast aluminium. Perhaps when you make your sensor in bulk it becomes more economical, but it is still an extra system to go wrong. I use my sensor only to level my bed manually as the duetwifi provides an easy to use graphical interface. In the end I would like to make this even simpler by having a graphical interface that tells me how many times to turn each levelling screw.

So, on a scale of cheap to reliable you have the 101hero on one end, and the Ultimaker 3 on the other. One is a joy to use -- Its software and hardware integrate well and the customer can fairly intuitively start a print with no experience and get something that looks like their CAD file. The other is cheap, has a higher learning curve and has a higher chance of problems occurring due to non flat beds, no heated bed, cheap motors that struggle to deal with the load placed on them and electronics that may be a hazard. The middle ground is probably the Prusa mk2 with good software compensation and integration and costs around £500. Prusa sells a ridiculous number of printers, and even he hasn't succumbed to those horrible toy motors. I don't agree with all the decisions prusa has made with his printer, But I have to admit the printer is a great place to start for new users.

I guess I've phrased what I'm asking badly. Given a low budget ($200 tops), something someone who either doesn't have much income or isn't sure if they want to commit to the hobby might want to spend. What features would you include and leave out? Or price aside, what features would be present in the MINIMUM viable product for a beginner have. This does not mean linear guides, 32 bit board with full color touch screen and wifi, servo motors and a temperature regulated enclosed build chamber for ultimate reliability and features.

Do you live in China Trakyan? as they are already doing it, cheap.
for $200 if it moves & squirts plastic and does it well what more do you need.

People who are new to the hobby need reliability... They have started with an idea of what they want to print, and if they can't print it the hobby will suck and they'll move on. If the machine doesn't work as they thought it would, £200 is too much. Any decision that impacts reliability needs to be seriously thought about. So, to me, heated beds and proper motors are a must. Linear rails and 32 bit electronics are not a necessity, but 32bit does allow for some very neat software, and this allows for a more intuitive UI. Setting up a duet is much nicer then settings up ramps, and it's safer in general. User experience is important for new products.

The 3d printer needs to be as reliable as the washing machine in your house for widespread usage to be a success. We arn't at that stage yet.

It needs to work out-of-the-box. The customer must have their first successful print (not calibration items) in their hand within 24 hours of opening the box. Or, if it's a kit, within 24 hours of the kit being fully assembled. So, very clear instructions on how to calibrate & set up the printer. It probably also means providing slicing software pre-configured with the right settings for their printer. Do not expect the new owner to print a required item like a fan shroud. Probably also provide a large-as-possible library of known-good objects to print; I'm thinking of a website or HTML file which has a 100-1000 images of guaranteed-printable-on-their-machine objects, each one graded by print difficulty, with an associated slicer configuration (support, perimeters, layer height, etc), and linked to the correct item in Thingiverse. Include some G-code examples too, so they don't even need to figure out how to slice. Include calibration items like a temperature tower. Throw in an entire kilogram (not 100g) of known-good PLA filament. Throw in everything they need to do their first print(s); blue tape, rubbing alcohol, rod lube, glue stick, etc.

+10 on the reliability.

I don't think you need a heated bed. Over the past several years, I've printed ABS only a couple of times (just to make sure I could); everything else has been PLA. And for PLA, a heated bed is no advantage.

Regarding automatic bed-levelling; I haven't used it in years. For me, it causes more problems than it solves (PrintrBot with capacitative sensor, Delta with mechanical switch on nozzle). The actual process of manually levelling the bed isn't all that difficult, and once done, it stays good for a long time. OTOH, I'm still looking for a good automatic levelling mechanism.

I think that it's much more important to have a truly flat (as well as level) bed, and these are rare in cheap printers. Actually, for me the main benefit of a glass bed is it's flatness; I only need a heated bed so my print will stick to the glass.

To get reliability, you really need an extruder which is very tolerant of high back-pressure from the hot-end, so that occasional nozzle-too-low or heat-creep occurrences don't kill the print; it will resume extruding when it can. Thus, uneven beds become less of an issue.

I'd leave out the LCD and SD card and just expect to run the printer directly from Repetier Host or Octoprint or similar.

Wifi does seem to me to be a useful simplifying thing, although an entry-level printer would probably start its life next to the PC it's being controlled from. If you do go Wifi, make sure the whole printing process (including slicing) can be run from a phone.

A lot of those ideas you mentioned do not make for a cheap printer, and I don't think are what define a minimum viable product. As for Chinese printers, they are cheap but again, I'm not asking where to find a cheap printer that "works" but what is the minimum hardware feature set for a printer. I find it odd how people are saying auto leveling shouldnt be included but a lot of other features for ease of use/reliability are apparently super important. How many people first getting into the hobby know how to properly level a bed or that they even need to do it after putting the printer together? If you want ease of use for a newcomer then auto level seems like a given. I would like to make the distinction that good auto bed level is good, bad auto level causes problems. Capacitive sensors and even inductive sensors are usually meant to be presence not distance detectors, so they're kind of iffy to use. Capacitive sensors in particular have a lot of drift and are affected by a lot of things.

If you want cheapest viable product, just copy China? They've already done it a hundred times.

The real issue here though, is that the product model doesn't work for general users who have no experience. This is the real reason we don't have a printer in every home - Chinese printers are rubbish. Every compromise you make is a compromise on the user experience, and that is fundamental to getting people into 3D printing.

Quote
Trakyan
........................................................ I find it odd how people are saying auto leveling shouldnt be included but a lot of other features for ease of use/reliability are apparently super important............................................

Most common (2D) printers go through a switch on checks and also before each page. Even if somebody thinks that users should be well disciplined and printers should be super rigid something that at least checks for gunk stuck between the heater and the build stage etc is mighty helpful. This is a big plus point for automatic bed checking - which of course will mean ABL as well.

Quote

............................................................... I would like to make the distinction that good auto bed level is good, bad auto level causes problems. Capacitive sensors and even inductive sensors are usually meant to be presence not distance detectors, so they're kind of iffy to use. Capacitive sensors in particular have a lot of drift and are affected by a lot of things.

I am still waiting for anybody who has tried bed leveling with piezo discs to give any negative feedback:

• Accurate
• Cheap
• Robust

etc., etc., etc..

Mike

What Mike said and .....

(you thought you were going to get a short answer haha)

I think you need to look at two scenarios:

1) Cheap, reliable printer for enthusiasts/hobbyists.
2) Cheap reliable mass market consumer printer.

These are completely different items.

Take 1 First - Hobbyist cheap (reliable?) printer.

If you are going to take up 3D printing as a hobby, you are not going to be able to do so without some sort of either knowledge of microcontrollers, robotics/motion systems, an ability to connect motors, sensors and wire things up correctly etc or at least a desire to learn these skills, and the patience to acquire them, which we all know is difficult and frustrating in the early stages where the learning curve is practically vertical, although not quite thankfully. I bought a second hand I3 rework for £150 2.5 years ago, knew nothing about 3D printers, learned what I needed to learn and have since gone on to start a 3D printing business, built one of the most accurate delta printers possible and bring accurate z probes to the reprap community. As you can tell, I really enjoyed it, and continue to do so.

A $200 printer is possible, buy all chinese components, modify and improve them (cloned hotends and chinese linear rails can be brought up to good standard with a little elbow grease and know-how), get a deal on 4 motors, buy a cloned arduino/mega combo, and heated bed, and build your $200 printer, salvage parts from other equipment. But you won't manage that without learning stuff, improvising and inventing solutions as you go along. If you weren't doing all these things then you are not a hobbyist/enthusiast, your're a consumer...

Which leads me on to 2 - cheap consumer 3D printer.

So ditch the stepper motors, and move to using DC motors with linear encoders as used in cheap and readily available 2D paper printers. These do not require endstops, they can offer closed loop control of position of the axis. They are cheap, not readily available but thats not a problem if you want to order several thousands.

Now use a one-piece frame, of some type, Tiko delta was a good starting point, how about a PVC tube, then its enclosed and has a frame?
Now you've got endstops, motors, motion control, you need a heated bed but if you go for a small printer say 100x100x100 volume then levelling isnt really that hard manually, and it stays level, and it's easier to get it flat.

Use some sort of intergrated controller like prusa do with mini rambo. Make enough units and its cost will be quite small.

How about some mechanism to autocalibrate like duet least squares? I'd say use a piezo probe on the hotend or under the bed, as they're cheap and I kinda like them, for calibration and maybe axis orthogonality compensation, but as I said above it might not need to autolevel, but then if you have the probe why not?.

How about a mechanism to auto calibrate the extruder steps so you can use whatever filament you want, or a combined filament monitor, presence detector and thickness measuring device?

There are sort-of consumer printers out there, but like paper printers, the hardware is more expensive than the sale price, and the money is recouped on the consumables. We don't really want to have to use the manufacturer's filament in cartridges, do we? Even though it means it can be tightly controlled for diameter or have this information encoded into the cartridge to adjust the extrusion multipler for slightly different filaments, this is a model to actively resist even if it makes more business sense, there are so many innovative filaments out there to try,

Then there is the issue of demand. Hobbyists love the machines to tinker with upgrade, and compare and all the stuff we do here on the forum, and pros love the machines to prototype stuff, maybe even do production, but your average Joe or Jane, probably isn't that interested in a noisy, possibly smelly and potentially dangerous machine that can make stuff he/she can buy on eBay or Ali for mere pennies. They certainly aren't going to start designing objects to print, if they did they would be dangerously close to being a hobbyist or pro, two groups which are currently rather well catered for.

Edited 1 time(s). Last edit at 11/24/2017 04:25AM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

Agree that good bed sensing and auto-levelling would be great, and probably a must for a consumer 3D printer. I just haven't seen one yet.

Really, in terms of development we're in the 1970s in terms of PCs. Back then, IBM would sell you a genuine PC/XT for $10K, or you could buy a Chinese knock-off for $1500. Colour was still somewhere in the future. No-one had developed a GUI, so everyone had to know DOS commands. And (in the case of my first computer) how to adjust the floppy drive rotation speed. And the ins and outs of RS-232 connectors. Being able to solder was a big plus. Steve Wosniak was giving away the design of the Apple I. It was a world inhabited by enthusiasts who could see that this was going to be great, if not see exactly *how* it was going to be great.

In the same way, you can buy a Stratasys for $20K nowadays, and colour is kindof becoming available, and everyone still needs to know arcane stuff like how to unclog a jammed hot-end and add LED lights to your printer and wire up a heated bed.

Good luck with inventing Windows for 3D-printing. When you're successful, can you please make me VP in charge of something or other, with stock options?

@DjDemonD
I wasn't expecting a short answer, I actually want to thank you for the long one, it was a good and informative read. Out of curiosity, is precision piezo your company's name, the product or both? Also curious to know how to got started up, even if it is a bit off topic.

I agree on a lot of those points, that all of the extra hardware to make things reliable would be great for newcomers, filament sensing, servo control and so on. The dilemma I have with this is that the people you're calling Joes and Janes. They aren't super interested in 3D printing, at least not to begin with, so the much larger price tag of including all these features will probably deter them. I think small, cheap, fairly reliable and easy to use printers are what's needed here. They can probably deal with the occasional print failing or some blobbing or stringing. Things like the monoprice printers. They're one of the more reputable companies from china who make cheap printers with a lot of features (bed level, full color displays, wifi etc.).

I think the hobbyist market is still missing something like the GUS Simpson, a printer that is super hackable with nothing but itself and some basic hardware like bearings, nuts and bolts. Some people don't have easy access to parts like aluminium extrusions, mills or even drill presses. For some people their only tool is a 3d printer and the printer is just that, a hobby to tinker and play with rather than the focus being on the parts it produces. The GUS Simpson has some issues which has held it back, but I'm working on solving those.

@frankvdh

I thin you're right, but we're maybe a bit further along. There are some advances in user friendliness with full color touchscreens, wifi integrations and good slicer profiles (from the bigger companies). Hardware wise Prusa has done some interesting things. I wouldn't agree with all their choices but they have done some things that I think other printers should pick up on. However, I don't think something like unjaming a hotend will ever go away. The hotends will always end up jamming for some reason or another, maybe less frequently, and people will always need to know how to do it. For instance, computers still freeze and crash occasionally, or get slow, fragmented or get viruses. Sure, you can take it to the tech guy but these are issues I think that are comparable to a jammed hotend.

The printer I'm working on is targeted to new users and hobbyists. It's a GUS Derivative so its main advantages are cost and mod-ability. Bar the basic hardware (motors, hotend, controller etc.) and some bolts, nuts and (radial) bearings the whole things is 3d printed so just about every part of the printer can be modified with nothing but the printer itself. You could make a larger version of the printer with nothing but the original hardware and new printed parts. I'm also working on making it simple mechanically, and having good, clear documentation and assembly instructions which will be a plus for beginners and hobbyists alike. I'd like to implement things like power loss recovery, missed step recover, jam detection and so on as well as a bed and steps/mm auto calibraton. Other than filament sensors (I think filament diameter shouldn't be much of an issue, and if it is I think better filament is the way to go, rather than more sensors. Jam and presence detection are both very good ideas though) and bed leveling, what else do you think is a good idea to have?

Precision Piezo is the company name, the products are Piezo20 sensor, and Universal Piezo Sensor kit (currently). Feel free to see our website its in my sig, but I'm not in this thread to spam.

I agree with your sentiment and your aims here, so please don't get me wrong when I say the issue with:

"I think small, cheap, fairly reliable and easy to use printers are what's needed here."

You can quite easily do small. Small means less issue with rigidity, levelling, flatness and lower cost etc..

But to do cheap, reliable and easy to use. Something has to give.

I watched a talk once by a former England Rugby player turned business coach who outlined the principle that in business you ideally want to be able to do low cost, convenient and high quality. But until the advent of online shopping with super efficient warehousing i.e. Amazon, no one had ever managed to achieve it, one of the three things always had to give. Think about any business and you can quickly decide which one of the three they dropped and thenhopefully they focussed on perfecting the other two and achieved success.

Same applies here, you can have Cheap, and Easy to use, but probably not reliable, or Reliable and easy to use, but definitely not cheap. You get the picture. Can you think of a way to "Amazon" the problem so you can get all three?

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I'd like a Strut/piston/column type Z(thats keyed so cant rotate), strong one not powered by air but maybe an internal screw, that you could attach an arm to and maybe that could be telescopic with print head on the end, and a bed, which maybe uses same mech as Z or another option i've already done.

Anybody paying $200 or less for a 3d printer is not very well informed. They will hate it if they lack DIY skills. And if they do have DIY skill they will upgrade the printer until it works which will increase the price they could have paid and gotten quality parts and pieces in the first place. Printing on a piece of glass with PLA only doesn't seem like a good hobby to me.

@DjDemonD

Thanks for the info, I think I'll have a look at them. I was planning to use microswitches but didn't quite like how the assembly would work out and the slight bit of slack they would introduce (can't fully tighten down the assembly otherwise you'd permanently trigger the microswitch).
Also, you're right, you can't really do all three, that was blind optimism on my part. I guess my main goal is cheap, since I'm not someone with a lot of money and I know that if the lowest cost 3D printer was $1000 or even $500, I couldn't have afforded to get into the hobby. I also know that price is a big deal, look at how many "cheapest 3d printers of..." lists you see, compared to "most easy to use/most reliable 3d printers of...". I am inevitably sacrificing on the other two for this, though.

Reliable to me means it works first time, every time. This printer will probably not be that unfortunately. It will be a hobbyist level printer which will work most of the time, enough so that anyone (especially new users) would be happy to use it and not get super frustrated with having two prints failing for every success. Low bar, I know, but given the nature of the hardware (mostly 3d printed), I can iterate easily and improve on this. I'd also spend a fair bit of time tuning and tweaking slicer settings which I feel plays a big part in reliability. You could have an extremely well built machine, but with a bad slicer configuration it will suck.

Easy to use to me had two different approaches, one being more 'easy to use' than the other. The first and 'ideal' is that the machine would come fully assembled, just plug and play. The machine I'm designing is a GUS Simpson derivative so it doesn't really lend itself well to shipping assembled I don't think. The second is having good documentation, so anyone who does have to assemble or tweak it has the least possible trouble doing it. For kits you can also work on minimizing parts and simplifying assemblies (which I have done) to make it easier to use. The 'ideal' of course changes if someone enjoys building and putting things together, but I assume most people would prefer a plug and play printer (at least new people would). Having pre made, well tuned slicer profiles also helps make things easy to use once you get to the printing stage.

For a company, because to them it is a significant cost, they may cut on things like documentation or well tuned slicer profiles. This is what really hurts the ease of use and reliability of some cheaper printers. Doing this costs me nothing but time, and I'm happy to spend that time to document my printer properly. Slicer profiles are something I'll be dialing in anyway, so why not share them? These would drive up the costs for a company, but not for a hobbyist/individual who is devoting their time to this.

@Mechabits
A screw powered piston? Wouldn't that basically be the leadscrews most people currently use?

@cwaa
I paid around that for my first printer. I was ill informed but in other regards. I chose a printer that was closed source (couldn't tweak firmware or anything, including to make use of the included but disabled sd card slot), and didn't lend itself much to modding due to being exclusively injection molded (bar linear rods, etc.). A printer with more printed parts, aluminium extrusions that can easily be mounted to and so on would have suited me personally better. I'm also not a particular fan of the i3 design now and wish I got something else. For some people the upgrading is part of the fun, and why they get the printer in the first place. A largely 3d printed printer caters perfectly to these people I think, and I don't imagine anyone would bother replacing the bolts or radial bearings in the printer unless they completely break (unlikely?). As for printing in PLA only, I do that anyway. My printer, like many people, sits in a living space and the smell from ABS is unpleasant to be around so I print in PLA quite happily. That being said, I'd would make upgrading to a heated bed as easy as possible (i.e. drop in and wire the bed, done).

Screw powered piston...yeah lead or ball I dont care(though would like something telescopic so unless the lead sinks through table or does something else), its the column & carriage also that needs figuring, there must be a number of ways to do it (but i've not sat & done any CAD on the problem) but i've not seen it done, maybe for good reason? But could make for a simple printer. I'd also settle for a wall based track that could be room height.

[en.wikipedia.org]

Edited 2 time(s). Last edit at 11/24/2017 07:24PM by MechaBits.

To be fair I'd say Prusa has probably cracked the reliable and easy to use with the Mk3, but not cheap - yet... Perhaps if they increased production volumes to really big numbers, injection moulded parts instead of print-farming them, they might get the cost down to £399 from £699... That would be the right level, but that's the kit, we'd have to talking about a ready-made machine. Perhaps they could have a frame made for them one-piece, in the same way the monocoque chassis made the car cheaper.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I think there are still (expensive) improvements to be made in reliability, like servo control and for ease of use I think some printers like the craftbot have prusa and the others beat in terms of easy to use. LCD displays with click wheels aren't hard to use, but a full color touch screen is much easier and more intuitive for people who have never touched a 3d printer. They have made some (in my opinion) questionable choices with their printers. I was never a big fan of the threaded rod frame, and I think auto squaring was a solution to a problem that shouldn't have been there to begin with, and the switch to extrusions was a good move. I like some of the features they've introduced with the trinamic drivers and power loss recovery. My biggest qualm about the prusa is the motion system itself, I prefer stationary bed or z bed printers.

Quote
Trakyan
I think there are still (expensive) improvements to be made in reliability, like servo control and for ease of use I think some printers like the craftbot have prusa and the others beat in terms of easy to use. LCD displays with click wheels aren't hard to use, but a full color touch screen is much easier and more intuitive for people who have never touched a 3d printer. They have made some (in my opinion) questionable choices with their printers. I was never a big fan of the threaded rod frame, and I think auto squaring was a solution to a problem that shouldn't have been there to begin with, and the switch to extrusions was a good move. I like some of the features they've introduced with the trinamic drivers and power loss recovery. My biggest qualm about the prusa is the motion system itself, I prefer stationary bed or z bed printers.

The simple frame of the prusa is one of its defining qualities, and the reason it is cheap. Even in mass production cuboid frames require multiple pieces and processes to assemble.

Quote
Trakyan
I think there are still (expensive) improvements to be made in reliability, like servo control and for ease of use I think some printers like the craftbot have prusa and the others beat in terms of easy to use. LCD displays with click wheels aren't hard to use, but a full color touch screen is much easier and more intuitive for people who have never touched a 3d printer. They have made some (in my opinion) questionable choices with their printers. I was never a big fan of the threaded rod frame, and I think auto squaring was a solution to a problem that shouldn't have been there to begin with, and the switch to extrusions was a good move. I like some of the features they've introduced with the trinamic drivers and power loss recovery. My biggest qualm about the prusa is the motion system itself, I prefer stationary bed or z bed printers.

Well I mentioned servo control only as a way to make a cheaper printer using cheap motors rather than steppers which are expensive by comparison. Servo/closed loop for 3D printing isn't really that useful if you're out of position your print is ruined so being able to get back into position doesn't fix it.

I agree that y moving beds don't seem to be the best solution, I prefer deltas and corexys for those reasons too. But the original prusa has sold thousands so they are onto something. Make the mk3 half the price and they will possibly entice some consumers to buy them not just enthusiasts/hobbyists. I think there is a lot to be said for delta's as a consumer printer the Tiko approach was promising, and with a one piece frame, good probes and clever firmware there is practically no issues with complex calibration 8-16 probes around the bed and you're good to go.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

If you can detect the loss of steps the print isn't lost, there will be a blemish but not a lost print. The prusa is doing this to detect layer shifts and recover. I personally don't like the idea of brushed servos, and I think a lot of industries are moving away from them for noise/dust issues. Another thing with these servos is they are usually high speed, geared down in order to allow precise movement. The gear trains add backlash, and from what I've seen there hasn't been much luck with backlash compensation in slicers so far. What I was meaning was adding an encoder strip or something to the movement axis to detect missed steps or collisions or something. With servos, brushed servos can be cheap, brushless ones not so much from what I've seen, and I think brushless would be the way to go.

The one piece idea of the Tiko is interesting, but I don't really see how it helps. You'd be limited to small printers because you'd need to injection mount it in one piece. Also with like shrinkage and warping as the plastic cools in injection molding, I don't actually see it being an improvement over a cnc cut base and top connected with some extrusions other than the lack of assembly. For a small scale printer like the Tiko maybe it makes sense, at least from a mass production point of view. Being a smaller printer also helps with calibration. But I feel they made too many compromises, with things like using 28-byj-48 stepper motors and plastic rails. The steppers were a flat out bad call, and with the price they were charging they could certainly afford to upgrade things to nema 17s or at least 14s (I think nema 17s are cheaper at this point from what I've seen when browsing aliexpress). The plastic rails are questionable, I imagine they'd develop a lot of slop fairly quickly. The Monoprice mini delta is an interesting printer to look at, I think they succeeded where Tiko failed.

my suggestion would look a little like this
[www.ebay.co.uk]

How about making a small delta printer using a piece of 200mm diameter Perspex tube as the body, with three MGN5 or MGN7 linear rails attached to the inside, and a round Perspex top sheet? Three Nema 14 or Nema 11 stepper motors would be mounted on the underside of the top sheet, with the electronics either below or above the top sheet, and a filament spool holder at the very top. The bed would be a piece of Perspex or aluminium with a circular groove milled in it for the body to sit in. Endstop switches could be omitted if stall detection is used, and a mini version of the Duet3D Smart Effector or a piezo-based nozzle contact sensor used for auto calibration.

Large diameter Perspex tube is quite expensive if purchased in small quantities on eBay, but should cost a lot less in production quantities.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

I'm looking for a suitable pvc tube now, drainpipe type stuff for now as you said transparent large diameter tubing is very expensive. But I think I'd leave the base open and not fit a bed as such, that way the printer could be placed onto whatever surface you wanted to print on and then auto calibrate it.

It would also be nice to cad where you wanted holes and have the precision drilled on some sort of cylindrical cnc. Failing that just paper print a template to improve the drilling precision.

Edited 2 time(s). Last edit at 11/28/2017 03:18PM by DjDemonD.

---

Simon Khoury

Co-founder of [www.precisionpiezo.co.uk] Accurate, repeatable, versatile Z-Probes
Published:Inventions

I wasn't meaning that you should attach the bed to the bottom of the tube, just that the tube could rest in a groove on the bed to prevent it slipping if the head hits a bump on the print. This arrangement would also pave the way for a heated bed upgrade.

That said, I've been amazed at how well my SCARA printer prints on to blue tape stuck to the desk.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].