Tinkerine Studio - Ditto 3D Printer

Hi everyone,

I'm here to announce that we have offically debuted our company's (Tinkerine Studio) flagship 3D printer at Vancouver's Mini Maker Faire 2012 this past weekend. We are a newly founded company with background in industrial and interaction design. However, I personally have been in the reprap community for about a year and a bit now.

Ditto came out from some limitations and inspirations of open source machines with an emphasis on addressing the usability and friendliness of current open source printers. We are proud to be able to provide the community with another printer and hope to create and help the community grow on both the mechanical and software side (UI design). What we plan on doing is releasing the files to the community on our website once we have finalized all the BOM and build orders. With that said, give us a little bit of time before that is all ready as we prep to launch our machine both on our website at www.tinkerines.com and on indiegogo.

The build space for Ditto is the following:
X: 190mm
Y: 180mm
Z: 220mm

With the open design, we are still able to get quality prints of 0.2mm layer height at a speed of 60mm/s. This isn't the limit of the machine and we feel it can definitely go faster and if wanted, go even further on layer height resolution down to about 0.1mm. I'll be posting more information regarding the specs tomorrow on the printer.

Since we just finished the Vancouver Mini Maker Faire I'll leave everyone with a few photos of both our machine and some of the others at the 3D Printer Village.

Ditto Beta- Pine by ECSuyu, on Flickr


Ditto Beta- Acrylic by ECSuyu, on Flickr


Darwin! by ECSuyu, on Flickr


BFB 3D Touch by ECSuyu, on Flickr


emaker huxely by ECSuyu, on Flickr


Tantillus by ECSuyu, on Flickr


Johns Prusa and Thing-o-matic by ECSuyu, on Flickr

Edited 1 time(s). Last edit at 06/25/2012 01:17PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

The acrylic one looks slick!

---

--
-Nudel
Blog with RepRap Comic

Looks like a nice machine, I'm glad to see that extruder attached to the head whereas others might be tempted by a bowden cable. It does however look direct drive which is a bit concerning as the lightweight XY head should lend itself to great speeds which may be limited by a direct drive setup.

Best of luck, looks like it has great potential!

@Nudel, Thanks! Everyone at the show said the same. My personal preference is the wood look but I can see why everyone says that about the acrylic version.

@yydoctt,Thanks for the feed back already! This is what I love about this community. A possible geared extruder is something that our beta team will be testing out, we skipped on the bowden system but might look into a geared system as it shouldnt be hard to add one to the setup and add the speed that some are looking for.

I'll be back on with more information in a bit, as we are currently cleaning up the mess in our office from the prep for makerfaire vancouver

Edited 1 time(s). Last edit at 06/25/2012 01:30PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

I like the fact that the front is open, it seems to allow for the possibility of making parts bigger than the print envelope.

Other than that I am having trouble seeing how this is different from ultimaker. The gantry is the same, the z-axis is the same..

Do you think that mounting the extruder on the head doesn't undermine the idea of the moving gantry? The point of that is to minimize moving weight but you just increased it. Is the moving weight still less than mendel?

What have you done about belt tensioning on the ultimaker gantry? That is a huge pain in the ass.

destroyer2012 Wrote:
-------------------------------------------------------
> I like the fact that the front is open, it seems
> to allow for the possibility of making parts
> bigger than the print envelope.
>
> Other than that I am having trouble seeing how
> this is different from ultimaker. The gantry is
> the same, the z-axis is the same..
>
> Do you think that mounting the extruder on the
> head doesn't undermine the idea of the moving
> gantry? The point of that is to minimize moving
> weight but you just increased it. Is the moving
> weight still less than mendel?
>
> What have you done about belt tensioning on the
> ultimaker gantry? That is a huge pain in the ass.


Thanks for the feedback and questions destroyer2012,

Our mechanics is indeed very similar to ultimakers but we believe we have improved on it enough to perhaps have made some contribution, on top of that we believe we have addressed what we believe will improve end user experience. That was one of our main design philosophy going into this, to create a product that is clean and uncluttered to the end user.

The main changes one would first notice is majority of the motors except for the extruder is now out of sight and tucked away. We believe that end users dont need to see what is driving the mechanics but only need to know that once the machine is good to go, that it prints. In order to properly address the weight issue of the direct drive extruder and to center the hotend so its not offset, dual rails were used for the floating system. A whole new carriage was designed around it, still with the mind set to keep the weight as light as possible even with a motor attached on top. Brand new sliding (delrin bushings for durability and characteristics) and belt tension system was designed to minimize weight but keep rigidity, but at the some time to be able to tension the belts as easily as possible, but what we have found is some of our testers have some trouble with it while others were perfectly fine.

For the ease of use of a direct or geared drive extruder we think having a moving mass on top of the carriage isnt too much of a problem. Yes it limits the maximum speed that the machine can run at, but at the same time, it gives the users an ease of calibration and use when they get it in their hands. Unlike with a bowden system's oozing issue it is much easier for a user to grasp the problem and be able to address it if we weren't around to aid them. Also we didnt leave out those that would want to try out the bowden system, at their own time and expense, the machine is designed in such a way that it wouldn't be too hard to add one if someone really wanted to.

We have never seen or touched an ultimaker in real life so its hard to say what we have done differently other than what pictures on the web has to show. I would have thought ultimaker would have addressed the issues if there was something wrong but thats not for me to judge.

I hope this has answered most of what you have asked. If there is more feel free to question.

-Eugene

Edited 1 time(s). Last edit at 06/27/2012 12:44AM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

I can agree that large moving mass is not a huge obstacle to reaching average performance (Heck, have you seen how slow the replicator is? And it prints pretty well) and direct drive is simpler in design and more established in the field. That said, I think a bowden drive, if done right, has a lot of potential. I have recently (not out of the box, mind) been able to get ooze-free prints at a pretty good speed from my ultimaker. So I now don't think (I was skeptical at first) that the bowden drive has any inherent disadvantages. So assuming you make a good filament feeder (with a large diameter hobbed wheel for better grip) and proper fastening for the bowden tube (screw it on, forget pressure fittings) you could get print quality equivalent to a direct drive system at a fraction of the moving mass.

I'm desperate for hints here, do you mind sharing at least some small details about your belt tensioning system? Are you using an idler wheel, or working an adjustable spring somewhere in the design? I think the lack of a good tensioner is a major drawback of the Ultimaker gantry, other than that the gantry design is very sleek (and by extension so is yours).

You used dual rails! To me that seems like an excessive amount of bearings. Eight points where linear bearings/ bushings must be used! Compare to ultimaker's six and mendel's four. This decision decreases your possible print area as well, since you are forced to have rods on all four sides of your nozzle instead of just two. What advantage does putting the nozzle in the middle give you that you are ready to add so much more complexity (and weight and cost) to the design? Just thinking about it now, it actually seems easier to make a direct drive extruder work with single rails and an offset nozzle, since the extruder stepper itself has to be offset from the path of the filament, and as a result the heavy motor would end up directly above the place where rods cross. In the ultimaker the total offset of the nozzle from each rod is approximately 1cm, which doesn't seem like all that much to me.

Hiding the motors is a good idea (watch out for those exposed gears on the tantillus! Hazard city heh), but should you be concerned about proper cooling? I guess the pictures you posted don't really show the back (and really there is a good opportunity for cooling there since the motors are relatively close) but at least on my ultimaker, the motors get pretty hot during normal operation and I've recently had to stick them outside the "chamber" just to make sure they get proper air flow.

Interested to hear more about this printer.

We agree that the bowden design, if tweaked right, can be just as good as a direct drive or geared extruder. Just like you had mentioned, at first you were skeptical, but with enough time invested in tweaking and making sure the retract and speed settings are right it can work wonders. In our humble opinion, the disadvantage of the bowden system is the fact that it does require that extra bit of tweaking in order to have it working reliably. In fact, we have tested the bowden system on Ditto, and even though we got it working, it was more time-consuming than other systems we've tried. This reason alone was more than enough for us to move away from it. That said, we wouldn't mind if anyone feels like tackling the bowden system and trying it on Ditto.

We decided on the dual rails for a few reasons, the majority of it was to keep the weight of a direct drive or geared drive system balanced. We didn't want to take the chance of having a wobbling carriage due to the shift in momentum from quick stops or jerks, especially in small movements. At the same time it addressed the "issue" of the offset print; to us. it only makes sense to have it in the center.

We aren't too concerned about the motors overheating. Heat rises so we simply planned for that by having vent holes under and above the motors. From our tests and being at Maker Faire Vancouver printing nonstop for 2 days, the motors became warm but were still comfortable to the touch. We are currently doing a few prints ranging from 6-17 hours per piece and the motors are doing just fine.

As for our belt tensioning system, there are no springs, just a simple clamp system. I'll be posting photos of it later today, so keep an eye out for that.

Edited 1 time(s). Last edit at 06/28/2012 08:54PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

Here are some photos of how the belt tensioning and bushing systems work. We did a couple prototype before determining on which length and style bushings to go for.

V1 on the left, V2 on the right

Slider Prototype v1 v2 by Tinkerine Studio, on Flickr

V3 on the left and V4 on the right

Slider Prototype v3 v4 by Tinkerine Studio, on Flickr
how the belt tensioning system works.

Slider + Belt by Tinkerine Studio, on Flickr

We are currently testing out Ditto with a geared extruder (gregs extruder) and we are liking the way it works! If it all goes well we will change Dittos extruder system from direct drive to a geared drive. Even with the geared drives added weight we are able to print perfectly good objects (golvends) at full scale with a speed of 80mm/s and get a perfect finish. So once I have more updates I'll post them up, meanwhile any comments or questions are always welcomed

Edited 1 time(s). Last edit at 07/05/2012 06:17PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

I don't understand how the belts are tensioned using that system. I can see they are clamped but I hope you can understand that clamping still allows the belts to stretch and thus backlash will slowly return even if it is nonexistant when the printer is first assembled. Most 2d printers use some kind of spring or leaf to actually apply tension to their belts/cables which eliminates the problems causes by stretch or loosening due to vibration. This lack of decent tensioning schemes is what is holding many "consumer 3d printers" back.

Yes you can loosen that clamp and pull the belt taut again but with nothing to maintain the correct tension the root of the problem, i.e., belts too loose or too tight, still remains unsolved at least from what I can see in that picture.

It is better than the ultimaker method though, so I'll give you props for that.

You are right it is a clamp system and we do see that it will stretch or perhaps slip overtime. It's something that we are looking into as a possible design improvement to the machine, to improve both the reliability and the durability of the mechanical components.

We have looked into some consumer grade 2D printers such as Epson and other brands for inspiration. However, in order to keep the gantry system the way it is, it won't be easy to add a spring loaded system to it without some major overhaul to the overall design. Nonetheless it's something we do keep in mind, and if needed, changes will have to be made.

edit: Even though we believe that there will be some stretching and slippage of the belt over time, it is by no means a design that is lacking in its ability to produce quality prints. We have been stress testing Ditto as much as we can at the studio and have just finished a print that took a total of 90 hours or so, with each piece taking 6-10 hours to print. The belts held up fine in that situation and we believe it will continue to do so. Having had a clamp based belt tensioning system (Prusa) for close to 1.5 years now with no re-adjustment to the belt tension. We believe that though this design has a flaw it will hold up that tension for quite some time.

destroyer2012 Wrote:
-------------------------------------------------------
> I don't understand how the belts are tensioned
> using that system. I can see they are clamped but
> I hope you can understand that clamping still
> allows the belts to stretch and thus backlash will
> slowly return even if it is nonexistant when the
> printer is first assembled. Most 2d printers use
> some kind of spring or leaf to actually apply
> tension to their belts/cables which eliminates the
> problems causes by stretch or loosening due to
> vibration. This lack of decent tensioning schemes
> is what is holding many "consumer 3d printers"
> back.
>
> Yes you can loosen that clamp and pull the belt
> taut again but with nothing to maintain the
> correct tension the root of the problem, i.e.,
> belts too loose or too tight, still remains
> unsolved at least from what I can see in that
> picture.
>
> It is better than the ultimaker method though, so
> I'll give you props for that.

Edited 1 time(s). Last edit at 07/05/2012 11:35PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

Quote

This lack of decent tensioning schemes is what is holding many "consumer 3d printers back.

I don't find belts stretch over time. Once they have settled in they run for years without adjustment. They just need a means of tensioning them and a secure fixing. A screw adjustment makes setting the tension easier than just having a clamp but I don't think springs are needed.

Edited 1 time(s). Last edit at 07/06/2012 07:08AM by nophead.

---

[www.hydraraptor.blogspot.com]

I find zipties a very efficient way to tension belts : cheap, easy.
The only drawback is you have to cut the old and use a new one if you want to lower the tension, but that's rarely the case.

sample usage :

---

Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.

@DeuxVis, we all know zip-ties too well, they work, but feel very hacked together. One of the core design values for our company is to try and stream line the the end user process. So we believe that anything that feels hacked should be address if possible, thus zip-ties were one of the first to be out the door for a mechanical tensioner. They do however work very well for cable management purposes like they were made for.

I'll be posting photos of our direct drive extruder test, to our surprise the gregs extruder fits perfectly on top of the carriage with zero modifications or the need for an additional mounting plate.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

As promised here are some of our testing with the gregs extruder, perfect fit!


Greg's Extruder Test by Tinkerine Studio, on Flickr


Greg's Extruder Test by Tinkerine Studio, on Flickr


Greg's Extruder Test by Tinkerine Studio, on Flickr

---

___________________________________________________
Kreativ-e
Tinkerine Studio

Ok, I don't know where I've been, but this is the first time I've seen a cam-lok on the pressure plate...... I LIKE that!
Not too proud to "steal" that idea.......Thanx. Oh and I agree with nixing the wire ties I've done some crazy things with wire-ties including holding motorcycles together to get out of the hills but accurate cnc? As an old retrofitter it just sends creepies up my spine. Your machine looks too nice for wire-ties. Good work guys.
Todd

To get the most out of your machine you might want to consider a lighter option than a Nema 17 extruder (direct or geared).

[www.thingiverse.com]

The motor in this 'thing' looks like it has great potential for bridging the gap between heavy extruders and oozing bowden cables. According to the description you could have almost three of these complete extruders (with hot-ends) for one Wade's extruder, I am yet to see one of these extruders in action though.

Get some videos of 'Ditto' up!

@tcoleman, thanks for the compliment and feel free to build upon what we have done, after all this is what this community is all about!

@yydoctt, thanks for the suggestion! We will look into that, we would love to lower the weight as much as possible as we are now looking towards a geared drive instead of direct drive, so that motor looks promising.

Give me until early next week to get a video of Ditto printing for everyone. meanwhile we do have our Maker Faire Vancouver video up on youtube. [youtu.be] which does have some actions of Ditto printing. It is a bit limited in the action footage but perhaps a start.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

Heres a short clip of Ditto in action printing a GT2 pulley that all the machines will be coming with for anyone interested. Enjoy!

[youtu.be]

and some end result photos

GT2 Pulley by Tinkerine Studio, on Flickr


GT2 Pulley by Tinkerine Studio, on Flickr

---

___________________________________________________
Kreativ-e
Tinkerine Studio

As we get closer to the launch of our Indiegogo campaign, I'd like to share with everyone one of the latest big change to come to the Ditto machine. The extruder system was initially a direct drive. Though it was good at speeds up to 50mm/s, we couldn't use it with faster prints as we tried to push the prints up to 120mm/s (thanks yydoctt for bringing this to our attention). The lack of printing speed made us go back to the drawing board. And from some quick tests with the Greg's extruder, it was clear that the change needed to be done.

Below is our geared extruder V1 with a ratio of 53:17 without sacrificing build space. It has a self tensioning guided channel for 3mm filament at the moment. The extruder currently weighs about 210 grams (without NEMA17), which for us is a little heavy but we have tested prints with perfect results up to 100mm/s print speed. We have yet to attempt 120mm/s.


Gear Extruder v1 by Tinkerine Studio, on Flickr

Edited 1 time(s). Last edit at 07/12/2012 09:23PM by ECSuyu.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

It's hard to tell scale. It looks like a NEMA17, but a NEMA17 would weigh more than 210 on its own.

btw. That seems like a good idea laser cutting the gears. Probably quite a bit less expensive than buying them.

Hi Billyzelnack,

Thanks for the correction, you are right, the weight is without the motors as we weighed just the plastic portions as the motor was mounted on the machine. The gears do have some problems at this moment due to the plastic characteristic of acrylic, we are most likely moving towards HDPE or delrin for the gears as they dont squeak during use.

billyzelsnack Wrote:
-------------------------------------------------------
> It's hard to tell scale. It looks like a NEMA17,
> but a NEMA17 would weigh more than 210 on its
> own.
>
> btw. That seems like a good idea laser cutting the
> gears. Probably quite a bit less expensive than
> buying them.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

I'm pleased to see the work going into this, I am still a little worried due to the weight. I suspect there will be resonance in this setup, you also wont be able to reach the high acceleration the ultimaker boasts. It's not so much a problem, but a missed opportunity. For the record, it is becoming commonplace for RepRaps to exceed 100mm/s (Prusa's) and many cases 200mm/s (MendelMax, Ultimaker), the latter should be the benchmark.

Again I would urge you to explore a low-weight (& fast) extruder option, such as Bowden or eMinebea direct mount. I also think I was a bit too harsh on bowden cables, and may be an option you should really consider. After a lot of fine tuning it is possible to minimise the ooze that they are renowned for. I also heard talk of Marlin implementing retract during travel, which should further reduce any ooze.

> but a missed opportunity. For the record, it is
> becoming commonplace for RepRaps to exceed 100mm/s
> (Prusa's) and many cases 200mm/s (MendelMax,
> Ultimaker), the latter should be the benchmark.

It is commonplace for people to post about their printers going that fast, but it is not commonplace for people to post the results of their printers going that fast.

Edited 1 time(s). Last edit at 07/13/2012 11:05AM by billyzelsnack.

billyzelsnack Wrote:
-------------------------------------------------------
> It is commonplace for people to post about their
> printers going that fast, but it is not
> commonplace for people to post the results of
> their printers going that fast.

Perhaps in these forums that is true, but on IRC it is certainly commonplace to post the result as a video or an image.
A lot of results are questionable, but 100mm/s infill should pose no problem to the majority of Prusa's and 200mm/s for more rigid machines, of course having a lower speed for perimeters and even higher speeds for travel is usually beneficial

What I'd like to see is people showing off their overall build time improvements. A video of two non-contrived prints. One at 20mm/s for a quality baseline. One at whatever. Both prints of similar quality, but one taking significantly less overall build time.

@yydoctt- Thanks again for the suggestions, though we could develop this non-stop, there has to be a point where we say, both the quality and print speed are good so the machine should be released to the general public. After which we can continue to develop and release improvements for the machine as it progresses. It's true that bowdens work, but it is definitely not user friendly for the people that are wanting to get into 3D printing for the first time, and the ease of use aspect of a machine is one of our main design philosophies going into this project.

We agree with billyzelsnack here in regards to print speed. It's rare to see someone actually printing at 200mm/s. Since we come from a design background we've had quite a bit of exposure to commercial 3D printers, and to this day we have yet to see one that is willing to print at 50mm/s+ to reduce time at the cost of print quality. Even though Ditto has successfully printed at 100mm/s, the average user would probably not do so. What we are after is a balance of print speed to print quality for what we give to the users as a baseline profile for settings; we are not going to constrain them to those settings so they are free to adjust it to their needs. Afterall, the firmware and software are all open source developed by the community.

---

___________________________________________________
Kreativ-e
Tinkerine Studio

yydoctt Wrote:
-------------------------------------------------------
> A lot of results are questionable, but 100mm/s
> infill should pose no problem to the majority of
> Prusa's and 200mm/s for more rigid machines, of
> course having a lower speed for perimeters and
> even higher speeds for travel is usually
> beneficial

In my experience the ability of the printer to just move the print head accurately enough at those speeds is not a problem at all, the problem is figuring out how to get the extruder to work reliably and accurately at those speeds. Coming up with fancy frame designs while there are no extruders to take advantage of them makes little sense. I spent some effort chasing the 100+ mm/s speeds but in the end reliably getting strong and accurate results at 50 mm/s won me over...

Actually moving the print head at those speeds IS the problem, a Prusa cannot accurately move the head due to resonance in the frame. A MendelMax can barely reach 200mm/s with a decent acceleration, in which it too see's problems due to throwing a huge Nema 17 extruder 100mm and stopping abruptly.
There is a lot of talk of the limiting factor being the extruder, but time and time again we see extruders up to the task. For one Lulzbot, for second the age old makergear. The true limiting factor is moving the print head with enough acceleration to reach these high speeds, but most importantly doing this accurately! This is where lightweight XY axes have the advantage and therefore lightweight extruders

I print at 80mm/s on my Prusa because the frame resonance impacts the print quality, not the extruder performance, which can handle printing well above 200mm/s (same extruder as on my MendelMax).

What layer height and nozzle size? Vids please. When I look around on youtube I don't find convincing evidence. I just see people claiming high speeds where the video obviously shows it is not achieving it. The videos are also terrible at showing decent shots of the final print.

This is the best evidence I have found.
[www.youtube.com]
It is printing very fast at "only" 100mm/s and the quality looks decent with good fill too, but I wonder how often the print is actually achieving 100mm/s. If that print took 5 minutes then how long would the print of taken at 50mm/s? Overall print time and quality is all that matters.

Also in my experience even if you can get the correct volume of plastic out of the hotend that does not mean it is going to properly adhere to the layer below at higher speeds. Thread overhangs and cornering are particularly troublesome. This means the slicing software needs to be smarter about going slower in trouble spots and going slower of course increases build time.