Sunday, April 18, 2021

Prusa MK3 vs Voron 2.4

Which is the better printer?

Recently I had an argument with Chris Warkocki the admin of the Prusa community group on facebook where he casted some aspersions on the Voron printer. so he claimed that every of his customer printers would print much better and faster than the Voron and the Afterburner would have a very bad cooling.

So there is a lot to unpack here. 

First, of course you cannot have an unbiased opinion about other printers if you have such close ties to Prusa and being the admin of the fb fan group. That is human nature and we know fanboys from other areas in life. But let us have a look at the substance of his claims and make some comparisons.

Costs

A stock Prusa MK3s as DIY kit comes with a price tag of 769,-€. As Chris already mentioned that his printers are heavily modded, let's add the most common mods. 

The bear frame kit without printed parts 89,99€

One spool of PETG for the printed parts 25€

Bondtech extruder upgrade 120€

SE Mosquito hot-end 160€

Seems he replaced the EinsyRambo as well and tried SKR and Duet boards, even is experimenting with klipper that requires and additional Raspberry Pi. So that could range from 100€ to 300€ for the electronics, but to be fair we just assume 100€.

That is a total cost of 1263,99€ for the heavily modded Prusa MK3. Not take into account other mods like Mitsumi, Vesconite LM8UU bearings or real linear rails, different stepper motors, larger frame (z height) etc. 

For my Voron 2.4 I have spent roughly 1400€ including LDO steppers, robotdig linear rails, 4Gb Raspberry Pi 4, Meanwell power supplies, panels with printed Voron Loge, prepared cabling harness from fermio labs and stainless steel bolts and nuts order in large quantity(hundred more than I needed).

I have also added some parts that are not listed in the BoM to have a more safe printer that pushed the overall costs to 1500€

So the Voron 2.4 is a bit more expensive than the modded MK3, given the doubt the no other mods are made. Still they are in the same ballpark in regards to cost.

Features

The official build volume for the Prusa is 250 x 210 x 210 mm that cannot be 100% utilized from my experience. 

My Voron has an official build volume of 300x300x300 mm that can be 100% utilized as the motion system has 350mm linear rails.

The heat bed on a Prusa is basically a PCB that is relatively thin compared to the 8mm aluminium slab I have on the Voron. The thick aluminum plate is not only more stable and precise(machined surface), it is also more resistant to warping with temperature changes, the aluminum has more mass and is a very good thermal conductor. So the bed temperature is more stable and more unified across the surface area.

Another advantage is that the Prusa PCB head bed is powered by 150W(24V6.25A), while the Keenovo heat mat on my Voron is has 400W (230V). It heats up quicker and does not stress the power supply or board electronics.

The Prusa motion system is still based on 8mm steel roods, 2 per axis. The Voron comes with linear rails, 4 for Z, 2 for Y and 2 for X. 

Also part of the motion system are the 5 1.8° stepper motors on the Prusa(2 for Z, 1 for X,Y,E each) while the Voron has 7 Steppers in total, 4 1,8° for Z that are reduced by pulls in a ratio of 20 to 80 teeth, 2 0.9° for Y and X movements and 1 for the extruder.

In numbers that means the Tr8x8 lead screw on the MK3 with the 1.8° stepper motors allow a resolution of 0.04 mm layer height. With the Voron also with 1.8° steppers and a GT2 fully with 16 teeth, but a ratio of 1:4 gets to the same layer height of 0.04 mm. But on the X and Y the 0.9° steppers on the Voron allow for more resolution that the Prusa.

Both design use printed parts, The Prusa comes with PETG parts, ABS is used for the Voron as PETG fails in a heated chamber.

I will not rub on the EinsyRambo board on the Prusa vs my 2 SKR V1.4 turbo boards as we included 100€ for a board upgrade in the Prusa in the cost calculation above. But I guess it is obvious that the combination of 2 SKR boards with Klipper doing the heavy lifting on a Raspberry Pi outperforms the old Prusa 8bit board.

Both can use steel sheets as build pate with PEI stickers or powder coating texture that are held in place by magnets. The Prusa has some needy magnets as inlays in the heat bed PCB( that like to fall off over time). The Voron uses a magnetic sticker on top of the aluminium bed. Both are rated up to 120°C before the magnetic force degrades.

Both printers have a inductive probe for bed leveling by default that require a steel sheet. But in addition the Voron has a probe for the nozzle. That means if you swap the nozzle or the hot-end(which is very easy on the Voron) then the nozzle probe can measure the new distance of the nozzle relative to the print bed and you don't need to adjust manually every time like on the Prusa(first layer calibration).

Per design the Voron has an enclosure. That helps with materials like ABS that tend to warp, but also keeps the fumes inside. The Voron even has a simple air filtering system. 

The Prusa uses 3 pin fans that allow the firmware to detect if a fan is not spinning which could lead to overheat in case of the hot-end fan. The Voron uses only 2 pin fans by default.

In terms of security features the Prusa has some self check abilities. The Voron has a thermal fuse for the heat bed by default that kills the power to the heat mat at 120°C. 

I have added a filtered C14 power inlet in my Voron that has 2 fuses. In addition I also added a 6A circuit breaker with built-in ground fault circuit interrupter. But that is not part of the standard Voron BoM.

Print Speed

I have to admit that I did no do my own tests yet as my Voron 2.4 is not finished yet. So I can only rely on information that other Voron users share. I guess a full video of a print with a clock running is a trustworthy source. 

Problem is that there was no real "standard" print. The often used bench boat does not say anything about size, infill percentage, material etc. The range of benchy times on a Voron 2.4 range from under 10 to 14min. I have seen videos from MK3 around 15min for a benchy, but again it is hard to compare as there is no standard for the benchy.

The actual speed you put into your slicer is a desired value. Just because you put in 800mm/s it does not mean your printer can do 800mm/s. Same goes for adjustments on the printer itself. 

So there are some videos from MirageC where he claims to print with 800mm/s with his HevORT CoreXY. Vez3D has a very fast coeXY, so does Anney_Engineering.

There is even a speed benchy contest on the Annex-Engineering discord channel. They have rules for the settings that lead to more comparable results.

One Voron 2.4 from joshmurrah (standard with afterburner) with 11:15min for the benchy and one Mk3 heavily modded with bear frame, linear rails, different hot-end/extruder and running klipper with 12:43min from Techromancer that looks a bit like cheating as he uses only one corner of the bed and thereby uses the advantage that the gantry if stiffer at the end as in the middle. And his MK3 is running klipper as well.

What I have not seen is a real fast benchy from Mr Warcocki. Not even a full video of his fast prints. Sometimes he shared short video sequences on FB and claimed a speed that cold not be verified in any way. 

To understand why the bed slingers like a Prusa Mk3 can hardly beat any CoreXY is very simple. There are physical limits. While the CoreXY design has the power of two stepper motors pulling at any X or Y movement, the MK3 design has a single stepper that must move the whole print bed plus the print itself(more mass).

Having that said the current best time is a delta printer.


Conclusion

To me the Voron 2.4 stock printer outperforms the Prusa MK3, even when the MK3 is heavily modded in many areas. You simply get a bigger, faster printer that already come with an enclosure.

When you want to opt for speed you can add a high flow hot-end from the beginning and don't have to throw money for a standard hot-end out of the window.

Changing the hot-end on the Voron works like a charm and is so easy. Compared to the mess on the MK3 where you have to completely disassemble the x-carriage and cabling to the board.

Why someone would spent >750€ on a printer kit, just to throw away most parts while replacing them and spend more money instead of directly building the printer of desire is a mystery to me.

I guess you have to be a real fanboy.




  






The end of the Dragon Hotend?

Is this the end? Let's try to unravel the mystery.


What is the Dragon hot-end?

The Dragon hot-end is a hot-end that was originally introduced by the Chinese company Phaetus. It combines elements from the E3D V6 with elements of the Mosquito by Slice Engendering. They have copied the design of the heartbreak and the way that the heat block is mounted to the heat sink. The E3D V6 uses the heat-break also as the mechanical mount between the heat block and heat sink. Unlike the V6 design the Mosquito uses small stainless steel pipes with small screws to heat block tot he heat sink, that allows to have a very thin metal tube with cooling fins as heat break. This part was obviously copied by the Dragon hot-end.

Mosquito by Slice Engineering

On the other side the Dragon kept the exact out dimensions of the E3D V6, including the mount ring which is optional on the Dragon(at least the TL versions I have). So in contrast to the Mosquito with its very distinct shapes that usually requires a different mount and even a different Extruder body (on the Prusa Mk3 for example), the Dragon is a drop-in hot-end that does not require any changes.

clone of the E3D V6

Not only the mount, also the heat block and the mounting for the heater cartridge and the thermistor are the same on the Dragon and the E3D V6.

Triangle Labs Dragon without mount ring

Slice Engineering has a number of patents on their products that can be found here, some are still pending.  https://www.sliceengineering.com/pages/ip

Now it seems like Triangle Labs fears that SE could sue them for patent infringement and being forced to pay royalties. This is why they have now stopped selling the Dragon hot-end.


If Phaetus will also stop selling the Dragon hot-end is not clear yet. But definitely bad news for all users of the TL Dragon as they might not get any spare parts in the future. 
According to Mr. Warkocki, the admin of the Prusa community group on facebook, SE has denied that they actually threatened Triangle Labs. But that does not mean that could go down that road in the future and Tl just proactively stopped selling it to avoid problems in the future.

Other sellers hoped up on AliExpress now, selling Dragon hot-ends for less than $20. If they are even close tot he quality of the TL Dragon hot-end, I don't know.

If there is enough substance for an infringement claim would be up to a court to rule on. Or if the SE design is that ground breaking different from the V6 that is justifies being protected by a patent could also be subject for discussion.

To me the fact that 3D printing was already invented in the early 1980s(FDM 1989), then protected by patents until the early 2000s held by companies like Stratasys until it became available for hobbyist because of the reprap project in 2008, is a point against patents. Especially as it seems we are now going back from open/free technology, back into the world of patents and closed development.

I guess Prusa is a good example that you can have development and make money while you open source your products, if you have a good product for a reasonable price and offer a good service.







Thursday, April 1, 2021

Mosaic Palette 2S Pro or 3 Pro?

Two weeks ago I finally ordered my Palette 2 Pro. I hate the Prusa MMU2 as it is so unreliable und far from being easy to use. So I was thinking to get a Palette for a while.

But I have so many other projects going on that 750€ was quite a sum for me right now. But while waiting for my Voron 2.4 parts I decided to get one, finally.



But this week Mosaic announced the Palette 3 Pro, for the same price as the 2S Pro and in addition you get a $100 off for pre-orders. So seems that was just bad timing.

But what really annoyed my was that there is no upgrade this time as the internals have changed so much. I would appreciate if they would give a bigger discount for existing customers.

I have not yet decided if the Palette will go on one of my MK3 clones or even on the new Voron. I guess the Palette would slow down the Voron.

I also got most of the mechanical parts for my Voron 2.4. The SKR boards are still missing, according to tracking they are in Germany since begin of March. The heat mat is still not available. I started to put the frame and rails together today and already ran out of M3x8mm screws. So I guess having the build mostly done over Easter will not happen.



A new project "automatic cat litter box"

 I did not post any updates for a long time. I spent some time making some money with 3D printing to recover some of the costs from this hob...