For the last month every plot in this studio has run on one machine, the UUNA TEK ArtStation 2436. It is a 36 by 24 inch CoreXY pen plotter, and it replaced my A3 machine as the daily driver the week it arrived. This review follows six plots that took it from a first squareness check to the seed of a new series.
Full disclosure first: this machine and this review are part of a UUNA TEK content partnership. They have had no say in the contents of this review, and the opinions expressed here are my own.
Some context for reading it. This is a data art studio, and pen plotting is the medium: the machines are not a novelty here, they are how the work gets made. So I do not evaluate a plotter on flashy features. I ask four things of a machine: precision, the same behavior every day, repairs I can do myself, and a fit with the tools and habits I have already standardized on. That is the lens for everything below.
The working area is 36 by 24 inches (915 by 610 millimeters), which covers an A1 sheet with the margins intact. The frame is a closed rectangle of extruded aluminum with the gantry supported on both sides. The bed is a magnetic steel plate: sheets are held flat with loose magnets, which is faster than tape and leaves nothing on the corners of the paper.
The motion system is CoreXY, and I like CoreXY plotters. Both stepper motors stay fixed at the frame's top corners and a single crossed belt moves the carriage, so almost nothing heavy travels: the motion is smooth, the design scales well in size, and the footprint stays fixed. Cantilevered plotters photograph well and look nice on a desk, but the unsupported arm can wobble and sag, and maintenance on them can be tough. I come from laboratory automation, and good old extruded aluminum rails work just fine for me. My iDraw H SE is the same closed-frame CoreXY layout at A3, so this is the geometry I am coming from; the 2436 scales it up rather than replacing it.
Three numbers from the spec sheet matter, translated into pen terms. Minimum motion resolution is 0.0125 millimeters, about one twenty-fourth of the line a fine fountain-pen nib leaves. Stated repeatability is plus or minus 0.05 millimeters at low speeds, one sixth of that same 0.3 millimeter line. Mechanical precision is 0.1 millimeters. Together they mean that if a line lands visibly wrong on this machine, the cause is the pen, the paper, or my file, not the positioning. A month of plotting bore that out.
The controller (DrawCore V1.3) runs EBB firmware over plain wired USB, the same protocol family my existing toolchain already speaks, so the machine slotted into my pipeline on day one with no new software. If you do not have your own toolchain it runs from Inkscape; I never needed to open it. And the whole thing works offline, with no account and nothing phoning home.
Shipping was a bit of a headache. I ordered the machine on July 15 and it arrived by ocean freight on August 26, six weeks later. The box arrived with a puncture, and the impact had knocked the frame out of square: the base had racked into a parallelogram relative to the top rail the head runs on.
The fix turned out to be simple, and the design deserves credit for that. The baseplate is detachable and adjustable, so the repair was: loosen the screws holding the base to release the tension the shift had put into the frame, realign the base against the top rail, check the diagonals until they match, and retighten. Under fifteen minutes, and the frame has held square through everything since, including the full-bed and stress-test plots below. UUNA TEK's support answered my damage photos within a day with exactly those steps, which is what I want from a vendor when freight goes badly.
Past that repair, setup was plugging in the stepper motor connectors and the servo wiring. That is the whole assembly story: the machine ships essentially built, and I was plotting the same day.
My previous daily driver is an iDraw H SE, an A3 machine on the same CoreXY closed-frame layout. It is the machine this studio grew up on, and it taught me exactly what to distrust in a plotter. Its bed developed a belly over time. I found it with a dial indicator swept across the plate and fixed it with shims, and since then I do not take a flat bed on faith from anyone.
On paper the two machines are close relatives. The differences that matter are size and rated speed; nearly everything else carries over. In the photo above, the white A3 sheet is the entire print area the H SE could give me; it sits on an A1 cutting mat, and the mat itself fits the ArtStation's bed with room to spare. This plotter is huge.
| iDraw H SE (A3) | ArtStation 2436 | |
|---|---|---|
| Working area | 297 × 420 mm (A3) | 610 × 915 mm (24 × 36 in) |
| Machine footprint | 560 × 460 × 115 mm | 1060 × 773 × 115 mm |
| Weight | 6.1 kg | 15 kg |
| Motion | CoreXY, closed frame | CoreXY, closed frame |
| Minimum motion resolution | 0.0125 mm | 0.0125 mm |
| Rated max speed | 200 mm/s | 280 mm/s |
| Controller | EBB firmware, wired USB | EBB firmware, wired USB |
| Bed | developed a belly, shimmed flat by me | magnetic steel plate, flat as delivered |
The H SE also set my registration baseline. I have run as many as 18 layers on a single A3 sheet on that machine, so I know what good layer-to-layer registration looks like at small format and what it costs in care. The question this month was whether a bed four and a half times larger keeps that behavior. The six plots below are the acceptance tests, in order, each chosen to answer one thing.
Before the tests, the thing that changed my day-to-day most. Plotting A3 on an A3 machine means working at the machine's limit all the time. There is no margin for error in placement, and the head and rail always need adjusting to clear whatever is on the stage. Registration in that regime is an exercise in squeezing: lining a new pen up as far up and over as the machine allows, because the piece already occupies everything the machine has.
A3 and the imperial sizes just beyond it are the bread and butter of my catalog, and on a 36 by 24 bed they sit in the middle of the plate with room on every side. I have room to align and move things around, and I stage the next layers' pens on the deck itself while a layer runs (which I do constantly and do not recommend). Registration gets fundamentally simpler when the print is nowhere near the envelope. It shows up most clearly in mixing pen types: fountain pens, 3 millimeter acrylic markers, and fineliners all sit at different resting heights and positions in the clamp, and on the old machine a swap at the edge of the envelope could push the new pen past what the machine could reach. On the 2436 I can run all three in one A3 piece without losing registration. That headroom is the most liberating thing about this plotter, more than the maximum sheet size itself.
The first plot was deliberately simple and deliberately huge: a torus I rendered from Turtletoy, drawn as a single layer with hatching, with a standard 0.7 millimeter gel ballpoint, on a full 36 by 24 sheet.
With one pen and one layer, any problem shows. I wanted to know whether the frame was still square after my adjustment, whether the rails were aligned, and whether the carriage could reach and draw cleanly everywhere on the sheet. A skipped line, visible ringing on the long sweeps, or hatching that drifts would have set this review on a different course.
None of that happened. The hatching is even across the full width, I found no skipped lines, and the curves show no ringing I can see. A gel ballpoint at 0.7 millimeters is the most forgiving tool I own, so this was the machine's easiest test, and it passed without a caveat. The full plot, start to finish, is on Instagram.
Next, color registration. For what would have been Freddie Mercury's 80th birthday I plotted an 18 by 24 portrait in six layers, one pen per layer: Staedtler fineliners on lightweight Strathmore drawing paper. The fineliners are my staple for multi-color work. They are reliable, they bleed very little, the tips hold up over hours, and on this paper they glide. They are less lightfast than I would like, which I accept for test pieces.
I was worried about the dense passages, the hair and the microphone, where six layers of marks pile up. They came out clean. End to end the piece took five hours at relaxed speeds, and the print matched the render. The plot video is on Instagram.
The third test inverted the scale question. Most of my catalog needs density and precision more than it needs size, and some of my early pieces used every millimeter the A3 machine had. So I plotted an A3 bird study, eight layers, about twelve hours of plotting, again at relaxed speeds. The question was whether a machine built for 36 inches loses anything when the work is small and fine.
It does not, as far as my eyes can tell. The finished sheet reads as delicate, which is not an adjective I expected to reach for with a 15 kilogram gantry machine. Eight layers registered cleanly, and the piece would not look out of place next to anything the A3 machine produced. The plot video is on Instagram.
From here the tests get less forgiving. I drafted a rhombille tiling with light hatching, the kind of geometry where perpendicularity is the aesthetic. If the axes are not true, a rhombille pattern tells you immediately, and at any viewing distance.
This was also my first piece on this machine with Pilot Kakuno fountain pens, fine nibs that draw a line around 0.3 millimeters. Fountain pens are notoriously fussy across layer changes in precise work. I did nothing special between layers, and the pattern closed up correctly anyway. I would encourage you to watch the plot video and zoom in on the intersections.
Then I tried to break it. The fountain hatch ladder is an 18 by 24 sheet, three fountain-pen layers, built as five sub-tests: a hatching ladder of increasing density at six angles, perpendicular cross-color hatching at increasing density, a dash and stipple spacing test, a circle filled with strokes at 5 degree increments across layers, and a spacing sweep from 0.45 to 4 millimeters. This is the plot where a machine with a bulging bed, uneven belt tension, or sloppy servo control has nowhere to hide. It is the test I wrote because of the old machine's bed.
The geometry passed. The ladder lines are parallel and evenly stepped at every angle, the radial fill shows no wobble and no uneven spacing, and there is no regional fading that would map to a low spot in the bed.
Two findings, and I attribute neither to the machine's motion. First, the ink density in the single-color ladders is uneven, and the purple, green, and red cross-hatch at 0.8 millimeters shows the same blotching. The strokes themselves are perfectly placed. My path optimization (I run vpype over every plot file) reorders strokes for plot time, so the pen revisits regions at different ink states, and fountain pens are unforgiving of that. I would still expect disjoint segments to look more uniform than this, so the pens themselves carry part of the blame. Second, the ends of long segments plot slightly darker and thicker than their middles. That is a pen-profile gap: speed, acceleration, and lift rates tuned to how a fountain pen delivers ink would close it. I tune profiles when a piece demands it rather than in advance, and no piece this month demanded it.
The last plot is the reason I wanted this machine, and the start of work I could not have made at A3. The TP53 gene codes the p53 protein, a tumor suppressor that guards the genome. p53 does not work alone: it functions as a tetramer, four identical copies assembled into one working unit.
The piece draws the DNA-binding domain of the canonical p53 sequence (UniProt P04637, residues 102 to 292) four times on an 18 by 24 sheet: one copy per monomer, one layer and one fountain pen per copy. A small Python pipeline pulls the sequence, scores each residue's side-chain polarity on the Kyte-Doolittle scale, and converts the scores directly into line pitch: hydrophobic stretches pack the lines tighter, the way they pack in the folded protein, and polar stretches open them up. The four copies are one family of curves repeated under the tetramer's own symmetry, mirrored left to right for the palindromic DNA site it grips and top to bottom for the dimer partner, with the twin layers tilted a few degrees apart. The pipeline emits the geometry as layered SVG and the machine does the rest.
One detail connects this piece to the stress test before it. Fountain pens have an ink floor: below a certain line spacing the hatch closes up into a wash. The hatch ladder measured that floor at 0.8 millimeters for these nibs, and the piece is built to respect it, dropping curves out as the weave tightens rather than ever crossing it. The test sheet calibrated the artwork.
I grew up with Afghan carpets, and their geometry, form built from many crossing runs of thread, is the visual grammar I wanted for a protein that only exists as a weave of four strands. Where the four staggered copies cross in the middle of the sheet, a form assembles that none of the single copies contains. That is the tetramer, drawn by its own sequence.
This is a first study. The machine held four near-identical dense line fields in register across just under three hours and two pen changes, and where the four families cross at the center of the sheet, moire fringes appear that exist only because the layers are in register. The interference is the artwork, which makes this about as direct a registration test as I can design. I will be building this idea into a plotted series on this machine, and there is an exhibition in the works around it. The plot video is on Instagram.
A month in, the machine has passed every test I designed to catch it out. The frame stayed square after one initial adjustment, the bed is flat by dial indicator and by print, reach is honest across all 36 inches, and multi-layer registration held from a six-layer portrait to four staggered copies of a protein sequence. The failures I did find, ink-state unevenness and heavy stroke ends, belong to fountain pens and to my own unfinished pen profiles, and they would have appeared on any machine.
The one design gripe I have is the servo ribbon. The flat cable that feeds the pen-lift servo stands in a tall arc above the carriage and sweeps the near side of the machine as it plots, so the 2436 wants about 18 inches of clearance above the deck and on the operator's side. It has never caught anything in a month of plotting, but I have kept my desk clearer than usual out of concern that it could, and on a machine this well behaved that is the complaint I am left with.
I used the phrase "relaxed speeds" all month, so here are the numbers: my standard fineliner profile commands about 49 millimeters per second pen down with acceleration at half scale, against a rated ceiling of 280 millimeters per second, and the fountain pen profile is gentler still. I run my machines that way on purpose. Rushing art has cost me more sheets than any hardware fault ever has, so I have never pushed a plotter toward its limits or toward some production quota, and I think that respect is a large part of why plotters have worked out for me. One month is also not enough time to test the extremes honestly, and this review does not claim to. If I am so inclined I will check in a year from now with a longer report on the harder questions, wear and sustained speed among them. What I can say is that at the pace I actually work, consistency and quality held from the first squareness check to the last layer of the p53 piece.
The value case, as someone who standardizes hard on plotter, pens, and paper: the 2436 is the H SE's quality at four and a half times the plotting area, reached without changing anything else. My workflow is frozen on purpose, and a new machine usually breaks that freeze: new software stack, new interface, new parts, new failure modes, and weeks of adjustment before you trust it. The 2436 behaves like a scaled-up iDraw. It speaks the same protocol over the same wired USB, my existing toolchain drove it on day one, and the only thing I rebuilt was my pen profiles, which are per-machine work on any plotter. For a studio that needs a solution that just works, that continuity is worth more than any single spec.
On price, it sits in a useful gap. Cheaper large-format plotters often ship without a bed, and fabricating a surface that is flat and coplanar to the rails is not hard, but it is exactly the kind of variable I standardize to avoid. The options that do include a bed tend to cost more and plot smaller. At $2,774 as I write this, with the magnetic steel bed included, the 2436 undercuts both paths.
It also travels better than a 36 inch machine has any right to. At 15 kilograms one person can pack it, and I kept the shipping material because I take my plotters to exhibits, fairs, and schools to run in public. The H SE went back into its box with the original cushioning and never lost calibration; loading the 2436 up works the same way and has been just as simple. I am pickier about where I bring it, since it needs real table space on display, but it makes the trip.
That is what has impressed me most about what UUNA TEK has built here. A studio like mine starts modestly, on a small machine, and standardizes everything around it: pens, paper, profiles, habits. The 2436 respects all of that. It asks an artist to change nothing except the size of the paper.
Buy the 2436 if you are upgrading from an iDraw H SE, as I did, because you want to plot larger and more sophisticated pieces. The simplicity of these machines is a feature, and it is what makes them maintainable and reliable.
