I've Burned Through 4 Test Boards Learning the xTool F1 Ultra—Here's the Checklist I Now Run Before Every Paid Job
- Who this checklist is for
- Step 1: Confirm the material (and the laser) before you quote the job
- Step 2: Never trust a default profile you haven't seen run
- Step 3: For photos on wood or leather, run a test grid, not a hero shot
- Step 4: Calibrate the rotary before you load the tumbler
- Step 5: For fiber marking, dial in focus and frequency before you dial in power
- Step 6: G-code—yes, you can touch it, but log every change
- Step 7: Air assist and exhaust—run them like they're part of the machine
- Common mistakes I still see (and have made)
Who this checklist is for
If you bought an xTool F1 Ultra to run small-batch metal or glass work for clients, or you're about to take delivery of one, this is for you. I run a small engraving shop in the Midwest—mostly Etsy sellers, three machinist shops, and a stubborn stainless tumbler guy who orders every month. I've been doing laser work since 2019, but I only picked up the F1 Ultra in April 2023. The first three months were expensive.
Below is the 7-step checklist I now run before I accept any paid job on the F1 Ultra. Not the marketing sequence, not the Facebook group's "just trust the presets" sequence. The one I actually built after burning through about $2,400 in stock and one very awkward client call.
Step 1: Confirm the material (and the laser) before you quote the job
The F1 Ultra runs two lasers: a 20W fiber source and a 20W diode. Same machine, very different material behavior. I lost my first $180 to this.
A local machinist asked me to mark some anodized aluminum tags. I threw them on the diode side because the presets were loaded, and the result was a chalky grey line that wouldn't hold. He was polite. He didn't come back.
Here's the check I now run:
- Bare/anodized metal, some plastics, certain ceramics → fiber source.
- Wood, leather, slate, glass (with prep), acrylic (with the right settings) → diode.
- Anything painted, coated, or laminated → confirm the coating before you commit. I now ask for a photo of the item and its spec sheet.
The F1 Ultra will technically fire on a lot of things. "Technically" is doing heavy lifting there. White acrylic in particular—it will cut, but with a diode+white combination that most people get wrong on the first try. I still test-grid it.
Step 2: Never trust a default profile you haven't seen run
In my first year (2019) with a different machine, I made the classic "trust the preset" mistake on a $600 order. On the F1 Ultra, the presets are better—they're actually pretty good—but they're a starting point, not a finish line.
What I do now:
- Open the XCS software profile for the closest material.
- Write the suggested speed and power into the job log (yes, on paper, on a clipboard).
- Cut a 1" × 1" swatch on scrap from the same batch.
- Adjust one variable at a time—power first, then speed, then frequency for fiber.
Took me maybe 9 months to understand that the material batch matters as much as the preset. Plywood from two suppliers will engrave differently on the same settings. Stainless from one distributor will mark at 60% frequency and another at 40%. Same supplier, different lot, different result.
Step 3: For photos on wood or leather, run a test grid, not a hero shot
This is the one people skip. It's the one that cost me most.
The F1 Ultra handles xtool f1 ultra photo engraving well, but the resolution you see on a 3"×3" sample doesn't always scale to a 10" plaque. Background gradients that look smooth on a small test get muddy on a big one.
My current photo routine:
- Convert the image to grayscale, then to a dot-density test pattern at three DPIs (254, 318, 381).
- Run the pattern on scraps from the same material batch.
- Pick the DPI that holds mid-tone detail, then engrave at that setting only.
That pre-check would have saved me the $340 pet-portrait redo in February—the customer's photo had a soft shadow gradient that only showed as banding at full size. I had checked it at 4" wide. It looked fine. It wasn't fine.
Step 4: Calibrate the rotary before you load the tumbler
The xtool f1 ultra rotary attachment is the part I see the most frustration around online, and honestly, most of that frustration comes from loading the workpiece before calibrating the rotary axis.
My rotation checklist:
- Mount rotary, tighten the chuck, and confirm the tumbler is actually centered—roll it by hand and look for wobble.
- Run the rotary calibration in XCS every time you change diameter. Every time. Even if it's the same brand of tumbler.
- Run a single-line test engrave on a hidden section of the item.
- Check that the seam line lines up after one full rotation.
The third tumbler order I ever ran had a 2mm offset at the seam. Two millimeters sounds small. On a ring of text, it looked like I couldn't spell. Customer noticed. I now refuse jobs without a hidden test area.
Step 5: For fiber marking, dial in focus and frequency before you dial in power
Fibre laser marking doesn't work like diode engraving, and I learned that the embarrassing way.
On stainless and anodized aluminum, the difference between a clean mark and a burnt halo isn't usually power—it's focus distance and pulse frequency. I mark at 30–60% power, but I'll spend five minutes getting focus to the spec (usually around 3.6–4.0" from the lens, depending on the setup) before I touch power at all.
I used to start at 80% power because that's what the video tutorials showed. That burned a $95 order of aluminum business cards to a crisp grey. Not scrap—just bad enough that the customer declined them.
A quick way to check: run a focus ladder on a scrap piece in 0.1" increments and pick the setting that gives the tightest line width. It takes four minutes and it saves re-runs.
Step 6: G-code—yes, you can touch it, but log every change
The gcode for laser engraver side of XCS is limited compared to LightBurn, but it's not nothing. If you're exporting or tuning passes, save the base G-code file before every experiment, not after. I lost two days in November trying to remember what I'd changed because I "just tweaked it in the software."
For fiber marking specifically, I keep a plain-text log of every successful custom G-code block—power, frequency, focus, passes, and the material batch it worked on. When a month-old job comes back for a re-order, I can rebuild it in two minutes instead of an hour.
Step 7: Air assist and exhaust—run them like they're part of the machine
The air assist on the F1 Ultra isn't optional, and it's not just for cutting. On engraving, it keeps char and resin off the lens. On fiber marking, it keeps smoke off the metal.
I made the mistake once of thinking air assist was a "cutting only" accessory. After about 40 minutes of engraving without it, I had a lens smudge that took an hour to clean and cost me a session.
Now it's part of the job setup sequence, not an afterthought. Exhaust too—if you're marking or engraving anything with a coating, exhaust is a safety item, not a comfort item.
Common mistakes I still see (and have made)
- Quoting jobs before running a material test. Even a 1" swatch. Even when the customer sends a spec sheet.
- Reusing a rotary profile across tumbler brands. The diameter is close—the calibration isn't.
- Judging photo quality on a small sample. Photo engraving fails at scale, not on a 4" swatch.
- Assuming "it fired" means "it worked." The fiber source will mark materials the diode shouldn't touch, and vice versa. Material choice is a discipline, not a suggestion.
- Skipping the lens clean between jobs. I do it every 4–6 hours of engraving now. The first time I skipped a week, I paid for it with a replacement job, not a replacement lens.
None of this is complicated. All of it is the kind of thing you only learn by getting it wrong once and then writing the correction into a checklist so you don't have to learn it twice.
5 minutes of verification beats 5 days of correction. That's not a slogan—it's what the $2,400 taught me.