Xtool F1 Ultra Glass Engraving: The Settings I Actually Use After 80+ Hours of Testing

Forget 100% Power — Your Glass Will Crack. The Right Setting Is Lower Than You Think.

After eight months and roughly 80 hours of running the xTool F1 Ultra 20W fiber & diode laser engraver on glass—tumblers, wine glasses, picture frames, flat panels—here's the short version: for clear glass, start at 60% power, 250 mm/s speed, 1 pass, with the 450nm diode module. Depth of engraving: 0.1-0.2mm max. Go higher on power, and you'll get micro-fractures. Go slower than 200 mm/s, and the thermal shock will chip the surface. I learned this the hard way. I assumed the fiber laser (the 20W IR source) would be better for glass because it's more powerful. Turned out that assumption cost me about $120 in cracked blanks in the first week.

I manage creative procurement for a 45-person company. We run a small in-house sign and promotional products shop. My job is to order supplies, vet equipment, and keep the production team happy. When I pushed for the xTool F1 Ultra, I promised it'd handle glass engraving without a separate CO2 unit. It does—but only if you ignore the manufacturer's 'maximum settings' tables and dial way back. This wasn't in the manual. It took three broken wine glasses and a very patient laser operator to figure out.

Here's exactly what I've landed on after systematically testing power, speed, and focus settings on six types of glass. If you're buying this for glass work, read this before you hit 'Engrave.'

Why I Chose the xTool F1 Ultra for Glass (And Almost Regretted It)

I'll be honest: when I compared quotes for a laser engraver kit last year, the F1 Ultra wasn't the cheapest option. A dedicated CO2 desktop unit from another brand was $400 less. But I was looking at total cost, not just unit price. We needed to engrave metal and glass and acrylic. A single diode unit can't do metal. A CO2 unit can't do metal either. We would've needed two machines. The F1 Ultra's dual source (20W fiber + 20W diode) meant one machine for everything. That $400 'savings' on the CO2 unit would have turned into a $2,000 problem when we realized we needed a fiber laser anyway.

That said, the first month was frustrating. The 1064nm fiber module can mark glass, but mostly creates a frosted surface with very shallow depth—about 0.05mm. It's great for serial numbers or barcodes. It's terrible for picture engraving on glass where you want contrast and tactile feel. The 450nm diode module, surprisingly, does a better job for photographic work. It produces a whiter, more opaque mark that actually shows up against the glass.

Here's the counterintuitive part: with the diode module, I started at 80% power (thinking 'more power = more contrast'). Wrong. The glass got so hot it micro-cracked along the engraving lines. Dropping to 60% actually improved the visual quality and eliminated fractures. The marks had better contrast at lower power because there was less scattering from surface damage.

Reference: Industry standard for laser safety glass (e.g., Schott Borofloat) recommends thermal gradient not exceeding 60°C/cm during processing. Above this, thermal stress exceeds glass fracture toughness (~0.75 MPa·m^1/2). The F1 Ultra's manual doesn't mention this. (Source: Schott Technical Glasses White Paper, accessed January 2025)

Glass Engraving Settings That Actually Work (xTool F1 Ultra)

These aren't theoretical. I've run each at least 10 times on different glass types. Adjust for your specific material—your mileage will vary.

Clear Soda-Lime Glass (e.g., standard wine glass, flat panel):

• Module: 450nm Diode (NOT fiber)
• Power: 60-65%
• Speed: 250-300 mm/s
• Passes: 1 (max 2 for deeper frosting, but risk increases)
• Focus: 0.0 mm (surface focus; slightly below surface for deeper mark)
• Result: Frosted white mark, 0.1-0.2mm depth, low fracture risk

Borosilicate Glass (e.g., Pyrex, lab glass):

• Module: 450nm Diode
• Power: 55-60%
• Speed: 300-350 mm/s
• Passes: 1
• Focus: 0.0 mm
• Result: Lighter mark. Borosilicate has higher thermal resistance, but it's still prone to chipping at higher power.

Coated/Tempered Glass (e.g., some drinkware):

• Module: 450nm Diode OR 1064nm Fiber (test both; coating affects absorption)
• Power: 50-55% (lower if coating is sensitive)
• Speed: 300-350 mm/s
• Passes: 1
• Result: Can burnish or remove coating, not really engrave. Fiber module sometimes marks through coating better.
• Warning: Tempered glass may explode if too much heat is applied. I've had it happen twice. Don't engrave tempered glass without a test piece.

The Picture Engraving Problem: Why My First Photos Looked Terrible

When I told our production team the F1 Ultra could do picture engraving on glass, they got excited. We have a client who wanted a photo of their dog on a crystal paperweight. I set up the image in LightBurn, ran it at the default 'glass' profile—which was set to 80% power, 150 mm/s, fiber module. The image came out patchy, with hot spots in the darks and almost no detail in the highlights. Looked like a bad photocopy from 1995.

Here's the thing: grayscale engraving on glass is not a simple conversion from image to laser. Glass doesn't 'hold' half-tones the way wood or coated metal does. The mark is either there (white frost) or not (clear). To get a photo effect, you have to use dithering—specifically, Jarvis or Stucki dithering at about 45 line per inch (LPI). The diode module handles dithering MUCH better than the fiber module, probably because the beam diameter is larger and creates more consistent coverage.

My current recipe for photo engraving on flat glass:

• Use the diode module, NOT fiber
• Power: 60%, Speed: 250 mm/s, 1 pass
• Dither: Jarvis, 45 LPI
• Focus: +0.5 mm (slightly above surface for softer mark)
• Result: Acceptable grayscale photo effect, about 70% of what a CO2 laser produces on acrylic, but serviceable for gifts and small-run production

Honestly, I'm Not Sure Why Some Glass Types Behave Differently

I've never fully understood the science of how soda-lime vs. borosilicate absorbs near-IR differently. My best guess is it has to do with iron content—borosilicate has less iron, so it's more transparent to the diode's 450nm wavelength, requiring lower power to avoid reflectivity issues. If someone has a materials science background, I'd genuinely love to hear your take. I'm just the guy who buys the stuff and watches what breaks.

One thing I have confirmed: the rotary attachment (sold separately for the xTool F1 Ultra) is mandatory for cylindrical glass. Trying to engrave a wine glass without rotary is basically impossible—the curvature distorts focus so badly that your image wraps around unevenly. Adding the rotary cost us another $199, but it eliminated the 50% reject rate we had on tumblers. That was a no-brainer investment.

On that note: if you're considering this machine for simple laser cutter projects on acrylic or wood, glass settings don't matter to you. The F1 Ultra is a decent engraver for those materials too, but its real value is the material flexibility. You're paying for one machine that does metal, glass, wood, and acrylic. For my budget, that single-machine capability justified the $1,499 price point (as of December 2024). I'd still buy it again, but I'd have insisted on better training from the vendor on glass settings before we put our first piece in the machine.

Bottom line: The xTool F1 Ultra 20W dual laser engraver can absolutely do glass. It's not magic—you need to test, adjust, and accept that some coatings won't engrave well. But if you dial back power, stick to the diode module for photos, and use the rotary for curved surfaces, you'll get good results. Just don't trust the default profiles. They're optimized for 'wow' demos, not production reliability.