ComMarker B4 30W: Six Weeks In, Honest Verdict

I bought the ComMarker B4 30W on a fall promo. ASIN B0C2PVXG4G, $1,299 from a $1,599 MSRP. Six weeks later the verdict is that the B4 does exactly what I bought it to do — and that I bought the wrong thing. Both of those are true and they don't cancel out. I sat with two browser tabs open for three days before I clicked. ComMarker B4 30W in one, OMTech 30W JPT MOPA in the other. Same wattage, same price bracket. The B4 has dual lenses in the box (110mm and 200mm), a built-in motorized Z, and a detachable head for handheld marking. The OMTech has variable pulse width — which means real color on stainless. I told myself I didn't need color. I was wrong, and we'll get there. For now, I bought the B4 because I went 30W (everything I'd read said 20W stalls on thicker steel, and I had knife blanks coming) and I liked the dual-lens setup more than I valued color I didn't yet need. Day one ended with me staring at a frame box LightBurn was drawing on a piece of 6061 aluminum that was roughly 30% bigger than the actual work area. I'd re-seated the lens twice, restarted the controller, swapped the USB cable, and watched a YouTube video where a guy with a clean shop confidently calls his B4 "plug and play." Mine wasn't. Not yet.

The B4 ships in two boxes. The main unit is 19.5 kg with the head attached. That weight is real — you don't one-hand it onto a bench. The head detaches with a single locking lever, which is the feature that makes "handheld mode" work, and it's also why everything inside has to be torqued correctly. In the box: the main column with the laser source, the detachable galvo head, the 110mm lens (already mounted), the 200mm lens (in a foam case), an EZCAD2 USB stick, a red dot module, the standard rotary chuck, a power brick, and a printed quickstart that does not mention LightBurn even once. Three things you only learn after you open it. One: there's a markcfg7 file on that USB stick — a LightBurn config file that sets up the laser source profile, lens calibration, and the corfile field size. You have to import it before LightBurn will frame correctly. If you skip this, you spend nine hours like I did. Two: the LightBurn Galvo license is not the LightBurn license you already own. It's a separate $60 SKU. Plan for it; the machine is functionally crippled on EZCAD2 alone. Three: swapping from the 110mm lens to the 200mm lens isn't just unscrewing one and screwing on the other. You have to reload the corfile in LightBurn so the field size matches the lens, or your frame box is wrong on every job.

The framing problem was the corfile. LightBurn's default galvo profile assumes a 125x125 mm field. The B4 with the 110mm lens is 110x110 mm. Until I went into Device Settings, edited the corfile field size to 110, and reloaded, the red-dot frame box was drawing a rectangle 13% larger than where the laser would actually fire. The fix took ninety seconds once I knew what I was doing. Finding the answer took most of a day. It's buried in a ComMarker forum thread and a single line in a LightBurn documentation page. ComMarker's quickstart doesn't mention it. The first time you frame a job and the box matches the workpiece — that's the moment the machine clicks. After that, no more drama. For anyone setting one up new: import markcfg7, set corfile field to 110, drop a known-size rectangle on a piece of anodized aluminum, and frame it. Measure the frame with calipers. It should be within 0.2 mm. If it isn't, run a galvo calibration scan; LightBurn handles this in five minutes. After that, you can trust the machine.

Once the framing is honest, the B4 is fast and it's deep. My working recipe for a permanent deep mark on 6061 aluminum: 40% power, 1500 mm/s, 40 kHz, 6 passes, 0.04 mm fill spacing. That gets me about 0.4 mm of depth on a 50x50 mm logo in roughly 90 seconds. Brass takes the same recipe but goes a touch deeper. Hardened tool steel — I marked a couple of knife blanks for a friend — wants 50% power and 8 passes; same speed and frequency. The 15,000 mm/s galvo speed isn't marketing fluff. On surface marks (single pass for a part number on an aluminum tag) I run at 3000 mm/s and the head moves like the workpiece is rolling underneath it. Production-pace work is genuinely production-pace. The failure mode at the top end is slag. I cranked it to 100% / 1000 mm/s / 6 passes on aluminum to see what would happen, and the surface around the engrave bloomed into recast splatter. Lesson: more power isn't deeper, it's messier. Stay under 60% on aluminum, lean on pass count for depth, and the marks are clean. This is the section where I'll say it plainly: if you're a knife maker, a deep-engrave shop, or a hobbyist marking tags, plates, and brass keepsakes, the B4 30W earns its bench space. Six weeks in, it's done every job I've thrown at it in this lane.

Here's the line I read four times and didn't actually understand: "limited color marking." That's ComMarker's own copy, on their own product page, halfway down a comparison chart with the B4 MOPA variants. Limited means none of the rainbow stuff you see on Reddit. The B4 30W is a Q-switched fiber laser. The source is a Raycus RFL-P30QS — the QS in the part number stands for Q-Switch — and it has a fixed pulse width somewhere around 130 nanoseconds at 40 kHz. You cannot shorten it, you cannot lengthen it. Color marking on stainless steel works by tuning pulse width down into the 2 to 200 ns range to control the thickness of the surface oxide layer; that oxide thickness is what your eye reads as a color. Fixed pulse width means fixed oxide. You get black. You get dark gray. You don't get blue, gold, red, or purple. The machine will produce a color-ish result on plastics and on anodized aluminum at low power, because the pigment in those materials reacts to heat. That's not the same thing. If you want to mark a stainless tumbler with a brewery's two-color logo, this machine is the wrong tool, and there's no firmware update or settings trick that changes the physics. ComMarker sells a separate B4 30W variant with a JPT MOPA source (ASIN B0DRY12JMW). Same body, totally different machine, $300 to $500 more depending on the day. If you want color, that's the SKU, or you go OMTech. See /guides/mopa-vs-q-switched for the full breakdown — I wrote it after I learned this the slow way.

I lasted three days on EZCAD2. EZCAD2 works. It also feels like a Windows 98 control panel grafted onto a CNC machine. It's free, it's Windows-only, and antivirus software flags the installer as suspicious because the signing certificate is unusual; I had to disable real-time protection during install and re-enable it after. LightBurn is mandatory. Not nice-to-have, mandatory. The interface is the same one you'd use for a CO2 or diode laser if you have one already; the galvo-specific controls (fill spacing, pulse settings, ramp lengths for 3D, hatch angles) are exposed cleanly. 3D sliced engraving — the technique where you do a depth-graded fill to make raised lettering — is workable in LightBurn and effectively broken in EZCAD2. The Galvo license is the trap. You buy LightBurn once for $60 and that license covers diode and CO2. The Galvo add-on is a separate $60. You have to buy it specifically; they don't bundle it. Total LightBurn cost to actually run a fiber: $120, one time. Build that into your machine budget before you click buy on the laser. Worth every penny — the difference between LightBurn and EZCAD2 on this hardware is more than the difference between a 20W and a 30W.

Six weeks of real use. Maybe 90 hours of actual run time, plus another 40 of setup, testing, and teardown. The wear list so far is shorter than I expected and longer than I'd like. The lens cap thumbscrew on the 110mm lens stripped at week three. It's an M3, brass, and the threads chewed themselves the third time I changed lenses. Replaced with a stainless M3x6 from a hobby box; problem solved. The rubber feet under the main column compressed unevenly. One foot is now slightly shorter than the other three. Doesn't affect marking, but the machine rocks on a flat bench. I'll shim or replace. The focus rod loosens. There's a single thumbscrew that locks the head height, and after about every 40 hours of run time it backs off enough that the head sags 0.3 mm during a session. The mark quality drifts before the rod visibly moves. Now I check and re-tighten that screw before any session over an hour. The red-dot pointer drifted out of alignment around hour 60. Two minutes with a hex key and a calibration scan in LightBurn put it back. This will probably keep happening; I budget for it. Nothing has actually broken. Nothing has needed a warranty claim. The Raycus source is rated for around 100,000 hours and there's no reason to think mine is any different. The machine is solid. The accessories around the source are where the wear lives.

Buy the ComMarker B4 30W if: you're marking metal in monochrome, you want depth and speed, you want two field sizes for the price of one, and color isn't on your roadmap. Knife makers, deep-engrave shops, jewelers doing serial numbers and maker's marks on gold and silver, electronics hobbyists doing PCB labels, makers running aluminum tags and brass plates. This machine earns the bench. Six weeks in, I'd buy it again for this work. Don't buy it if: you want color on stainless tumblers, you want anneal-black on titanium with controlled tone, or you might want any of those things in the next year. Buy the OMTech 30W MOPA at the same price point, or step up to the B4 30W MOPA variant. Don't try to grow into color on a Q-switched machine; the physics doesn't bend. Don't buy the cheaper B6 20W as a way to save money on this category either. The B6 is a fine 20W with a bigger 150 mm field and autofocus, but it's also Q-switched, and the 10W of headroom on the 30W is the difference between marking 6061 in 6 passes and grinding through it in 12. If your budget can stretch from the B6 to the B4 30W, stretch. The deeper read on this: I should have bought a MOPA. I knew it three days into ownership when I tried to mark a friend's stainless growler and produced a perfectly clean gray logo that would have been a perfectly clean gold logo on the OMTech. The B4 is doing exactly what I bought it to do. I just bought the wrong thing. That's the most useful sentence in this review, and it's the reason post #2 in this series is going to be a head-to-head between this machine and the OMTech.