36 Hours, 200 Flasks, and One xTool S1: A Rush Order Story

The call came in at 9:47 PM on a Thursday. I remember the time because I was already in bed, and my phone lit up the whole room.

It was a client — a mid-size corporate events company we'd done work for twice before. Their client's annual gala was Saturday night. Two hundred stainless steel flasks, each laser engraved with guest names and a company logo. Their original vendor had just bailed. Not a slow-down. A full bail, twelve hours before the engraving window opened.

Normal turnaround for a job like that is 5-7 business days. We had 36 hours, and only two of those hours overlapped with standard business hours.

The First Question: Can We Physically Do This?

I've handled 200+ rush orders over the past four years. The first thing I do — before I even think about price — is figure out whether the job is possible. Not "possible with heroic effort." Possible, period.

Two hundred flasks meant two hundred individual engravings. No batch shortcuts, because each one had a different name. At 4-5 minutes per pass on a diode laser, that's roughly 15 hours of continuous machine time, minimum. Plus setup, plus reloading, plus the inevitable redo of three or four units where the alignment was slightly off.

We had one xTool S1 40W in the shop, plus a smaller desktop unit that wasn't going to cut it for stainless. The S1's cutting area — roughly 19.6" × 12" with the standard bed — was tight for a flask lying flat, but workable with a jig. I'd done singles before; never 200 in a night.

The upside was clear: a $6,800 job, and a client who would remember we showed up. The risk was equally clear: if the machine choked halfway through, or tripped the breaker in our rented space, we'd have 100 half-engraved flasks and a client on the phone at 6 AM.

The Power Problem (and Why It Almost Killed the Order)

Here's the thing nobody tells you about running a high-wattage diode laser for 15 straight hours: your wall outlet matters.

Our shop is in an older commercial building. We'd run the S1 for 3-4 hour stretches before without issue, but never a marathon. I pulled up the xTool S1 power consumption specs around 10:30 PM — 40W module draws around 200W during operation, but the peak draw with the air assist and enclosure running was higher than I expected. If our circuit was already sharing load with the space heater in the back room, we were going to trip.

I called our landlord at 10:50 PM. He wasn't thrilled. He also didn't say no. We agreed to kill every non-essential circuit in the building except the one outlet feeding the S1. (Note to self: buy a dedicated 20A circuit for the laser bay before next winter.)

Finding Files at Midnight

The client sent over a logo in JPG. Low-res, white background, the kind of file that looks fine on a screen and terrible on metal. We needed vector.

So at 11:15 PM I'm on three different sites looking for laser cutter files free downloads — not because I wanted to use someone else's design, but because I needed a flask outline template that matched the specific shape they'd ordered. Rim curvature, base radius, that sort of thing. Tracing it manually would've eaten an hour I didn't have.

Found a decent SVG on a maker forum (this was back before the big Etsy file crackdown, so options were wider). Modified it in LightBurn. Traced the logo. Ran a test engraving on a scrap flask I keep around for exactly this reason.

Looked good. Maybe slightly shallow. Bumped the power up 8%.

The 3 AM Doubt

We started the production run at 12:40 AM. By 3 AM, we'd done 62 flasks. The machine was holding steady, the air assist was keeping the lens clean, and honestly, the engravings looked sharp — better than the test piece.

But I kept second-guessing. What if the jig shifted on flask 140 and we didn't notice until the morning? What if the client's expectations for the laser engraved flask finish were more "deep etch" than what we were producing? The two hours between 3 AM and 5 AM were the longest of the night.

Around 5:15 AM, my partner — who'd been reloading flasks while I monitored the software — pulled flask 118 off the bed and said, "This one's actually really nice." She held it under the shop light. The logo caught the metal cleanly. No burn marks, no blur.

That was the moment I relaxed. Not because we were done — we had 80 more to go — but because I knew the process was solid. If we could do 118 well, we could do 200 well.

The Last 80 and the Delivery

We finished at 11:20 AM Friday. 200 flasks, 196 perfect, 4 sent back for redo (all from a single misaligned jig batch around flask 155 — caught it, fixed it, re-ran them in 25 minutes). Total machine time: 14 hours and 40 minutes, plus change for setup and redo.

I delivered them to the client's office at 2 PM. The events coordinator opened one box, pulled out a flask, and just said, "Oh, these are nice." That's the whole review. I'll take it.

Final cost breakdown: $6,800 for the job. $200 in rush shipping for extra blanks (our supplier came through at 7 AM — don't ask me how). $340 in electricity for the building that month, up from our usual $180. The margin was thinner than a standard order, but the client has already sent us two more jobs since.

What I'd Do Differently

Three things.

One: I'd have tested the circuit load before 10:30 PM. That was an avoidable hour of stress. If you're running an S1 or any 40W-class diode laser for long stretches, figure out your power situation before the rush order lands. As of early 2025, most residential and light-commercial circuits can handle the S1's steady draw, but the peak draw with accessories is the thing that trips breakers. Check your amperage.

Two: I'd have a flask jig pre-built. Making one from scrap MDF at midnight cost us 40 minutes I could've spent sleeping.

Three: I'd stop apologizing for taking small jobs. The client who called us at 9:47 PM on a Thursday had placed a $400 order with us two years ago. That $400 order is why they called us instead of someone else. Small doesn't mean unimportant — it means potential. (I really should put that on the shop wall.)

As for the question people always ask after a story like this — is the S1 the best laser for cutting wood or the best laser for anything? It's not. It's a diode laser, and it has limits. Thick metal, deep acrylic cuts, industrial volume — that's not its lane. But for engraving stainless flasks, cutting 3mm plywood jigs at 2 AM, and running 15 hours straight without a hiccup? It did the job. That's all I needed it to do.

If you're staring down a rush order of your own, the math is simple: figure out your machine time, check your power, build your jig, and start. The flasks don't engrave themselves.