Our daughter, who got me into GF to begin with, upgraded her GF+ to a Thunder. She loves it, and it too cut the cutting time to about half what the GF does. Plus she does a lot of the tumblers and stuff on it.
Either way, until I pay off bills, I can’t afford to get anything new, and just keep praying that my GF stays working as well as it is!
it would be interesting to know the design. i don’t think the GF would necessarily be a lot slower on cut time compared to something with 10w more power. but the GF is significantly slower on engrave times because the head just won’t move that fast. the universal i have at work is 75w, so nearly 2x the power, but the engrave times are closer to 4-5x faster because the head on the universal can just move so much faster since the rails carry so much less weight.
It was a sheet of large words to cut, no engraving.
Can’t wait to hear more about it and your experience! I hope this laser turns out to meet the hype.
We might have more success if we ask them about the " box moistening fluid" .
All 3 of the machines I have received have had a notable fluid leak during shipping.
thanks for the review. very helpful.
It’s been a week, so I guess I can share some more early thoughts.
There’s almost no learning curve switching from the Glowforge webapp to xTool Creative Space, which is also actually a webapp, just wrapped up as a desktop install via Electron. It has built-in settings for common materials, design tools don’t require an extra paid subscription, and I can use all the same SVG files I made for the Glowforge. It’s even got an AI art generator.
I haven’t reached for the Glowforge much this week since the P2 does all the same stuff, but faster. Autofocus for material height, camera-based alignment, built-in air assist and exhaust fans, built-in coolant tank/pump, big button that lights up that you press to start a job… I feel like it’s another pretty direct Glowforge clone though the xTool owners community never seems to feel that way when I mention it there.
It has a secondary camera on the print head, like the Glowforge does, but unlike Glowforge you can ask it to go take a picture (or multiple) anywhere on your material for a super accurate close-up shot to align your design against. That’s nice and I don’t know why Glowforge doesn’t let you use its print head camera for anything yourself.
The machine offers a wifi connection, but your computer will be directly connecting to the laser via your local network, not via a cloud server intermediating like Glowforge does it. The upside is that you can do everything even if the internet’s out. The downside is that even a momentary connection dropout disconnects the software from the machine. I prefer how Glowforge does it TBH. I don’t know why people so often complain about Glowforge being web-based – who wants to update desktop software and hardware drivers every time there’s a new feature?
The P2’s early customer base appears to include many Glowforge owners based on photos people are sharing of their workspaces as their new laser arrives… and xTool knows this, as one of their first software updates was to treat colors in SVG files as different layers, so that files made for the Glowforge can be easily used with their machine.
Maintenance on the Glowforge is easier. Fully cleaning the optics on the P2 requires pulling out a hose from the air assist and undoing two screws to access the lens. I’m sorely missing the ease of Glowforge’s metal lens holder that you can pop out with a magnet. I’m hoping that with a nozzle-type print head, the lens won’t get dirty as fast so I won’t need to clean it as often as I do in the Glowforge – which was around every other day.
Why would this happen? Router malfunction?
They copied Glowforge a bit too well by also using a flaky wifi chipset and antenna? That’s how it’s feeling based on my usage so far and the kinds of posts other owners are making about wifi issues. The machine beeps when it connects to wifi, and it beeps rather often in the middle of print jobs, meaning it is reconnecting itself. I have a mesh wifi system and wonder if it’s connecting to one of the further access points instead of the one in the same room. But I can only speculate. It does have Ethernet and USB ports as well, so flaky wifi is never going to stop someone from being able to work at least.
This is something I have also experienced with my new Carvera. Another customer pointed out that it looks like they left the PCB antenna connected on the board, which you’re not supposed to do when using an external antenna, and this degrades the signal.
Nevertheless, it doesn’t matter as much why this happens – network connections, and especially wireless connections, are inherently unreliable. Software must be designed to assume the connection will be flaky. Ideally, by continuing to work offline and quietly reconnecting in the background, not hanging, crashing, exiting, or popping up error messages.
Glowforge’s design is pretty good this way. While the machine’s WiFi itself could obviously be more reliable, the app doesn’t need the Glowforge to be online to work, and vice-versa. Carvera’s is somewhat worse, in that the app behaves badly if its connection to the machine goes away, but at least it runs its jobs from local storage, so you only need to be connected to make adjustments and start it running. I’d be terrified of a machine which depended on an active WiFi connection for the entire job.
This.
The module used in the GF is ubiquitous, you’ll find it in many wifi-connected devices, and there is nothing inheritablly bad about it.
Sure you can get better modules but this one works just fine.
I think the print head camera is extremely limited to only see the red laser in a certain range. The lid camera can be extremely accurate with a good calibration, with the major downside that it cannot see the work when the printhead is in the way. I do not find that to be a terrible downside 
I don’t think that’s the case, since it’s also used for passthrough alignment. Also, Scott was able to grab some images from it and they appeared to be full color.
It takes about 1"-square images at typical focus range, 1080p or so. It’s pretty good quality, and used for pass-thru alignment and camera calibration.
1080p is 1920 x1080 so it’s not 1” square. Unless you mean total area which I can’t be bothered to do the math on the real dimensions.
Anyway I’ve always wondered about pixel pitch at that point because it’s a fundamental limit as to the camera’s accuracy.
If we assume you want a 1 square inch 1080p resolution you should be able to figure out dimensions and therefore pixel pitch.
x2 + y 2 =1
x / y = 1920/1080
Someone go figure out x and y. K thx!
(This assumes square pixels)
To calculate the pixel size of a full HD screen with a total area of one square inch, we need to know the resolution of the screen. Full HD resolution is typically 1920 pixels in width and 1080 pixels in height.
Given that the screen has a total area of one square inch, we can calculate the pixel size by dividing the width and height by the square root of the total number of pixels. In this case, the total number of pixels is 1920 * 1080 = 2,073,600.
Let’s perform the calculation:
Pixel size = sqrt(area / total number of pixels)
= sqrt(1 / 2,073,600)
≈ 0.0000165 inches
Therefore, the pixel size of a full HD screen with a total area of one square inch, assuming square pixels, is approximately 0.0000165 inches.
It’s about an inch from the pics I have, but I can’t extract them myself to verify the precise dimensions.
I’ve never given it a second’s thought. I just know it’s a typical 1080p or so sensor that has a fixed focus that works for our needs.
Seems to me that pixel pitch of 0.001 is needed to get the reported 1/1000th accuracy, yet as we all know that isn’t what we actually see in practice. I think you’re lucky if you get 1/100th in passthrough alignment. Still pretty good but an order of magnitude worse than their claimed abilities.
Like I said, never given it a second’s thought. Cuts work perfectly. Engraves fail between steps because the laser does not fire on the first pass of the next step. Reported, nothing ever done about it.
I have used the passthrough alignment maybe three times, and each using Snapmarks. Instead I use the technique we had before snapmarks, but I thought that was done with the lid camera and not the one on the head.
I just set up my XTool P2 this week and so far I can’t be more pleased. Mostly because of the accuracy. It will print a design where you tell it too. Not somewhere within a 1/4" like GF. It’s faster and quieter. It takes getting used to there Creative Space software but all and all. I love it. The true test will be if it breaks down a month after my warranty expires like GF pro did. and the week after that and the week after that …
