What does i3 think of their G. Weike lasers?

[Brad Luyster]

LVL1 recently received a grant, part of which we're going to use to upgrade our laser cutter. Visiting i3 during Makerfaire was a great experience, and most of our members were super impressed by the enormous laser cutters in your shop. We're considering spending our money on a laser of roughly the same scale. What is the general i3 opinion of these cutters? If you could chose again, would you pick the same laser?  Did you import them directly, or buy from the importer, and who could I ask about the importation process?

Thanks in advance, and I welcome any and all to stop by LVL1 if you're ever passing through Louisville!

-Brad Luyster

LVL1 Hackerspace

[Nathaniel Bezanson]

I'll let others speak to the import process, but I'll give my technical impressions of the machine. Tl;dr: It's fairly easy to work on, which is good, because you're gonna have to.

There's nothing about CO2 laser cutters that should inherently require a lot of maintenance or attention. Keep the optics clean and the slides lubricated, and the rest should Just Work. Unfortunately, so many corners were cut in these machines, you'll end up repairing and replacing things you didn't even realize could fail. On the plus side, the things that've been breaking on ours haven't been particularly exotic, so it's more of a hassle than anything.

Here's an email I wrote a few months ago in response to a similar query from another space:

-------- Original Message --------
Subject: Re: Re: LC1290 questions
Date: Thu, 16 May 2013 12:35:27 -0400

The early problems we had were power-supply-related. There are NTC inrush current limiters on the AC input, but they blow after a few days. Our guess is that the 240v-centric design was only partially reengineered for 120v operation, and they forgot to account for the doubled current. The seller's tech rep suggested just bridging past them with wire, but if they're not needed, why were they included in the first place? Roger found some suitable replacements and it's been fine since.

More recently, there's the machine's propensity to smash itself. The limit switches are poorly placed and/or malfunctioning, and the auto-Z sensor is stupidly placed so it can crash into things during a home-axes operation. The auto-Z operation also should never be performed with the head in home position, as there the Z sensor is not positioned over the bed, so the bed's upward travel is never sensed, and continues until it crashes into the head, the motors grind, and a human slams the red button.

So much stupid. I'd say CNC machines should be smart enough not to smash themselves, but I've seen the Bridgeport drive its quill right through its own toolchanger, so perhaps I expect too much. 

Also, doublecheck all the safety switches. Ironically there are lockout switches on the main cover but not on the laser tube doors. These should inhibit the tube power supply. Also make sure the beam can't fire when the chiller tube is pinched or the water flow sensor is otherwise annoyed. 

You'll also want to reengineer the airflow inside the machine. Slats and honeycomb are pointless unless the air is being drawn downward through the table, which it isn't by default. We removed one of the planks which forms the solid bed beneath the slats, and piped a flex duct into that space directly. Could provide photos if ya like. That, combined with covering up the inactive honeycomb area, serves to "focus" the vent-suction and does a much better job of keeping the smoke away from the beam path and machine slides.

That's a point worth reiterating. The "easy" air connection results in the smoky air being drawn back towards the tube, so plastic vapor and wood rosins are condensing on all the important parts. It takes some work to change this, but it's worth it in machine and optics longevity. 

We also discovered last week that vapors condense in all sorts of places. We have our exhaust blower mounted remotely and powered through an SSR, with a little switch by the laser. The big noisy blower motor is on the back wall away from everyone. That's good, but it means the exhaust duct is quite long, and everything's nice and cold by the time it hits the blower's intake grille. Said grille had thus managed to almost-completely stop itself up with plastic and wood crud, in a spectacular pattern. So, make it a regular maintenance item to inspect your blower! 

The other main maintenance task is cleaning the mirrors. Personally I'd love to rig some air-curtain nozzles in front of the mirrors to make sure they're blanketed in fresh air even when the chamber is smoky, but haven't gotten around to it. The tube aperture window is perhaps more important, because it's not replaceable like the mirrors. 

The machine controller's USB interface is flaky/picky about its host. We tried mounting a hub inside the machine, with the LaserCut software dongle plugged into the hub alongside the Leetro controller itself, and had nothing but trouble with that scheme. Tried a few hubs, they didn't last long. I realize upon typing this that the controlling PC and the laser machine might be on different circuits and I hope they weren't doing evil with the USB cable, but that's worth investigating. Anyway, get a high-quality USB cable to run between the controller and the PC, you'll save yourself some hassle.

Which brings us to the software. It's awful. It works, but boy the various efforts to reverse-engineer the MOL format and write an Inkscape plugin are super promising. The fact that LaserCut can't directly import SVG is pretty much all you need to know about how awful it is.

Also, there's some more info about the machine on its wiki page:

http://i3detroit.org/wi/index.php?title=Bumblebee

That's my braindump for now,

-Nate B-

Since writing that, the blower intake clogged itself again, and I had the foresight to grab a pic before cleaning it:

http://imgur.com/rGa63qC

We've been contemplating a huge-surface-area prefilter to make these operations less frequent. That'll probably come along with the rest of the exhaust revamp, which we really need to do because the included blower is woefully inadequate to suck a reasonable volume through the length of flex duct we're using. Smoothwall should help a lot and might mean we can continue using the existing blower, but if that's not enough improvement, we'll end up with a bigger blower too. These things want a *lot* of airflow.

Also since writing that, the IEC power inlet socket and fuseholder melted. Judging by the construction of the socket and the ratings of similar parts, it was probably designed to pass 10 or 13 amps, and this machine will easy draw 14 or 15 for hours at a stretch. The high-intensity WTF, though, was the fact that the machine came with a 20A fuse installed in there. The 5x20mm fuse isn't even supposed to be *available* in a 20A variant. Improving the electrical safety of the machine occasionally means redesigning parts of it.

All that being said, I don't know if you could do much better without spending a tremendous amount more. All these machines are made in 240V countries, so their attempts at 120V conversion are shaky at best. I wonder if the situation might be better if we'd simply ordered the 240 version and run a new circuit for it. 

Okay, second braindump completed.

-Nate B-

[Brad Luyster]

Thanks for the response! These challenges are not unfamiliar to us, based on our experiences with an earlier 40W Full Spectrum Laser.  I'm glad to hear the warnings about the 120v conversion issues, if we go with a G Weike, we'll probably just go for a 240v unit.