What about tube replacement if GF folds? I know they don't manufacture the tubes themselves, but I wonder if there is any GF IP right that means Yongli can't sell the tube directly? From what I understand it is a bespoke tube design.
What about tube replacement if GF folds? I know they don't manufacture the tubes themselves, but I wonder if there is any GF IP right that means Yongli can't sell the tube directly? From what I understand it is a bespoke tube design.
G. Weike LG900N 100W RECI RDWorks V8
Leiming LM2513FL 1kW Raycus fiber laser cutter
Wisely 50W Raycus engraver
What is it that you can't cut .5 inch material with the tray installed
When this subject first came up, one of the now-departed fanboys (whose name I have purged from memory) claimed that the cloud server was directly controlling the motion-system motors. What I pointed out at the time was that the latencies of the interface between the cloud server and the motors make that impossible, and yes, I am still of that opinion. Now, if you want to define "cloud motion control" as "a cloud server sending vector endpoint coordinates to onboard motion-control circuitry", we can discuss that, but that definition is about as meaningful as calling the GF a 3D printer.
I'm also under the impression that several of the more unique advertised software features (e.g. auto-alignment) of the GF still don't work. That's pretty much the definition of vaporware: advertising/selling something that doesn't yet exist.
Yoga class makes me feel like a total stud, mostly because I'm about as flexible as a 2x4.
"Design"? Possibly. "Intelligent"? Sure doesn't look like it from this angle.
We used to be hunter gatherers. Now we're shopper borrowers.
The three most important words in the English language: "Front Towards Enemy".
The world makes a lot more sense when you remember that Butthead was the smart one.
You can never be too rich, too thin, or have too much ammo.
It's nothing special Mike, it's just a DC tube that uses the same methodology as pretty much every other one in the world. The PSU however is pure GF designed and built even though they were warned by YongLi not to do it as it may cause stability issues.From what I understand it is a bespoke tube design.
You did what !
And Lee I am also wondering that is the purpose of the Cloud? Is it just to prepare the gcode?
There are thousands of plain O laser engravers running just fine on existing software. If the camera does not work, the 3D does not seem to be that functional so now its just a slow laser engraver? Better ones can be purchased anywhere.... even eBay.
Retired Guy- Central Iowa.HVAC/R , Cloudray Galvo Fiber , -Windows 10
I suspect one of the reasons for using a Cloud based solution was that they knew they were not going to have the software tried and tested before starting to ship machines. That way they can make code changes on the fly and not have users download and run frequent updates on their PC's. They are probably using an Agile approach where you release code as sprints rather than following a traditional release strategy. It has it's advantages but I don't like the dependency on a live internet connection and I'm not sure how resilient their cloud solution is.
G. Weike LG900N 100W RECI RDWorks V8
Leiming LM2513FL 1kW Raycus fiber laser cutter
Wisely 50W Raycus engraver
Good morning,
think you created the latency theory to then shoot holes in that straw man. That was not what GF advertised or implemented, to the best of my knowledge. It is/was my understanding that a cloud-based motion control was used in lieu of a control card in the machine. This was to reduce cost and to implement cloud-technology tech to update and make improvements on-the-fly. Regardless, it doesn't matter how or if you understand the exact implementation, but it seems to work and folks are using it to create things, contrary to your original opinion.
I would disagree with your definition of vaporware. They did build a laser and are delivering into the hands of customers. They will probably continue to make software improvements and refine the machine. You're just moving the goalposts now.
That's the lay of the land now Bill.
What are you cutting with a laser that is 1/2 inch? Just curious.
See above.
Sounds nefarious.
It is funny to read opposition to internet connection on the internet. It's sometimes about access in remote low-population locations, but I think it's mostly about control.
Again, I wasn't responding to what "GF advertised or implemented", I was responding to what one of the GF apologists claimed. And using your "understanding" of GF's implementation to accuse me of a straw-man attack is a bit rich unless you have inside information we do not regarding that implementation. What you seem to be saying is that my observations about cloud-based motion-system control are wrong because the machine appears to work at some level now regardless of how it's implemented.
Bill George's conjecture that they use the cloud server just to prepare gcode (or some set of much simpler control parameters) is probably correct, but, as I said, that is not "cloud-based motion control" by any definition I'm prepared to accept. Fun fact: when I say "much simpler control parameters", that's exactly how my ULS works. It transmits HPGL-like commands (actually formatted as a small subset of EMF) to a rather primitive control board, where it is translated by an ASIC into low-level stepper commands. Note that most of this control board most certainly exists on the GF, since all but one chip and a handful of discrete components are necessary to actually transmit step/direction pulses to the motors. The complexity of that function is orders of magnitude smaller than implementing whatever hypothetical protocol might exist to communicate with the cloud.
So some of the things that I deemed vaporware two years ago now work, and some don't. Ok, it's less egregious vaporware than it was two year ago. When/if the machine works as advertised at that time, it will (finally) be vaporware-free and you can say "I told you so"...I'm not holding my breath.
Last edited by Lee DeRaud; 11-05-2017 at 11:22 AM.
Yoga class makes me feel like a total stud, mostly because I'm about as flexible as a 2x4.
"Design"? Possibly. "Intelligent"? Sure doesn't look like it from this angle.
We used to be hunter gatherers. Now we're shopper borrowers.
The three most important words in the English language: "Front Towards Enemy".
The world makes a lot more sense when you remember that Butthead was the smart one.
You can never be too rich, too thin, or have too much ammo.
I cut .75 acrylic with the beam focused to the center of it. using a 4 inch lens.
Right. You were speculating and offering conjecture that turned out to be incorrect, in hindsight.
Don't need my understanding. From GF:...using your "understanding" of GF's implementation to accuse me of a straw-man attack is a bit rich unless you have inside information we do not regarding that implementation.
"We do both CAM and motion planning in the cloud. We just send the machine the actual waveforms to be fed to the motors (X, Y, each fan, focus, etc), laser, chiller, etc. There's a bit of decompression and then the uP sprays the results at the right I/O lines."
No, that is not what I'm saying. This machine has cloud motion-control and does operate. This is what was offered in relation to this and the original discussion. You opined that it was not possible and got lost in the weeds conjuring a made-up scenario about latency. Any rational and reasonable person would probably guess that you were referring to the GF and not just speaking academically.What you seem to be saying is that my observations about cloud-based motion-system control are wrong because the machine appears to work at some level now regardless of how it's implemented.
See above.Bill George's conjecture that they use the cloud server just to prepare gcode (or some set of much simpler control parameters) is probably correct, but, as I said, that is not "cloud-based motion control" by any definition I'm prepared to accept.
There appears to be no control board on the GF, so you might be incorrect here.Fun fact: when I say "much simpler control parameters", that's exactly how my ULS works. It transmits HPGL-like commands (actually formatted as a small subset of EMF) to a rather primitive control board, where it is translated by an ASIC into low-level stepper commands. Note that most of this control board most certainly exists on the GF, since all but one chip and a handful of discrete components are necessary to actually transmit step/direction pulses to the motors. The complexity of that function is orders of magnitude smaller than implementing whatever hypothetical protocol might exist to communicate with the cloud.
Well, I think you're creating a new definition to suit you opinion. Something is vaporware or it's not.So some of the things that I deemed vaporware two years ago now work, and some don't. Ok, it's less egregious vaporware than it was two year ago. When/if the machine works as advertised at that time, it will (finally) be vaporware-free and you can say "I told you so"...I'm not holding my breath.
Go learn something about motion control, then come back and tell me how that differs from the ULS implementation I described, except that they implement the CAM/motion-planning functions in a cloud-based server instead of a printer driver. Hint: motion control starts after motion planning...you can't control the motion until after the path is determined.
So the "uP" GF refers to is, what, duct-taped to the inside of the cover? Pick a story and stick to it.
Yoga class makes me feel like a total stud, mostly because I'm about as flexible as a 2x4.
"Design"? Possibly. "Intelligent"? Sure doesn't look like it from this angle.
We used to be hunter gatherers. Now we're shopper borrowers.
The three most important words in the English language: "Front Towards Enemy".
The world makes a lot more sense when you remember that Butthead was the smart one.
You can never be too rich, too thin, or have too much ammo.
Correct, real time motion control would be next to impossible with the controller off premises. I do not think that could even be implemented if the wireless feed was in the next room let alone over the real internet. Anyone who has worked with, repaired or built CNC knows it depends on constantly knowing where the carriage, milling head or laser is at and sending reliable feedback to the controller. Yes I know steppers do not have feedback, but it depends on knowing the last command was received.
Retired Guy- Central Iowa.HVAC/R , Cloudray Galvo Fiber , -Windows 10
I design, engineer and program all sorts of things.
Oh, and I use Adobe Illustrator with an Epilog Mini.