The “competition would love me to reveal that” excuse has certainly ran its course.
We’ve already confirmed that the motors, pulleys, belts and control board are the same. I have to concur that the only real remaining “component” in the motion system is the software, hence the claim “enabled by a software update”.
They’ll likely stretch this claim the same as they did for the focus capability.
As to the optics, I haven’t opened the magic box for fear of screwing with the alignment, so I am not sure what it contains. This is from the patent application:
The box just seems to be two mirrors, a window and a gasket. It aligns the mirrors perfectly at 90 degrees, which means the beam comes out parallel to the beam coming in. They say rotating about the pivot point displaces the beam in the Y direction but maintains it parallel in the XY plain.
That implies the tube needs to be perfectly parallel to the gantry. Is there any way of adjusting it? I would not expect the beam to be perfectly aligned with its glass envelope as I can’t imagine glass blowing to be that accurate. I may be wrong though. The mirrors on my tube are factory adjustable with grub screws as they need to be perfectly parallel to each other. I don’t know if that allows them to be made perfectly orthogonal to the envelope when held at specific attachment points.
99% of the patent seems to be waffle. They seem to be trying to patent having one or more fixed items in a CNC machine!
Can someone with a Pro see if they can snap a picture to see if they are same value? Should be able to stick a smart phone in there or use a mirror to avoid disassembly.
As I am currently working with the laser drive/power system, I’d be interested to see what the Full Power settings are in the motion files for various units.
When I set a score for 100% on my Basic, the power is set to 0x66 in the motion file. The maximum value is 0x7F, so it’s 80% of the possible output from the power supply.
If some people want to volunteer: import this svg OpenGlow_Text_Plain.zip (1021 Bytes)
into the GFUI, and set it for a manual score at FULL Power (not 100%). Speed doesn’t matter.
Then follow these directions to grab the motion file. Send me the file, and I’ll pull the power value out of it.
My hunch is that besides the obvious difference in values between Basic’s and Pro’s, there will also be variations between units to act as a calibration for individual tube performance.
Note: The power setting can only be found in motion files with scores or cuts. Engraves set the power at 100% and then pulse the laser to achieve the desired power level.
EDIT: Corrected the power setting. Should be “FULL” not 100%.
Or perhaps the tubes do have different max power, but they target each of them to output at their rated max power at around 80% of actual max power (to prolong tube lifetime).
The percentage I am calculating is the percentage of the maximum output of the power supply, which is the same for both models (i.e. 80% power output in a Basic is equal to 80% power output in a Pro).
My initial guess, without having any direct measurements yet, is that the PWM resolution is low - meaning that only 3 steps represents a significant power increase.
This may explain why they don’t use the PWM setting to control lower power values, and instead use a dithering scheme. It would give them greater resolution, especially if it is non-linear.
PWM probably controls the current linearly. Although a tube is more or less constant voltage, as long as it has some positive slope resistance the voltage will also increase, so the power increase will be a little more than the increase in current.
Maybe the 45W tubes are more efficient. Possibly they are the same tube but some work better than others, so they grade them.
A interesting test would be to also compare them at 100% to see the relative efficiencies.
In the Pro pulse file, the X/Y motion starts immediately after the header.
In the Basic pulse file, it starts 21,436 cycles later. A delay of 2.14 seconds.
I haven’t found a technical reason for this, other than to maybe deliberately degrade the apparent performance/speed of the Basic model. If you are watching them side by side, you would say “WOW! Look how much faster the Pro is!!”
They both take 7448 cycles from the start of X/Y motion until they reach the point where the laser first fires. So, at least they haven’t artificially slowed the ‘G0’ rapid move speed.
When I was looking at files from your basic some time ago there wasn’t any delay at the start. I think it was mentioned in the forums that a fan spin up delay had been added as it was much requested to stop smoke escaping at the start. I also think I remember some people saying they didn’t have it. Perhaps they simply forgot to add it to the pro? Their level of incompetence never ceases to amaze me. They still haven’t managed to refund the correct amount to me yet. They don’t seem to understand overseas wire transfer fees. Odd because I am sure I am not the first OS customer to cancel. Did everybody else just accept losing some Dollars?
45W * 5.04 / 5.71 = 39.7, which is very close to 40W. When they use software PWM the average power must be directly proportional to mark space ratio, so it seems that the pro tube produces more output power for the same input power unless there is some other difference, such as a PSU setting.
Perhaps the pro tube just has more expensive mirrors and everything else is the same. That would be hard to spot visually. It also explains the new plus model because that has pro power but not pro cooling. That shows the pro tube doesn’t need more cooling, which would be the case if it is more efficient.
Possible. The Pro file is from yesterday, and the Basic file is from today.
I don’t recall ever not having the long space at the beginning of the pulse files I looked at, but I could be wrong. Do you still have any of the earlier files?
I haven’t found any other settings for the HV power supply other than the watch dog, enable, and pwm. The pinout accounting is pretty thorough at this point.
Yes I still have all the files I looked at. I checked a random one to make sure. Starts with Z steps and then XY motion. The Z steps can take some time if the focus is high as they step it very slowly.
Discourse doesn’t like conversations with only two people!
I have no doubt that some Glowforge staff are lurking. (Hi guys! We do really like some aspects of the product, honest! Please keep adding new features!)
