Engrave off?

[Robert Silvers]

I keep on hearing how RF C02 is so much better than DC. I think we should have an engrave off - hopefully with US and Chinese machines.

Test 1:

Engrave 1mm high text in Arial. Type to engrave is "1mm tall Arial" Take close-up photo to show sharpness. Post machine, speed, and power. Substrate - painted metal or anodized aluminum.

Test 2:

Pick a common image to engrave, say a photograph with lots of grayscale, and see who can do best with what machine.

I looked around for a photo that has good exposure, a nice range of grays, and is a portrait. This is what I came up with:

http://www.washingtonpost.com/blogs/style-blog/files/2013/06/10011242H16190604.jpg

If we do this, we would need a common substrate. Black anodized aluminum would be best, but that is rare and expensive.

I propose mirror tiles from Home Depot because a six pack is $10 as you can get sharp small dots on it:

http://www.homedepot.com/p/Erias-Home-Designs-12-in-x-12-in-Plain-Edge-Mirror-Tiles-6-Pack-201010/202300825#.UliL42RATHI

Lowes has them also.

Engrave the back 10 inches tall, and then take a photo with it lit from behind. It provides extreme contrast because it can block the light 100%, or let a lot through. Note that you will have to reverse the image left-right, and probably invert it to not get a negative image. This will have to be half-toned. Post machine, speed, power, and how you made the image file.

Anyone?

[Dan Hintz]

These kind of "throwdowns" have been cast from time to time, but they really serve no purpose other than to get people riled up. Save your energy.

If you really must, Steve (Scott) has posted some sweet detail in the past from his machines... usually with sub-mm fonts, if memory serves.

[Kev Williams]

Robert, before reading this thread, I just posted in your other thread, a couple of pictures of some sample engraving from my LS900, and asked if you wouldn't mind trying to duplicate the results! I'm not interested in this as a "my machine's better than yours" thing, it's simply because I need to know if your laser will engrave as well as mine, since I'm about THIS CLOSE to buying one! Because of the work I do I need a machine that will produce similar results to mine (or better will do too!)

<edit> thought I may as well post the pics here too!

272729272730

[Scott Shepherd]

Kev, that's pretty darn good on the small scale. I'd also be interested to see if the Chinese machines are at that level of quality on things that small. I've not seen it myself, but I've seen numerous threads by Chinese laser owners that have said the detail on really small stuff isn't that great, so I'd love to see that example sample you posted from someone with some of the various Chinese machines.

[Robert Silvers]

I will refine my statement to:


You said in another thread that you would be hard pressed to tell the difference between fiber and C02 for marking aluminum, unless it was a 2 point font. It would be good to see what the difference is with the same mark.

[Scott Shepherd]

Here's a rough one from yesterday. This was the test piece so the speed/power wasn't quite right. Too slow so it's slightly blown out.

Cermark on brass : Font is .030" tall on the capital letters, if I recall correctly. The production pieces are sharper than this one, but it shows an example.

272731

[Robert Silvers]

I need the typeface and the units you used for size.

[Robert Silvers]

Nice wallet.

[Kev Williams]

Robert, font is just Arial regular, NOT bold, and the .12 high is 12 point, .01 high was 10 point, etc, the sizes in inches come to within .002" of the point size-

Scott, that's some nice engraving-- what lens you get that with? 2" on mine--

Oh yeah, settings-- 800 dpi both ways, 40% power and 30% speed I think it was--

[Robert Silvers]

There are 72 points to an inch.

So 12 points is 0.167 inches.

2.5 points is 0.035 inches.

Just type into google: "12 points in inches", and it will tell you.

I will print one tonight in points instead of inches so that it will match yours, as soon as my current run ends.

BTW, my software does a horrendous job at spacing characters. I just put this into the software, and it had too much space. I will try it in Illustrator and convert to DXF.

[Robert Silvers]

I made the engraving. It was useful as it helped me work out my work chain. This is the first time I tried to create a file in AI and import it. I can say that I was not able to import from the Mac version of AI either from AI, EPS, or DXF format even trying just about every option of using older versions of the file format and creating font outlines. I was able to save the file, load it into the PC version of AI, and the software could read those files.

I am photographing the result now.

[Kev Williams]

There's points, and then there's points-- all I know is Corel's points are nearly equal to .010". Below is a screenshot from Corel, the "XYZ" text is grouped, Corel shows actual text height is .199", and text point size is 20- It is what it is!

Regardless of that, I'm still very interested to see if you can get the same or close to the same quality of lettering with your machine! - it doesn't really matter on what material, anything handy-- I'll know if I like it when I see it--

I translated the corel job into actual size in MM, from top to bottom the MM sizes are:
.12 = 3.32mm
.10 = 2.76mm
.08 = 2.21mm
.06 = 1.66mm
.05 = 1.38mm
.04 = 1.11mm
.03 = 0.83mm
.025 = 0.65mm

272739

[Robert Silvers]

http://imageshack.com/a/img855/9616/guf7.jpg

[Rodne Gold]

When I got my shenhuis , I compared them to my GCC spirts and explorers , which can do fine text real well
My 60w cheapy tubed shenhui did as well as any of my other lasers , my 80w Reci tubed was not as good on very small lettering.
Considering the price , I was very impressed.
The chinese lasers actually need to be set up for backlash compensation at various speeds to get the absolute best out of them , some mainstream lasers also have this , tho it's called tuning or something like that.
Thruput on my chinese lasers is around 1/2-2/3rds of my fastest spirits (which are super quick)

[Robert Silvers]

Note that 4 points on mine is actually 0.056 inches (there are 72 points to an inch in the real world. Not sure what Corel is doing, but 20 points is always 0.278 inches).

5 points is 0.069" - about 1/16.

I did this at 200 mm/sec. Maybe I should do one at 50 mm/sec and see if that is any better.

I can email you my file in AI, EPS, or DXF if you want to output it. on yours.

[Robert Silvers]

Here it is at 100mm/sec and 15% power. Both of these were done at 600 dpi scanning. I did a 300dpi one (0.085mm scan pass) and it was not as good - at least under my Nikon stereo microscope.

http://imageshack.com/a/img51/9599/z3h2.jpg

[Robert Silvers]

You asked about how 0.050" would look. That is 3.6 points, so imagine 1/2 way between my 3 and 4. I would not call it crisp, but you may decide it is ok given the low cost of the C02 system. The letters are smaller than the LIBERTY on a dime and about as crisp as the dime.

The 2.5 point is 0.88mm or 0.035" looks crisp to my bare eyes with no magnification, which to me makes it usable.

[Robert Silvers]

Just an iPhone photo in the kitchen so far of a backlit mirror. I used a 300dpi / 0.085 pass and a Photoshop circular halftone at about a 50 screen. I also tried diffusion dither and it gave higher resolution but poorer grey scales. I believe the laser just cannot burn individual pixels very well, so maybe if I did it again with a 300 dpi dither pattern but output it at 600dpi I could make it improve the tonal values. My PHCad software will import photos and then also make a halftone screen on its own.

http://imageshack.com/scaled/1600x1200/593/vd3x.jpg

[Dan Hintz]

I will refine my statement to:


You said in another thread that you would be hard pressed to tell the difference between fiber and C02 for marking aluminum, unless it was a 2 point font. It would be good to see what the difference is with the same mark.


This is what I said:

Yes, you can certainly get finer detail with a YAG/fiber, but most would be hard-pressed to know the difference unless they were looking for 2-pt fonts.
I specifically pointed out that more detail was certainly possible with a fiber system. My point was most customers aren't bringing in microscopes to check your work, nor are they often bringing in projects that require that level of detail. And as I mentioned before, if they do, you move to the fiber system, as appropriate.

[Dave Sheldrake]

The question of RF excited Vs DC excited is one of physics fact and not opinion or comparison.

A good DC will be better than a bad RF, a good RF will be better than a good DC. The beam mode is what matters and DC excited tubes don't have good beam modes. Having a poor RF that is out performed by a very good DC proves nothing.

Spherical aberration makes a big difference and the larger input beam of a DC tube WILL cause aberrations.

AS Dan rightly says Fiber / YAG are much shorter wavelengths ergo they can produce smaller spots (by a factor of 10) but unless that is a requirement the point is moot.

I have Fibers, DC , RF , YAG and Flow lasers, all do different jobs, all do what I need them for very well. the fact the Mitsubishi is better than the ULS is a given, it should be it cost 10X as much as the ULS and nearly 100x as much as my Chinese machines.


Yes, you can certainly get finer detail with a YAG/fiber, but most would be hard-pressed to know the difference unless they were looking for 2-pt fonts.

This isn't only Dan's opinion, it's a statement of fact of optical physics.

cheers

Dave

[Robert Silvers]

It would be nice to see this fiber marked to see exactly how much brighter and sharper it is.

[Scott Shepherd]

Looks like serious banding in that photo of the girl. On the right hand side, it looks like diagonal banding next to her head. Looks like horizontal banding on her left shoulder and working down (right hand shoulder if looking at the photo).

[Robert Silvers]

Half-tone patterns should be a great test to show banding. This is the kind of thing we can discover from a comparison.

[Kev Williams]

Robert, many thanks for the test engravings! What material did you engrave? It's obvious your laser's spot size is larger than mine, however, I'm actually very impressed with the results of the smaller engraving! The overall quality of the text is great in relation to the dime. For 90% of my engraving needs, what I see in your sample would be fine! One of my steady jobs is Cermark-etching large stainless steel operator panels, complete with several 1/8" and 1/4" thick border and sectional outlines, 100 sq inches of bounding-boxed bold logo text, and about 50 words of identification. The bigger panels are 31" tall so I have to rotate and match up the 2 halves of the job. It takes my 40w laser over 3 hours of pure running time to complete one. I could probably cut that time down to a little over an hour with your machine. The larger spot size would likely be a time saver on these panels. And I'm sure a tighter lens could provide a bit smaller beam. And I always have my other two machines...

Doing a little simple math tells me that if it takes 15% power to fire a 100w laser at minimum power, that's roughly the equivalent to a 15 watt laser at FULL power-- if that math is correct, then I'm also interested to know how well your machine engraves typical Rowmark Lasermax or equivalent? My 25 watt Universal etches black/white Romark nicely at 45% power, or at least the 45% power SETTING-

Thanks again Robert, much appreciated!

[Robert Silvers]

I engraved black anodized aluminum.

I will drive to a Rowmark distributor on Monday if they are open, or Tuesday.

[Robert Silvers]

One of my steady jobs is Cermark-etching large stainless steel operator panels

When I said that fiber would be better for aluminum, I didn't know you meant Cermark, and I didn't know you meant large. Most fiber are limited to a small imaging area, though they also make them with travel, but I don't think they can do continuous large pieces.

Yes, this seems good for you. The quality exceeds a typical laser printer.

[Dave Sheldrake]

Fiber is the method of delivery of a 1064nm source to the coupler, Galvo/Gantry is the method of delivery to the workpiece.

Fibers can be had up to 3m x 2m off the shelf with sizes up to 8m x 3m to order in gantry format with 1200mm x 1200mm not that unusual for makes like Vytek Galvo's.

Remember Kev western lasers are usually servo or high quality steppers, they accelerate a lot faster and have better profile beams (if RF) so the benefits of a bigger bed may not be as much as you think when comparing western to Chinese.

I remember doing a test a little while ago that Scott took part in comparing engraving an A4 sheet on a Western Vs Chinese, the Western was about 3 times faster. The bigger the bed the more mass has to be moved, speed is limited by gantry weight / stepper power.

If you are used to using a ULS machine you will likely be disappointed with a Chinese equivalent :(

cheers

Dave

[Chuck Stone]

Note that 4 points on mine is actually 0.056 inches (there are 72 points to an inch in the real world. Not sure what Corel is doing, but 20 points is always 0.278 inches)..

Not really. 72 points per inch is a PostScript standard.
Traditional typesetting is 72.27 points per inch.

[Robert Silvers]

In any case, we know that Corel showing 100.5 point to an inch cannot be correct.

[Scott Shepherd]

Here's some smaller text on anodized aluminum. I didn't do anything but run it with the factory settings from the database. Run on a 5 year old Universal laser. The text is a little bold. There are settings to correct that and I can, but I didn't want to spend a lot of time on it.

The last line is a 1.5pt font. Also lost a little of the photo quality in resizing it to get it small enough to post. It's a little sharper than it looks in the photo. I shot it holding the camera by hand.

272748

[Robert Silvers]

What is different about the Universal that gives that quality?

Beam diameter?

Lens?

RF?

[Scott Shepherd]

What is different about the Universal that gives that quality?

Beam diameter?

Lens?

RF?

I suspect a lot of things. The quality of the beam, the ability to fine tune the details (although I didn't fine tune anything here- it should be able to be a little better if I tuned it), higher quality parts, etc. It's just a well engineered machine and well built.

I haven't tried it on the Trotec yet, it was running a job when I ran these tests.

I might add, that's a 2.0 lens on that 1.5pt font.

[Dave Sheldrake]

Beam profile, polarisation,lens,source,rigidity,pulse rate.

cheers

Dave

[Scott Shepherd]

Just ran it on the Trotec and had very similar results to the above photo. I've not been into the tuning side of the Trotec much, but with a little tuning, I suspect I could have a very well defined 1.5pt font. It was very good with the factory tuning set. On the Universal, you can openly tune and tweak every job and material. On the Trotec, the tuning settings are more in the machine side of things than the job and material side, so the settings are deeper in the controls of the "settings" part of the software. I'd rather not tweak anything at this point as we've got a number of repeat jobs going on at the moment and last thing I need to do is screw it up :)

[Robert Silvers]

I knew the US machines would have better software, more safety features, etc. But now I have a sense for the difference in engraving quality, and it is more than I expected.

[Dave Sheldrake]

That's just the surface details Rob,

On the optical train / source you also have,
Beam propagation ratio
Divergence angle
Waist Location
Rayleigh range
Ellipticity
PSF
Wavefront
Zernike aberration modes

From there on it starts to get complicated.(Gas Pump State ratios are a mind numbing set of figures before you even consider the optical train itself)

Contrary to common belief Laser beams are not all equal, even when they are emitted by the same type of source in the same wavelength. Western quality control is magnitudes of order above anything from the far east.

Take the Z4 you have, even a simple change of the final lens for a different one from the same batch will change the way the machine cuts. Change between Plano - Convex and Meniscus of the same focal length will change cut rates, kerf size and focal position as well as power density.

Lasers are an exact science, the machines they go into ...aren't.

cheers

Dave

[Robert Silvers]

I could probably fix or reduce the banding by picking a step gap that is evenly divisible with the steps per mm of the gantry. I suspect that error accumulates and then catches up. It is visible because of the line screen pattern.

[Scott Shepherd]

Just for clarity, here's what it looks like normal size. It's easy to think they look bad when zoomed in, but in real life, they look quite readable. Saving this photo to upload made it lose some of it's quality.
272770

[Chuck Stone]

"Lasers are an exact science, the machines they go into ...aren't."


Now that's quotable..

[Robert Silvers]

I don't think I have ever used a font less than 6 points.

[Scott Shepherd]

I don't think I have ever used a font less than 6 points.

You've only had your machine a week and a half, haven't you? Give it time, you'll see some text jobs come your way. I see them a lot. I've probably engraved several hundred thousand items with small text on them over the years. I like them. They take so little time to engrave and they pricing is good on those small item jobs with high quantities.

[Robert Silvers]

I meant on anything ever in my life with a laser printer.

[Kev Williams]

I just had to see how low my 10 year old LS900 could go, with upper and lower case-- Kinda surprised me!

I used a piece of trophy material, font I used is Swiss 721 Thin.
I also cheated a bit and spaced the characters 25% more than default.
Point sizes entered in Corel are 4, 3, 2.5 and 2 --
I did a couple of tries and ended up with 38 power/12 speed, 1200/1200 dpi

I measured (Corel measured) the cap T and the lower e on each line-

In inches, T/e top to bottom:
.04/.03 -- .03/.023 -- .025/.019 -- .02/.015

In mm:
1.015/.768 -- .761/.576 -- .634/.48 -- .507/.384

Not that I've never needed this small... :)


272851

[Scott Shepherd]

It's not about "needing" text that small. The small text is just a demonstration for the quality in detailed items. Forget text and assume it's a ornate filigree or just a simple thin line that's a border for something. We've done a fair amount of fire evacuation maps for commercial buildings. They consist of line drawings of each floor with escape paths in case of fire. They are composed of nothing but fine lines. Typically .020"-.030" thick. What these examples show is how sharp you can expect things like that to be. That's a selling point of our business. You business may be something else where it doesn't matter, so it might not apply, but this has nothing to do with small text, it has everything thing to do with holding the smallest of detail so the end product is as clear and sharp as it can be.

[Robert Silvers]

I tuned up the machine, and this is realistically the best it can do - maybe a hair better because rather than set the focus dead center of the optimal, I biased it inward about 0.5mm so that it will never focus above the surface.

I got my minimum spot size down to about 0.006.

http://imageshack.com/a/img17/812/u1et.jpg

[Dave Sheldrake]

In inches, T/e top to bottom:
.04/.03 -- .03/.023 -- .025/.019 -- .02/.015

In mm:
1.015/.768 -- .761/.576 -- .634/.48 -- .507/.384



For a 10 year old machine Kev that's tidy, nice bit of work at sizes that small.

Much as I like Chinese lasers it's unlikely a DC tubed gantry machine will match that quality.

cheers

Dave

[Robert Silvers]

Yes, that is the kind of proof I need about some of what I am giving up vs a better machine.

Is that machine galvo or gantry?

[Dave Sheldrake]

Gantry I'd bet, Galvos weren't in the public domain as much 10 years ago.

cheers

Dave

[matthew knott]

Galvos wont allow smaller engraving, if anything its the other way round, i will try some small letters on a fiber tomorrow, the small spot size due to the laser should allow me to beat anything posted thus fair. A gantry fibre would allowing even smaller marking as you can use a shorter focal length lens!

[Scott Shepherd]

i will try some small letters on a fiber tomorrow, the small spot size due to the laser should allow me to beat anything posted thus fair.

CHEATER! :p

[matthew knott]

As Dan pointed out is all becomes pointless as you need a microscope to see it after a while, but it will give you a comparison.