View Full Version : Newbie: Chinese machines that cut perpendicular to the surface of the material?

john banks
11-04-2011, 1:02 PM
I hope as a n00b to this forum I'm posting in the right section although it says "Engravers", apologies if not, as it is a question about vector cutting materials such as hardwoods or laser ply up to 0.2 inch thickness with Chinese glass tube lasers of say 80W, but considerations for machines are in the 50-150W range. This forum seems to be the place for good tech input from all the posts which I've spent hours reading and appreciating - thanks all.

What we found on our demo of a Chinese machine (and also an Epilog) is that cuts through wood are not always perpendicular with some being up to 10-15 degrees out which can be obvious to the naked eye and would personally make me as a purchaser of a product produced on it think it was poor quality. If you cut a gear or a circle it might not be able to be refitted in orientations different to the original one cut. This is a problem to us as we want to make some "kinetic art" pieces and sell them for a premium price. One machine made by Cadcam (FB 700) cuts perpendicular all the time, but the price is similar to Epilog and a bit more than we want to pay in a flat economy with a new startup business run only by my wife with me offering her tech support (whilst new to lasers, I am reasonably competent as an electronic engineer and have reverse engineered the Nissan GTR engine computer for the leading supplier of tuning tools for this car, and will enjoy the challenge in my spare time away from my other work of keeping a machine nicely aligned and running well for her to do her art on and she can do the design and marketing, I'd love to build my own machine if I had the time and the CNC control would not phase me I think, nor the wiring).

Anyway, is it wobble or flex that means the beam direction through the final mirror/lens is inconsistent? Cuts just 0.25 inch apart are at different angles, so how could it be static alignment issues? Given that the Epilog demos also didn't cut consistently through fairly thin wood, I don't think it is a Chinese glass tube vs American metal tube issue either. Sometimes I wonder though how much attention is made to alignment by sales people when they demonstrate machines and move them between locations regularly. Maybe they don't bother if it engraves OK?

http://www.ctemag.com/dynamic.articles.php?id=213 is interesting reading...

Joe Hillmann
11-04-2011, 1:42 PM
The reason it cuts at an angle is because the cut is wider at the top then the bottom. A 2 inch lens in theory should have slightly less slant then a 1.5 inch lens. Some materials are worse than others. Plexiglass doesn't seem to be cut on an angle at all but that is probably because the edges get flame polished when it is cut. 1/4 inch birch or luan which are actually .207" thick appear to be cut at right angles when looked at with the naked eye but if you stack to identical parts you will find that it has a slant of about .002" from the bottom to the top. 1/8 inch hard board is the worst It has a very obvious slant that can easly be seen with the naked eye. I think the hardboard is so bad because it takes so much power to cut and the top keeps being burned away and the bottom is only barely cut through.

As far as cutting gears I just recently built an orrery (model of the solar system) that had gears stacked 3 high in places and the slant of the cuts on the 1/4 birch didn't cause any problem but I did have to go through with 220 sandpaper and give one stroke to each side of each tooth to get them to turn smoothly(the probably would have worked as the were out of the machine, without sanding, if I had used metal bushings instead of wood bushings).

If you have your doubts about whether a laser will do what you need why don't you create one of your sculptures in corel draw and find someone who has a laser to cut it out for you. That way you will know if it will work, how much work it takes to finish the piece and how long it takes to cut the parts.

Gary Hair
11-04-2011, 1:46 PM
If the cut is angled then the alignment of the beam to the lens is off. That's not a sign of a poorly manufactured machine but definately a sign of one that isn't adjust properly. Any laser can be adjusted to cut perpendicular to the substrate but sometimes it take a bit of adjustment to get it right - then you have to make sure it's right in all four quandrants of the bed, not an easy task if you don't have the initial beam alignment done properly. My laser has 4 mirrors before the lens and if any of them are off a bit then all bets are off as to getting a perpendicular cut.


Rodne Gold
11-04-2011, 2:08 PM
Apart from alignment which is most likrly out , even if perfectly aligned , the laser beam is not shaped like this || , it is shaped like this X - you focus on the top of the material with the waist (narrowest part) of the X , the X flattens with shorter focal lens and elongates with longer ones . On most materials the cut follows the shape of the ---X---- - the more elongaded the X is , the more like a straight cut you get. The other issue is that the beam gets weaker as it approaches the bottom of the X , it often happens that due to the weakness it can no longer vaporise and it starts creating a lot of heat affected zone (char , melt etc) alongside the cut , also leading to a slanty cut. Acrylic acts as a wave guide , so the beam actually DOES appear like || to it.

Michael Hunter
11-04-2011, 2:27 PM
With a typical 2" focal length lens cutting 0.2" thick material on a machine that has been properly aligned, you should have to look quite hard to see that the sides of the cut are not quite perpendicular to the top and bottom.

If the sides of the cut are very obviously out, it is because of mis-alignment, not lens characteristics.
This happens when the laser beam is offset from the centre of the lens, making the focused beam come out at an angle.
If the machine has a red-spot facility that is well aligned to the laser beam itself, then it is normally fairly simple to adjust the mirrors so that the beam hits the lens exactly in the middle.
On some Chinese machines the red-spot does not follow the laser beam, but comes in below the focus lens - in this case proper beam alignment can be time-consuming (but still possible to do well).

Dan Hintz
11-04-2011, 2:49 PM
If the angled kerf is a problem in your application, have you considered a CNC machine? You're limited a bit more on how tight of an inner corner you can create, but outer corners will be as perfect as with a laser... and with a CNC, the kerf will be completely vertical. As the wood gets thicker, you'll be more effective (i.e., faster) with a CNC compared to a laser.

Something to consider...

Richard Rumancik
11-05-2011, 12:39 PM
Thanks for the link, John.

You probably can't get rid of the angle completely, but 10-15 degrees seems like way too much. Did you actually measure the angle with a machinists protractor? Some responses suggest that you can get perpendicular sides but I guess it is a matter of what kind of accuracy you need. I don't think I can get better than a couple degrees for wood or acrylic. If you are making acrylic ornaments the angle cut is pretty much irrelevant but if you are cutting gears to mesh it is a different story.

When I cut acrylic it usually has some some angle to it. It might not be visible just looking at an edge, but you will notice it if you construct a box of acrylic and try to bond the joints. You will find that the edges are not perfectly perpendicular and the bond gaps are wedge-shaped.

You need to ensure that the optics are aligned. Do the circle-test and see if the disk is "tilted" all the way around. If it is, then your beam is not hitting the last lens in the center. If it looks more like a section of a cone (smaller diameter on the bottom) then this is normal and all you can do is play with focus (i.e. try focusing into the part) or you can try different focal length lenses. A longer FL lens will reduce angularity at the expense of spot size and intensity of the spot. If you have enough power you might be able to use this approach.

If the angle is random (not the same on every left edge, for example) then you probably have a mechanical problem.

If your disk looks like the one in Figure 3a in the link, then you have a polarization problem. Normally the lierature suggests that with organic materials you won't see it. But I always had a hunch that polarization was an issue in low power lasers with organic materials. Then I came across an article that suggested that acrylic was susceptable to polarization problems. The lasers below 50 watts are usually linearly polarized. If a laser system has two tubes, like some ULS systems, or if there are two tubes internally (like some higher-power Synrad products) then you will get circular polarization or a close-enough approximation. You can also get additional optics that can generate a circular polarized beam from linear (using quarter-wave plates) but I don't know of any small systems that have this as OEM. That's because small laser systems were designed for making "art" not high-tech parts. Those of us that are trying to get more precision and accuracy from our lasers have to deal with the limits, or find work-arounds.

I would suggest that you cut some squares and circles (be sure to mark them so you know the orientation on the bed when cut) and analyze the perpendicularity of the edges so you can come up with some theories and an approach to solve it.

Do you know if the glass tubes on the Chinese lasers are circular or linearly polarized?

john banks
11-05-2011, 6:50 PM
Thanks all. Dan we are considering following the laser with a CNC router and combining both techniques were useful, but the noise puts us off slightly, plus it is a steeper learning curve for my wife to operate, but will be much better for removal of materal, larger items, kerf angle, no charring (hopefully!) Regarding the noise, we have put sound absorbing dry wall and insulation in the workshop that we've renovated from old stables, but there is just one house nearby so we need to not be too noisy. So the laser will have a quiet compressor. I suppose we could make a compartment for the router.

Richard, I'll add few image rather quickly snapped now. This is 3mm laser ply. 1 is from above, 2 is from above angled so you can see through the "7" area (it is a mirrored 7), 3 shows the angle to the horizontal required to see through the "7" from above. I'm looking for my protractor, but the angle is not small.212063212064212065

Of note the FB700 had two 50W lasers, and on some of our more ambitious cuts on thick hardwood it was in 100W mode and it was remarkable, but would cost near to 40000. Your explanation of the polarization is most interesting in this respect.

Rich Harman
11-05-2011, 7:53 PM
I used a machinist square against the table to check the alignment of the lens holder assembly on my machine. It was off quite a bit both fore and aft and side to side, it is fairly easy to adjust.

To check that the beam is centered through the lens, lower the table a few inches from focused and test fire the laser onto a piece of thermal paper (receipts are often made of thermal paper). If the beam is centered it will make an evenly filled circular mark on the paper.

George M. Perzel
11-06-2011, 6:22 AM
Hi John;
Welcome to the SMC. I noticed on the pics you sent that the objects you cut have severe smoke damage and wonder if part of your problem may be the power, speed, and ppi settings you are using. It appears as if you are burning thru the material rather than cutting it cleanly. Try faster speed and ppi not more than 150/inch.
Best Regards,

Kay Bengtson
11-06-2011, 12:27 PM
I cut balsa wood up to 1/4" thick and 3mm poplar plywood. The side wall profile is visually perpendicular when I focus halfway into the material.


Richard Rumancik
11-07-2011, 10:20 AM
John, if I understand correctly you have to tilt the shape quite significantly to be able to view through the "7" cutout (such that it's walls are parallel to the eye).

That would suggest a significant error but I don't see a corresponding error on the outer cut line. This is puzzling.

I would suggest that you do your tests with simpler shapes such as rectangles and circles. Using complex shapes for trials can be confusing.

One way to get a better visual indication of the angular error (and provides a visual demonstration to the manufacturer) is to make a rectangle with a hole in it (let's say .150" diameter). Then get a set of numbered and fractional drill bits that you can use as a poor-man's gage pins. Find a bit that is a snug fit with the hole and insert the shank. With the plate horizonatal you can get a qualitative idea of the angle of the hole. If no bit fits snuggly, adjust the cut circle diameter in your file by .0005 or .0010 or whatever till you get a snug fit (or else make several holes in the rectangle stepped .001").

I really can't understand how an internal feature could be skewed without a corresponding skew on an external line.

Rich Harman
11-08-2011, 12:22 AM
If the outside were cut before the internal features it could be skewed due to the piece tilting after it has been freed from the material.

Richard Rumancik
11-08-2011, 11:17 AM
Rich, that's an possibility. Never thought of that . . . but you are correct; if the material and/or the material were not flat, and there was a gap below the shape, it could fall out and rest at an angle. Then when the "7" was cut it would cut on a different axis.

John, is that a possibility? Are you cutting ouside first? How repeatable is the problem you are seeing?

(As an aside, it is usually better to cut all internal features before cutting the perimeter for a number of reasons. When a shape falls out it can twist, move slightly laterally, change focus, etc. This is more likely a problem on things like baltic birch which are not always flat to start with.)

john banks
11-08-2011, 11:34 AM
Thanks for the further thoughts. The smoke damage was due to poor air assist as well as perhaps excessive power for this material. We had visited the company for a demo to see how it would handle thicker wood. At my request (which was a bit odd for me to make since I had only seen my first laser machine the day before and it cut better with far more air assist), some other trials were using much more powerful air assist and the smoke damage was reduced. Other trials with or without the more powerful air assist produced circles that were not circular. They would not fit well into the original material except in the original orientation, and I could only see apparent randomness to the kerf angles, hence my post here as we also saw the same with an Epilog cutting wood. Since we were just cutting simple shapes and the parts didn't visibly move I don't think that is the explanation for the chaotic kerf angles. I shall post some photos of simple shapes from the Chinese and the Epilog machine...

I should also mention that after the kerf angles were produced at our Epilog demo the salesperson went away to their tech support and mailed some more samples which were better - but this is the one I've photographed below. Some were worse.

john banks
11-08-2011, 11:47 AM
Here are two photos which show interesting kerf angles. I haven't bought a machine yet, I didn't want to be stuck with these sorts of results, but if it is a case of setup then I'm happy to put in the attention to detail to get it right, but don't want to spend on a machine that can't be setup to give quality straight cuts, hence this thread to find out if it is all alignment/setup, instead of a Chinese machine thing which I thought it was until I saw angles from the Epilog. Sorry the photos are upside down, you are looking at the side view of the laser cut circle and rectangle in the middle of the photos, not the lamp base in the left of the photo!