Possible causes of fuzzy engraving

[Darren Wilson]

Hello All,

Please see photos, this was done at 60P/45S 400dpi on 36EXT 75W

323724323725323726

Known issues:

Tube is weak as I struggle to cut .125" acrylic in one pass, @ 100P/8S/5000F
Encoder strip and reader are new
Servos/encoders have been bench tested and are functioning normally
Motherboard is new, and did not have any original settings saved.


I have a new tube en-route, but would like to now if there is anything else that could be causing this, such as laser match settings or any calibration? The belts and cogs are clean and in good nick, no cracks or build up. I have loosened tension bolt cycled by hand and tightened.

Thanks in advance

[Mike Null]

Darren

That could be caused by a number of things. I would first make sure that the material is secure on the table or grid so that it does not move during the engraving.

On my machine there are adjustments I can make to the motion system through Job Control which can fix that issue assuming that the motion system is in good order. You may have the same type of settings available to you. Check tech support. It is possible that bearings need to be replaced and a thorough cleaning of your motion system would help.

Are your mirrors and lens secure?

Try running the job at half the speed and power and see what your results look like.

[Rich Harman]

Can you set your machine so that the engraving only occurs on the left-to-right pass, instead of bi-directional? What you show there looks just like when the bi-directional settings are off on my machine.

[Darren Wilson]

Hi Mike,
Material wan definitely not moving around. If I do a second pass it fills the gaps so to say bit edge is still not smooth. Will double check the linear rails, bearings and belts, as well as put in call ro tech support. With MB change i lost any values for lase match which i think controls timing on on -x, +x firing.
Will run one at half speed power and post results.
Cheers

[Darren Wilson]

Hi Rich,
Not sure if my Epilog has that function.
Anyone out there know how/if epilog will fire in described mode?

[Kev Williams]

If you know the tube is weak, I might blame the tube for what I see, those edges could be a symptom of the tube being slow/late to fire on the leading edges.

Could be an X left/right alignment issue, but usually that leads to fatter, 'double-vision' type text.

Could be a loose belt or something other mechanical, loose lens holder maybe?

I'm still leaning toward the tube is firing late...

[Bruce Volden]

Darren,

Make certain you don't have "crap" building up in the peaks and valleys of your belts and pulleys!
This will cause all kinds of "fuzzing".
Since you stated the encoder strip is clean, ------how about the reader? (the little black thingy the strip goes through).?
I'm pretty certain you're dealing with a mechanical problem.

Bruce

[Darren Wilson]

Hello Kev/Bruce
Belts and gears on servos are clean, no build up and the belts seem to be tensioned enough. The encoder and reader are both new, and have been cleaned.
I was thinking the same with the tube not ramping up fast enough, will know for sure in a few days.
There also seems to be a slight misalignment of vector if I add outline to the text, but it is hard to tell due to the fuzziness. Will run a few more at different speeds and power when I get home and post.
Cheers,
Darren

[Samuel Espy]

Darren, I would look into your "tickle" settings for your tube. I have a ULS machine and it has a tickle test you can run to determine if the tickle settings are optimal. You will have to research this setting for Epilogs. On RF laser tubes, tickle is essentially a low-duty secondary modulation signal when the laser is not firing to keep the plasma slightly excited to improve response time for when you do need it to fire. If the tickle is too low, your response time is slow (laser fires late) so that when it actually does fire the motion system has already moved farther in X than where you wanted it to fire. I strongly suspect Epilog has some sort of built-in tickle test. If not you can draw 5-10 parallel vertical lines that are between say 0.010" to 0.020" thick and maybe 0.030" between lines. Engrave these and what you want to look for is a tickle setting which makes the engraved vertical lines have a minimum width (in x). Hope that helps.

[Darren Wilson]

Hi Sam,
There is definitely an adjustment for tickle. As i had to change my MB i dont know what original setting was. I am not aware of a tickele test and just thought this adjustment was to help laser fire after sittling idle. I also have issue with this as first run of day or after a while between jobs there is a delay in laser firing. I usually have to manually fire through align mode to get it perked up.
Cheers
Daz

[Dan Hintz]

This is a relatively easy one to figure out. Vector a square spiral at a slow speed, like this:
323829

If the jagged marks appear at the corners, it's mechanical (belts, pulleys, loose optics). If not, or the jaggedness appears elsewhere, it's the tube/power supply.

[Mike Null]

I believe the issue is more likely to occur in a raster mode and may be fixable with some software adjustments. If it occurs in a vector mode then I'd lean toward a mechanical issue.

Dan, you can replicate this problem on your Trotec by changing some settings in your Job Control software.

[Michael Hunter]

I think Kev Williams is probably right - it is likely to be the tube.

For the time being (until the new tube arrives) try upping the resolution to 600 and halving the speed (with the power adjusted to suit).

[Darren Wilson]

Hi Guys,
Sorry I didn't get a chance to engrave anything yesterday, I did check for looseness of mirrors and lens assy, all nice and snug. Hopefully will get a chance tonight to run Dan's pattern and another pass at higher res, lower power/speed.
cheers,
Daz

[Dan Hintz]

Hopefully will get a chance tonight to run Dan's pattern and another pass at higher res, lower power/speed.

Remember, vector not raster... resolution doesn't matter. I'm trying to determine if there is slop in the mechanicals that are leading to your non-rigid lines. If it exists, you'll see it every time the machine changes direction. Going slow removes a weak tube from the test.

[Rich Harman]

I don't see how a spiral like that will diagnose slop, especially at a slow speed. For example, if there were lots of backlash it would not be evident since you are not joining up a line that is approached from separate directions. I would think a circle would be better for that, or a series of circles - each starting at a different quadrant.

[Kev Williams]

Here's an enlargement of some of the engraving, and I've adjusted the contrast--

IMO, again, this is not an alignment or mechanical problem. Alignment problems will have runout like this on each opposite pass, but the each half of the runout will be lined up. This runout is all over the place. This might indicate a mechanical problem, except that if it was mechanical, the runout would be pretty much the same on each horizontal line where the laser fires, and that would result in 'wobbly' looking engraving...

The runout is totally ragged, and the engraving isn't wobbly- the "i" is straight, the curves are consistent. The only wobble apparent at all is the top of the "d" appears a little bent.

Finally, the white that shows up by lightening up the contrast, means the white is there, but I'm not sure if that means anything more than more power could've been used.

It looks like not only is the laser firing late, it also seems to be turning off late too. Might be some tickle or other adjustments to be made, but I'm still leaning towards the tube is the main problem...




http://www.engraver1.com/erase2/fuzzy.jpg

[Darren Wilson]

Ok got a chance to pay around first Mike's test:
323911
Next Samuel's:
323912

[Darren Wilson]

Some others:
323913
slower diff tickle settings
323914
40s60p600dp1 TM1TI0
323915
a second pass
323916
20s25p600dpi TM1TI0
323917
vector pass

[Kev Williams]

Your "squaral" tests- the 10 power looks pretty consistent, but the 20 power is pretty heavy on the left side- material bowing up out of focus? Aside from that, they look perfect...

Your tickle tests look horrible, but then I'm not sure what they're supposed to look like, however, I would expect the edges to be a lot less raggely!

[Scott Marquez]

You have done some great tests, what does Epilog think it is?
Scott

[Darren Wilson]

Yeah the Sq vector test at 20P cut thru so material is slightly warped, that's why I ran again @ 10P. The tickle test are pretty average, can't really pick one better that the other, to my eyes TM1/TI0 looks best but not by much. I haven't had a chance to get a call into Epilog, due to time difference and current project at day job. Hope to get a chance early this coming week. I am really hoping that a new tube will fix this up.
Daz

[Scott Shepherd]

Have you reloaded the driver Darren? To me, your tests prove it's not mechanical. I also don't agree that it's a tickle setting. I've never seen tickle cause mismatch in left to right engraving. My guess is it's software related, but that's just a guess.

I'd personally try reloading all the software, deleting the existing and starting over and see if that fixes it. That might not be it, but it's where I'd start.

[Darren Wilson]

Hi Scott,
Reloading the driver is easy enough, I will give that a shot.
Has anyone played with the "Laser Match" settings? As I had to replace my MB whatever the factory settings were they have been lost.

"Calib X encoder" is what I'm afraid to run as this is how I blew my original MB... I damaged my linear reader while cleaning, so there was not a valid signal from the encoder strip... this somehow destroyed the MB. I have steered clear of this function since, as I don't have $2k lying around to replace the MB again. So since replacing the MB, Linear Strip and Reader, it has not been recalibrated. Has anyone successfully run "Calib X Encoder" and if so what precautions were taken?

Cheers Daz

[Darren Wilson]

Finally should have some time to reload drivers and re-network engraver. My replacement tube is currently being released from AU Customs, I am desperately trying to get it held for pick up this afternoon, if not it should be at my doorstep by Monday. Then the fun of alignment begins and hopefully crisp engraving, Fingers crossed...

[Phil Vernon]

Darren, I would look into your "tickle" settings for your tube. I have a ULS machine and it has a tickle test you can run to determine if the tickle settings are optimal. You will have to research this setting for Epilogs. On RF laser tubes, tickle is essentially a low-duty secondary modulation signal when the laser is not firing to keep the plasma slightly excited to improve response time for when you do need it to fire. If the tickle is too low, your response time is slow (laser fires late) so that when it actually does fire the motion system has already moved farther in X than where you wanted it to fire. I strongly suspect Epilog has some sort of built-in tickle test. If not you can draw 5-10 parallel vertical lines that are between say 0.010" to 0.020" thick and maybe 0.030" between lines. Engrave these and what you want to look for is a tickle setting which makes the engraved vertical lines have a minimum width (in x). Hope that helps.

Agree with this prognosis. Looks like Tickle to me as well.
You should have access to a tickle test.

[Darren Wilson]

Hello All,
I've installed my replacement tube from Evergreen Laser, was quite easy. After aligning I ran couple of test...
324745
as you can see the engraving has vastly improved, this was run at 100S/40P
324746
not perfect though, there is still ever so slight fuzzy on the small text .157" (4mm)
324747
however now there is definitely a mismatch in Vector to Raster. It always seemed that there was a slight mismatch, but this seem to be more exaggerated that before.

As I mentioned earlier, the encoder and reader were changed, but I have not run X-Calib due to it blowing my motherboard last time. Anyone with an Epilog 36EXT ever run this calibration with successful results?

Thanks,
Daz

[Darren Wilson]

Hey where have my post gone?

[Darren Wilson]

Vector/Raster mismatch fixed!

Wow what a difference a fully functioning tube makes!