Scan Offset - not lining up on both sides?

LayzarBeams · February 16, 2025, 5:41pm

Good’ay y’all!

Let me first say I have 100% scoured the forums/Dr. Google and have had no luck finding this exact thing i’m seeing. Your input is greatly appreciated.

I have an Aeon Mira 7 with closed rail gantry. New tube/power supply about 2 months ago

I was experiencing some odd issues that have since been resolved with cleaning the snot out of all the snotty places. What I thought was an easy moving gantry… after cleaning i could send it to the back with an easy push… lol.

My lines are almost straight so I know the belts/gantry are functioning correctly… To the naked eye, they are straight, microscope… you can see some things. My engraves after the pat down have gone back to the insane quality I capture before.

I decided to do a scan offset from 0 again to just start fresh. When looking at the .5LPI tests, they look perfectly aligned. When looking under microscope, I can get them to line up on the right… but the left side(starting side) never lines up… Or I can get them to align on the left, but not the right. Like I said, to the eye- minor shifts under microscope can’t be seen with the eye.

Initial offset doesn’t seem to cover this? What setting can fix this, timing? Would initial power % help fix this at the beginning of the engave line? Before I jump into slowing the accel down on X or Y… I would like to try and remedy what looks like a timing issue first.

Bi-Direction is on.

Am I being to critical and send it? I’d like to try and hone in more detail… I run the machine at 400/317 and pull of some incredible detail for a Co2 machine.

3 pictures below.

Right Side

Left Side

The circle outside of the letters is the size of a dime, thats a 65/9 commercial sewing needle which is slightly larger than your average house sewing needle. I started to see some of the aforementioned issues around this job, however - all of the letters are 100% readable for being that small. Was run at 400/317 - 1 pass.

thanks for taking the time to read!

jkwilborn · February 16, 2025, 6:26pm

Since the Aeon Mira 7 comes in different options for the laser, it’s nice if you are more specific in your profile. Such as a dc or rf excited laser and wattage.

I am lost now… what is a snotty place?

With any mechanical movement between parts, there needs some kind of clearance for it to move, such as a pair of gears. This is termed backlash and is entirely mechanical.

Scan offset is to adjust for backlash on that specific axes. If you can’t adjust it out, then it likely isn’t backlash.

Yours doesn’t look like I’d chase it any further. Other things start to come into the picture… such as the response time of the lps, which you can sometimes see.

I think it looks pretty good as is … if your eye can’t see it, it’s likely not an issue.

Most glass tube lasers (dc excited) have a spot size of about 0.20mm. The best resolution you can expect is 127dpi. You could use shorter focal length lenses and get better resolution.

It’s nice to know what you are entering, such as 400/317, if I didn’t know you had a Mira, then I’d be guessing if it’s 400mm/s or 400mm/m… I have to assume the dpi is 317.

Much like having an end mill that’s 1 inch, the best resolution you can attain is 1dpi. What limits detail is the tool (beam) size.

For a 317 dpi, you need to produce a beam of 0.0801mm. Simple calculation show that 0.20/0.0801 is 2.496. You are lasing the same area almost 2.5 times.

The output can also be related to speed, when it comes to detail. If you want to control the lps, you should keep in mind it’s response time related to your speed.

Most lps have a specified response time of <=1mS (1000mS = 1 S). If you take the worst response time for a value, then at 1000mm/s you can change the laser every mm, so the best dpi you can expect and keep control of the laser, is 25.4dpi.

If you cut the speed in half, you can turn the off/on twice a mm, so the dpi is doubled to 50.8dpi… 250mm/s it doubles again and it’s now 101.6dpi…

My machine has an acceleration value of 45,000mm/s^2 in the controller and has been measured at 1650mm/s, but it clearly outruns the lps at a much slower speed. I usually don’t run over 300mm/s.

If you have other questions, sing out.

LayzarBeams · February 16, 2025, 7:27pm

Jack - thanks for taking the time to respond.

My apologies - Mira 7, 60w. 1,200mm/s max.

Snotty places… like the places your kids snot winds up while you wonder how it got there… AKA engrave dust.

I prefer 10w non detergent oil for the rails, it’s whats used for commercial sewing machines, and air cylinders on screen printing presses. Light weight, no drag. It’s been a game changer as far as movement goes.

I am fully aware of what backlash is, in a past life - Certified Master Auto Tech. I get the backlash to Scan Offset correlation - but what does the actual backlash settings in the controller handle if we are making up for it with offset? How does one test for this to see if there is an issue happening beyond offset? The belts/gantry are square and correctly tightened, too tight and I get stutter seen by the eye. I have read what feels like months worth of posts about this whole topic, but nothing that discusses what i’m seeing specifically. Outside of the lines not lining up on both ends, What i’m seeing via microscope is the First line of engrave(left to right) starts late, finishes, moves to the next line and starts what seems like semi-late(like scoring and then a full burn) as seen from the ends of the engrave line above. Is this normal latency of a glass tube even at 400mm/s? It 100% explains what i’m seeing from a fine tuning standpoint if so.

I appreciate the end mill analogy. Ever seen one fly across a room? lol.

The reason for continuing down the detail path is to try to lower LPI in relation to power. I should have noted I am using a 2" Lens, defocused 1mm to achieve the results @ 317lpi in the last photo. I haven’t played with a 1.5" yet, would this be the better option to retain more detail? I’d prefer to not have to raise LPI and create more run time. Again… I realize that i’m using a Co2 and response time has a major factor, but I 100% feel there is more to be had through simple tweaks?

I have tried everything down to 200mm/s, oddly - best results is 400/317. If I lower the power by say half, and run a second pass - the detail is even better but the vertical lines of say a border become thin where the horizontal lines stay the correct width. IF I run the design at the correct power level and use a single pass - the vertical lines are the correct width. This is what has pointed me towards a potential timing issue, or initial power %. The current settings in the controller for initial power/Hz are 5000hz, 1%. I did bump the 1% to 99 just to see what would happen and the beginning of the engrave(offset test) did become much clearer, but the lines arent lining up still on both sides. It’s either one side or the other, or split the different and be done with it.

last tidbit… If I take the same high detail file and burn it to a flat business card - the detail is incredible, but vertical border lines are still thinner than the horizontal. More power, problem goes away. While I know the surface it flat and super thin via anodizing… I mainly run leatherette on this machine.

all mirrors are brand new, lens is new.

sorry for the word vomit.

jkwilborn · February 16, 2025, 8:42pm

Glass tube or metal tube rf?

I’ve used white grease for as long as I’ve had the machine.

If I understand you, if you can’t adjust it out with the offset adjustment, then you’re stuck with it being something else.

At 400mm/s it will travel 400mm * 1mS (1000/sec) response and you could have as much as 0.4mm before it can lase. This isn’t relative to the tube, it’s an lps issue. The lps has to raise the voltage to the tubes trigger voltage, where it can start lasing, this takes time.

We already spoke about spot size. All lenses have a depth of focus (dof), in photography we call it depth of field, but it’s the same animal. We think of the many photos we see about how a lens focuses to a spot. In reality the lens focus to a particular spot has a dof associated with it. The left one is the main thought, but the right one is more of a reality.

The dof for a 2" lens is approximately 2.75mm. If you’re correctly focused, with a 1 mm change, you will still be within the dof of the lens and have changed nothing.

How can you get better detail from a 2nd pass? It’s the same mill end size, so something else must be occurring.

If you’re running in fill mode, I can see this, but it should be consistent.

My Ruida and all the others I’ve seen have an initial pwm period of 50 microseconds (20kHz). Where is the power percentage you are referning?

We don’t use these words… we use lol

LayzarBeams · February 16, 2025, 10:12pm

As always, enjoy reading the detailed responses.

Whoops… glass tube, 60w

At the bottom will be a picture of the settings in question. The starting frequency is set to 5kHz and running Hz is 20kz

Offset hasn’t changed anything except the ends lining up. The offset from left side to right is strange still. Both sides should technically be the same…

White lith is great, 10w is just preference. Try it sometime, you might be pleasantly surprised at how much less drag there is. Also… much easier to clean the rails when needed. I do realize this is preference.

Vertical lines- yes fill mode.

I follow on the focal distance being +/- 1mm, 1mm out has improved the engraves at lower lpi/higher power. My ramp test shows 7.5ish mm. I defocus to 8.5, I’m aware the 7.5 must be the “very near” focal point when I run my ramp test. My question on this now becomes- if I toss a 1.5” in, my Lpi would have to raise to account for the smaller dot, yes? In turn this raises the run time collectively… not complaining, because what I provide for leatherwork is more about high detail one off’s. If I defocus on the 1.5, with the increase in concentrated power… I should be able to use it like a 2” but with less blow out from higher power/lower lpi(2”)?

LPS… I have multi meter I can test with, this machine doesn’t come with an ammeter. How does one find the correct info to set starting voltage/max voltage?

jkwilborn · February 16, 2025, 11:14pm

When you do a ramp test, you can see and measure both beginning and end of the dof. This usually makes it much easier than using an on-line calculator.

The main idea of using a short lens is for detail, de-focusing loses detail, it seems to be kind of illogical to me.

Don’t know where you’d do this, might want to explain it to us.

Put in a mA meter, not an ampmeter, they are cheap.

Pre-ignition is only an RF metal tube option.

Nothing is written in stone, so to speak. You have to look at it and decide what you like more. Sometimes I overlap scans, but this is generally for the fiber, to reach a specific temperature or effect.

There is nothing saying you can use a smaller interval than spot, but most of us find there isn’t much advantage to doing that with the co2.

ednisley · February 16, 2025, 11:17pm

You’re looking at the turn-on and turn-off delays of the power supply + tube, which are almost symmetrical. Your pictures show some differences in adjacent lines, but overall they look just fine.

Here’s a somewhat gnarlier example of the turn-on and turn-off effects from my 60 W laser:

The start of each line is much pointier than the end, with an occasional extraneous dot appearing off the end where the tube has some trouble shutting off.

The Scan Offset Adjustment assumes symmetrical times and can’t compensate for any differences, so stop when you’ve gotten results Good Enough for naked-eye inspection.

jkwilborn · February 16, 2025, 11:22pm

Forgot to mention this… done with a 5.5W diode on leather.

system · March 18, 2025, 11:22pm

This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.