Cutting foam sheets... with a needle!

Springer

Member
I have a question for the gurus! I have been using the 2.5watt laser on my CNC machine (had to temporarily remove the needler) to burn sayings into some wood disks for the wife. I had done this on a trial basis a few months ago and didn't see any issues other than sorting out the speed and power mix. Now running the actual files, I noticed that the insides of letters (like the center of an o or b) are displaced in the x axis. It doesn't seem to affect the Y or Z axis, but it occurs after moving from one completed loop to the next one, at the time when the laser turns back on. I thought it was always shifting left, but today, saw it sometimes shifts right. The first loop is burnt correctly, and the second appears to be correct in each case, but just shifted left. I took some video today (the vimeo file), and while it is somewhat hard to see, it does show the smooth motion in each loop, and the jerk at the start of the next one. I checked the code and there is no spurious step in it, the visualization is smooth. I thought that maybe i needed more power, so switched from the 2.5amp max power supply that came with the original laser cutter i made the machine from to an AT powersupply based unit. No change. I also thought that maybe the laser wiring was too close to the stepper wiring, sending a pulse, and separated them from the spiral wrapping, also no change. Am I dealing with a bad stepper? the fact that it tracks perfectly in each loop seems to belie that. What d'yall think?

The picture shows how the letter centers are displaced.

 

Attachments

  • 20191205_115828.jpg
    20191205_115828.jpg
    900.3 KB · Views: 0

jhitesma

Some guy in the desert
Mentor
Ok, could be a dumb question. And it could have been answered somewhere in this huge thread. But if 8500 rpm seems to be the magic speed for these needle cutters. Could I get a 2s esc, and a 1700kv motor and run it on 5v ?

I would advise against just assuming the math will work out. When I first started building my needle cutter I had months of frustration because I was trusting the math and ratings but it turned out I was running WAY faster than I thought I was which was leading to my issues. Once I got a tach and was able to dial in a more reasonable RPM all kinds of issues went away.

The big issue is not all ESC's are linear in their throttle response, and not all servo testers are consistent about what percent they output based on their markings. With the setup I was using the math said I should be at about 75% "throttle" but once I got a tach I found out that it was more like 20% to get the RPM I was after. Slight variation in motor Kv may also play a roll there, but overall I found that it's better to measure than assume the ESC is going to output what you think it will at a given throttle setting.
 

rockyboy

Skill Collector
Mentor
I have a question for the gurus! I have been using the 2.5watt laser on my CNC machine (had to temporarily remove the needler) to burn sayings into some wood disks for the wife. I had done this on a trial basis a few months ago and didn't see any issues other than sorting out the speed and power mix. Now running the actual files, I noticed that the insides of letters (like the center of an o or b) are displaced in the x axis. It doesn't seem to affect the Y or Z axis, but it occurs after moving from one completed loop to the next one, at the time when the laser turns back on. I thought it was always shifting left, but today, saw it sometimes shifts right. The first loop is burnt correctly, and the second appears to be correct in each case, but just shifted left. I took some video today (the vimeo file), and while it is somewhat hard to see, it does show the smooth motion in each loop, and the jerk at the start of the next one. I checked the code and there is no spurious step in it, the visualization is smooth. I thought that maybe i needed more power, so switched from the 2.5amp max power supply that came with the original laser cutter i made the machine from to an AT powersupply based unit. No change. I also thought that maybe the laser wiring was too close to the stepper wiring, sending a pulse, and separated them from the spiral wrapping, also no change. Am I dealing with a bad stepper? the fact that it tracks perfectly in each loop seems to belie that. What d'yall think?

The picture shows how the letter centers are displaced.

Maybe need to increase voltage to the steppers slightly - sounds like she's missing a step or two in some scenarios.
 

Springer

Member
It did occur to me that perhaps I should revisit the Vref on that stepper. It is one of the original steppers that came with the eleksmaker laser engraver and I guessed that it was a 2 amp max motor. I'll try going up a little and down a little. I have the stepper that came with my z axis unit, and may swap it in to see if it works as well though I hate to do that if I don't know what is going on.
 

Michael9865

Elite member
It did occur to me that perhaps I should revisit the Vref on that stepper. It is one of the original steppers that came with the eleksmaker laser engraver and I guessed that it was a 2 amp max motor. I'll try going up a little and down a little. I have the stepper that came with my z axis unit, and may swap it in to see if it works as well though I hate to do that if I don't know what is going on.

When had issues I found out that the pot on my stepper driver was in need of re-calibrating. You may want to check that out.
 

dkj4linux

Elite member
Yes cutting out a part completely before doing scores would indeed be problematic I think, but cutting scores while a part is only partially cut out may indeed save tool travel time overall. From what i'm seeing so far in fusion it looks like i'll be able to select the paths I want cut and in what order. So for instance you could cut out the front and back of a wing while leaving the ends, then do the score cuts, then cut the ends, then move on to the next part. I'm just diving into all this so i'm slowly picking up on how fusion works. My experience prior has all been in 123d and 3d printing.

One thing I have a question about is with edwards cnc for instance, why are tool travel movements when the cutter is not down and cutting so slow? On my 3d printers I can setup very fast travel movements and then the machine slows down when its actually laying down filament. Is this not possible with his design/GRBL?
I'm sure there'll be a lot of folks interested in what you come up with... there's a lot of interest in Fusion, nowadays.

Yeah, leaving tabs is the normal way of keeping parts in place with most tooling. In my experience with cutting foam, however, I've rarely seen the need for tabs as there are usually thousands(?) of microscopic threads left, bridging the kerf.... especially with papered foamboard.
 

Jonny360

Junior Member
Yeah, leaving tabs is the normal way of keeping parts in place with most tooling. In my experience with cutting foam, however, I've rarely seen the need for tabs as there are usually thousands(?) of microscopic threads left, bridging the kerf.... especially with papered foamboard.

Not really suggesting tabs as you would normally see, just suggesting that the scores are cut before the parts are completely cut...
 

Guy S.

Well-known member
Not really suggesting tabs as you would normally see, just suggesting that the scores are cut before the parts are completely cut...
Using the methods most people use to program these cutters. It seems doing all the score cuts, then inside, then outside is the easiest fastest way to program. The time spent programming to minimize the non cutting transit time would be much longer than just doing it the easy way.unless you are using cam that will allow you to do it easily. My software does, and I still don’t do it. Although my cnc rapids at 35,000mm/min. If you want to do all the operations on each component using tabs, then removing them and so forth, I would recommend forgetting tabs and using a vacuum table. Just a non expert opinion.
 

Jonny360

Junior Member
Using the methods most people use to program these cutters. It seems doing all the score cuts, then inside, then outside is the easiest fastest way to program. The time spent programming to minimize the non cutting transit time would be much longer than just doing it the easy way.unless you are using cam that will allow you to do it easily. My software does, and I still don’t do it. Although my cnc rapids at 35,000mm/min. If you want to do all the operations on each component using tabs, then removing them and so forth, I would recommend forgetting tabs and using a vacuum table. Just a non expert opinion.

I agree its probably the fastest way to program, but I dont think its the fastest way to cut. The programming only needs to be done once. A little more time spent programming to save time when creating multiple copies seems worthwhile.
 

Guy S.

Well-known member
I agree its probably the fastest way to program, but I dont think its the fastest way to cut. The programming only needs to be done once. A little more time spent programming to save time when creating multiple copies seems worthwhile.
I also like a very light coating of 3M Super 77 on whatever substrate you are cutting on. I put just enough on my vacuum plate to give the foam lateral grip but not enough to make it want to stick down. Really helped me prevent part movement.
 

dkj4linux

Elite member
For you guys new to the thread and discussion... and speaking of vacuum hold-down, you might find this of interest...

A number of us are using a vacuum pad consisting of 3 sheets of DTFB; slots east-west on one sheet, slots north-south on the second sheet, and holes at the intersections on the third sheet, which doubles as the spoilboard. A clean workpiece is then placed on top and since the kerf is narrow, sufficient vacuum is maintained for the duration of the cut. Here it is in action


I'm not sure the Z-adjustment on Edward's TimSav will allow 4 layers of DTFB under the tip of the needle guide yet but, if so... this is a pretty quick way to hold your workpiece flat and secure.
 

Jonny360

Junior Member
Feeling confident in getting my process working after some fooling around today. I'm using adobe illustrator and finding it extremely easy to process the PDF's to DXF. Taking me around 3 minutes to process a sheet of plans and export. The i'm able to import the DXF layers into Fusion 360 and generate tool paths for both the cut and score lines at once. I need to figure out the post process to generate the correct gcode for servo movement for the cut vs score lines. But there are a lot of post processes available and it doesn't appear that difficult to create one manually if needed.

 

dkj4linux

Elite member
I have a question for the gurus! I have been using the 2.5watt laser on my CNC machine (had to temporarily remove the needler) to burn sayings into some wood disks for the wife. I had done this on a trial basis a few months ago and didn't see any issues other than sorting out the speed and power mix. Now running the actual files, I noticed that the insides of letters (like the center of an o or b) are displaced in the x axis. It doesn't seem to affect the Y or Z axis, but it occurs after moving from one completed loop to the next one, at the time when the laser turns back on. I thought it was always shifting left, but today, saw it sometimes shifts right. The first loop is burnt correctly, and the second appears to be correct in each case, but just shifted left. I took some video today (the vimeo file), and while it is somewhat hard to see, it does show the smooth motion in each loop, and the jerk at the start of the next one. I checked the code and there is no spurious step in it, the visualization is smooth. I thought that maybe i needed more power, so switched from the 2.5amp max power supply that came with the original laser cutter i made the machine from to an AT powersupply based unit. No change. I also thought that maybe the laser wiring was too close to the stepper wiring, sending a pulse, and separated them from the spiral wrapping, also no change. Am I dealing with a bad stepper? the fact that it tracks perfectly in each loop seems to belie that. What d'yall think?

The picture shows how the letter centers are displaced.

Hey, Mike!

If the "trial basis" runs from before seemed okay... and they don't now... and the code looks okay... and you are running the same gcode files from before? It sure sounds like the grub screw in contact with the flat of the X motor shaft has loosened and "jumps" from one side of the flat to the other with a change of direction. Kind of a "backlash" sort of thing... but due to loose screws rather than part tolerances?
 

jhitesma

Some guy in the desert
Mentor
I used to be careful to do my CAM so the score and marking cuts would be done first. But the kerf is so small that pieces can't really move after they're cut and add in the vac pad I now use and they're not going anywhere. So I stopped worrying about it and just let estlcam pick the order to do the operations in. Sometimes it will do the scores/marks first then cut the part...sometimes it will cut the part first. I haven't had any issues with that but it doesn't really give a big speed increase. It does make generating the gcode slightly quicker for me...but again not a major difference.
 

Jonny360

Junior Member
I used to be careful to do my CAM so the score and marking cuts would be done first. But the kerf is so small that pieces can't really move after they're cut and add in the vac pad I now use and they're not going anywhere. So I stopped worrying about it and just let estlcam pick the order to do the operations in. Sometimes it will do the scores/marks first then cut the part...sometimes it will cut the part first. I haven't had any issues with that but it doesn't really give a big speed increase. It does make generating the gcode slightly quicker for me...but again not a major difference.

We're actually off fusion360 for a moment and following this lead of generating the code directly from illustrator. I'm having some issues on my mac but others seem to have this working on pc. This may be the fastest method we've found to go from plans to code.

https://diegomonzon.com/illustrator-to-g-code-panel-v2
 

lostphoenix

New member
Hey Everyone! Just getting started on building my MPCNC. So far it's working out pretty well. I do have a question though and it's probably been answered in the 140 pages before this but I haven't found it yet. Is there a recommended motor for the needle cutter that someone can point me to? None of the motors I have on hand have mounting holes in the face and I haven't been able to find one easily. If someone could point me to one that's the right size for jihtesma's mount, that would be super helpful! Thanks!
 

dkj4linux

Elite member
Hey Everyone! Just getting started on building my MPCNC. So far it's working out pretty well. I do have a question though and it's probably been answered in the 140 pages before this but I haven't found it yet. Is there a recommended motor for the needle cutter that someone can point me to? None of the motors I have on hand have mounting holes in the face and I haven't been able to find one easily. If someone could point me to one that's the right size for jihtesma's mount, that would be super helpful! Thanks!
I had hoped Jason (jhitesma) might see this... or someone who's built his cutter. Many have... hopefully someone will speak up and let you know what they used for a motor. I personally have never built or used that particular cutter, so really can't help. Did you check his Thingiverse page and read it thoroughly... maybe the comments will give you a clue?

BTW welcome to the MPCNC club. You'll have fun! -- David
 

jhitesma

Some guy in the desert
Mentor
Unfortunately it seems the motor I designed mine around isn't available anymore :( And since I still have a couple more of them on hand and mine works I haven't had much cause to investigate other options. There should be some options out there.. I know I have a few motors from hobbyking that use a bolt on prop shaft and I think sunnysky had some. But I haven't really shopped lately.
 

dkj4linux

Elite member
Unfortunately it seems the motor I designed mine around isn't available anymore :( And since I still have a couple more of them on hand and mine works I haven't had much cause to investigate other options. There should be some options out there.. I know I have a few motors from hobbyking that use a bolt on prop shaft and I think sunnysky had some. But I haven't really shopped lately.
Enough folks have built your cutter that I was sure you'd have a suggestion for another, similar, motor that might work? What was the original one? Maybe that will be enough of a clue regarding specs and dimensions for someone to chase.... :unsure:
 

lostphoenix

New member
Thanks for the response! I'll have to keep looking and see what I can find. I've looked around and found a variety of ideas but it's hard to figure out what the latest and greatest is. For example, the latest thingiverse model I've seen users the flywheel, but there's an instructables out there that says to not use the flywheel. Any advice on that? For now, I'm still focused on getting the electronics wired up for the CNC so it might be a while before I'm cutting foam but I hope not too long.