New to lasers and SLA and need some info

[Jasonpgr]

I've built several FDM printers and am fairly experienced and familiar with them and how they work, but SLA is all totally new to me. So I have a few inquiries before I dive all the way in to this project.

Now that I've found some really reasonably priced laser scanner heads, complete with 405nm laser, and f-theta lens for up to 300mm x 300mm print area, I think this is definitely the way I have to go. The supplier I've been talking to has asked me some questions that extend beyond my knowledge of lasers and galvanometers, seeing as how this is my first project that involves either one that's not a huge surprise. So I thought I'd relay those questions here and see if someone can give me some more direction.

1. Laser power? I'm assuming approx 150 mw would be plenty?

2. digital or analogue galvo? I'm assuming analogue, but I haven't done a whole lot of research on how exactly I'll be driving the galvo? Will I need a separate galvo control board in addition to the ramps 1.4 I was planning on using? Can I control it directly from a raspberry pi? Again, totally new to galvos and how they work and all that, so feel free to enlighten me as to the electronics necessary to pull this off.

3. connecting table? Not even sure what they mean by this, but hopefully someone here knows what they're talking about?

The supplier is a company from alibaba.com and the model they're offering is a Model JD2203 Galvanometer, with a 405nm laser and an f-theta lens lens specifically designed for a 405nm laser and a 300mm x 300mm working area. I'm expecting the total scanning head to come in around $500 or so, which is insanely cheap compared to the other options I've found online from American Suppliers. It's a big enough savings that I'm happy to drop the $500 on one of these scanners and give them a shot. Any feedback or input of any kind would be greatly appreciated since I'm in way over my head on this build.

I'm also trying to figure out exactly HOW all the components will play together. I've been hunting around for some info about what the lasershark board actually does, correct me if I'm wrong, and it looks like it essentially scans an image, line by line, and converts it translates that into movements so the laser will draw the image that is sent (via usb) to the board. So in this instance, it will essentially be working like a DLP Projector. Whatever control software I'm using (would love recomendations for slicers and control software for sla set up) would tell the stepper to move the Z axis up/down, then send an image (rather than a series of gcode commands for all the movements of an FDM printer) to the lasershark board and it will "project" the image onto the resin. Does this mean I could use software designed for DLP printers such as NanoDLP or others if I'm using a lasershark board to control the laser?

I'm attaching the small PDF that the laser scanner company sent me with info about the galvo/scanner head thing I'm considering purchasing. I'd love all the help/guidance I can get!

[Macpod]

1. Laser power? I'm assuming approx 150 mw would be plenty?

405nm 50mW-150mW lasers seems popular for SLA printers. Would those work? Maybe...
This is a challenging question to answer without other system details. Some of these include:
-accounting for resin cure characteristics
-how fast the laser will be scanned
-beam focus
-beam diameter.
I see you plan to use an f-theta lense and that will help minimize some variable complexities.

2. digital or analogue galvo? I'm assuming analogue, but I haven't done a whole lot of research on how exactly I'll be driving the galvo? Will I need a separate galvo control board in addition to the ramps 1.4 I was planning on using? Can I control it directly from a raspberry pi? Again, totally new to galvos and how they work and all that, so feel free to enlighten me as to the electronics necessary to pull this off.

I suspect they are asking if you need a digital scanning head or analog scanning head vs galvo. If you purchase a digital scanning head I presume it will use the XY2-100 protocol and you will not need a LaserShark (nor will the LaserShark be compatible/make sense). If you purchase an analog scanner head you will need an analog controller board such as the LaserShark but first you will need to determine if the input characteristics of the heads input match. I would expect most things would follow the ILDA standard so this would be true but you would need to confirm this. I do not see analog specifications in the attached document.

3. connecting table? Not even sure what they mean by this, but hopefully someone here knows what they're talking about?

The documents suggest this might be a poor translation and correspond to the mounting interface between the scanning head and laser.

The supplier is a company from alibaba.com and the model they're offering is a Model JD2203 Galvanometer, with a 405nm laser and an f-theta lens lens specifically designed for a 405nm laser and a 300mm x 300mm working area. I'm expecting the total scanning head to come in around $500 or so, which is insanely cheap compared to the other options I've found online from American Suppliers. It's a big enough savings that I'm happy to drop the $500 on one of these scanners and give them a shot. Any feedback or input of any kind would be greatly appreciated since I'm in way over my head on this build.

That is quite low in cost. Generally this is an indicator that R&D costs could be avoided or that the product has been cost-optimized. I would be curious to know the KPPS rating of his scanning head as well as the stability characteristics in comparison other companies around the world.

I'm also trying to figure out exactly HOW all the components will play together. I've been hunting around for some info about what the lasershark board actually does, correct me if I'm wrong, and it looks like it essentially scans an image, line by line, and converts it translates that into movements so the laser will draw the image that is sent (via usb) to the board. So in this instance, it will essentially be working like a DLP Projector. Whatever control software I'm using (would love recomendations for slicers and control software for sla set up) would tell the stepper to move the Z axis up/down, then send an image (rather than a series of gcode commands for all the movements of an FDM printer) to the lasershark board and it will "project" the image onto the resin. Does this mean I could use software designed for DLP printers such as NanoDLP or others if I'm using a lasershark board to control the laser?

The lasershark driver (that runs on your computer) accepts x/y coordinates and power intensities for the laser channels, stuffs these into buffers and sends them over usb to the LaserShark. The lasershark in turn prints out these coordinates at the designated kpps rate you configured the points to be displayed at.

the lasershark_hostapp github repo provides two means to use this:
1. You can use lasersharklib to integrate the lasershark libraries into your codebase and drive the LaserShark as fast as possible
2. You can compile lasershark_stdin and pipe samples in a simple text format into the application. This latter approach is less efficient, but much easier as you do not have to learn how to use libusb/etc.

In this repo are two examples of how to use lasershark_stdin. One is lasershark_stdin_circlemaker. If you pipe this into lasershark_stdin, a circle will be drawn. The other is lasershark_stdin_displayimage which will line-by-line display a png image if you pipe the output into lasershark_stdin.


Your software would need to control the z-stepper motor (i.e. you could write a python app that controls this between layers). If you can get software to present you with a png image of each layer you could theoretically adapt it, but you would need to have some integration glue on and (all of which you should be able to confirm before buying any hardware).

[Jasonpgr]

Thank you for the extremely thorough reply! So I'm slightly modifying my laser build as the company from China I was working with ended up being a lot more expensive than I had previously expected. Their lenses are incredibly affordable though, so I'll definitely be getting their f-theta lens that covers a 300x300 print area.

The total cost ended up being around $2400, but they were also charging $700 for their laser alone, so I said no thanks on that one considering the galvo doesn't need to move all that fast since it needs to go slow enough to be able to cure, and it doesn't take a super powerful or fancy laser to cure UV resin. So I figured it wouldn't hurt to save some money on these two areas.

As for the rest of the components, I've been hunting around on ebay/amazon and all that and these are the ones I've come up with instead.

for a cheap simple laser
https://www.ebay.com/itm/405nm-100mW-Blue-Purple-Focusable-Dot-Laser-Module-TTL-Long-time-Working-12V/322293392564?epid=1794342367&hash=item4b0a2ee4b4:g:0lUAAOSwCGVYAG34

and for a laser galvo
https://www.ebay.com/itm/30Kpps-High-Speed-Laser-Lighting-Galvo-Scanner/112841533160?epid=13015689010&hash=item1a45e126e8:g:tr8AAOSwyTValkuQ

If you have any components you'd recommend for a 3d printer that you've used or seen used, I'd appreciate whatever advice you've got.

[Jasonpgr]

Actually, the original company got back to me and they suggested another product they have that is a galvo scanner with the housing, the f-theta lens for a 405nm laser, set up for a 300x300mm scan area, all the mirrors, and galvo bits and a laser that mounts up to the head, and they said they could do that one for $700 total. That's about the cost of just the lens from other companies I've seen, and these guys sell tons and tons of scanners and their reviews look pretty solid. I think I'm going to go this route as long as you think it should work for a high accuracy 3d printer set up. I'm going to attach the manual for your to review and let me know what you think.

Also, it turns out they can do it in either a digital, or analogue. Which option would be easier to integrate into the software? If I went the digital route rather than analogue with a lasershark board would I need some kind of usb dongle to send the xy coordinates out to the galvo motors? Could I use a standard arduino set up of some kind like an fdm printer? But rather than sending steps to a stepper driver, it's basically sending steps to a galvo motor? I've got a lot to learn before I'm able to make this happen so I appreciate whatever help I can get.

[Macpod]

Actually, the original company got back to me and they suggested another product they have that is a galvo scanner with the housing, the f-theta lens for a 405nm laser, set up for a 300x300mm scan area, all the mirrors, and galvo bits and a laser that mounts up to the head, and they said they could do that one for $700 total. That's about the cost of just the lens from other companies I've seen, and these guys sell tons and tons of scanners and their reviews look pretty solid. I think I'm going to go this route as long as you think it should work for a high accuracy 3d printer set up. I'm going to attach the manual for your to review and let me know what you think.

It sounds interesting for everything and having a pre-assembled head with galvo/power supply/galvo amplifiers/laser/controller could save you headaches. I couldn't comment on if there were any areas of concern without a datasheet but hopefully from my first reply you know about some of the things to look out for (i.e. monotonicity/linearity)

Also, it turns out they can do it in either a digital, or analogue. Which option would be easier to integrate into the software? If I went the digital route rather than analogue with a lasershark board would I need some kind of usb dongle to send the xy coordinates out to the galvo motors? Could I use a standard arduino set up of some kind like an fdm printer? But rather than sending steps to a stepper driver, it's basically sending steps to a galvo motor? I've got a lot to learn before I'm able to make this happen so I appreciate whatever help I can get.

This will be dependent on the software you are integrating with and the speeds needed. If starting from scratch digital would be easier. Going digital may increase you price estimates as the equivalent of a Lasershark will need to be integrated into the head but even so it may be worth it.

If you go the digital route you will still need something to output XY2-100 protocol. An Arduino could be sufficient or in conjunction with line level shifting/etc according to the head's input requirements.

[Jasonpgr]

Sorry, I totally forgot to include the datasheet/manual they sent me with the stats on it. It seems like if it can interpret the digital signals, that it would already have a similar board type thing in the head, but I could be wrong. If not, I think going to the lasershark route is definitely the best and simplest way to go. I'm probably going to be purchasing the CW software license since that looks like the smoothest way to itegrate everything since I'm already way out of my knowledge base on this one.

I was able to build some really great and intricate FDM printers using marlin firmware and Cura for slicing, so I assumed it would be just as easy to use open source software for a laser SLA set up, I was way wrong. I have little to no software/coding experience at all beyond very very basic arduino stuff, and mostly that is modifying existing codes like I have done when setting up Marlin for my printers specs.

It was so user friendly and simple to figure out, I'm really wishing there was a marlin firmware designed for SLA that would take the gcode that is sent out by the slicing software and convert it to galvo angles, rather than steps for a stepper driver. In conjunction with options for controlling the Z axis the in different ways we would need for wiping resin and all that. But alas, we haven't reached that point yet.

All that aside, I'm still determined to sort this out and get it up and running. So here's the data sheet, let me know if you think this would work and if you have anything else for me to look at to get the CW software talking to the lasershark and an arduino to move all z axis bits around at the right time, I'd really appreciate it.

[Macpod]

Sorry, I totally forgot to include the datasheet/manual they sent me with the stats on it. It seems like if it can interpret the digital signals, that it would already have a similar board type thing in the head, but I could be wrong. If not, I think going to the lasershark route is definitely the best and simplest way to go. I'm probably going to be purchasing the CW software license since that looks like the smoothest way to itegrate everything since I'm already way out of my knowledge base on this one.

It should for a digital head. The logic board I am referring to that would attach to an Arduino/etc would be for the purpose of level shifting (think 3.3v to 5v TTL Level conversion or TTL to RS232.. very simple/common) vs a processing or analog to digital module.

Before purchasing a copy of CW make I would encourage you to reach out to the developer to verify it still supports the LaserShark. CW used to be freely downloadable and supported the LaserShark but this support was removed and the product went to a license-based model. According to another post on this forum the LaserShark is not supported by the latest CW.

[Jasonpgr]

That's really unfortunate. I sent them an email a couple days ago and I'm still waiting for a response.

It kinda seems like there's no good way go about making a laser driven SLA printer. At least from a software/firmware perspective. I see a bunch of companies are doing it, because you can buy one. And most of them are driven by a small board that doesn't seem much different than a raspberry pi or arduino type board. A lot will print from an SD card much like a typical FDM style printer, which leads me to believe they are just streaming gcode from a slicer software.

Maybe I should phrase this differently. Since you have a much stronger knowledge of galvos, and coding and all the areas where I'm lacking, if someone came to you and said "I have this galvo head (the one I included the manual for in the last post) can you turn it into a laser sla 3d printer?" How would you go about it? Would you prefer a digital or analogue head since it can come as either one? What are the missing components and where would I source them in order to pull it all together.

With my limited knowledge and experience it seems like the best way would be to find some way to reconfigure a basic marlin/repetier firmware to drive a galvo instead of steppers on the XY axis, and swap the extruder bits for laser ttl on/off commands, and just use a ramps board to do it all. But I have no idea how to go about doing that, and I don't want to have to manually modify massive amounts of Gcode manually before doing a print. The solutions are definitely out there because people are doing it fairly inexpensively. I just keep running into brick walls, so maybe I need to take a whole new approach all together.

[Macpod]

Maybe I should phrase this differently. Since you have a much stronger knowledge of galvos, and coding and all the areas where I'm lacking, if someone came to you and said "I have this galvo head (the one I included the manual for in the last post) can you turn it into a laser sla 3d printer?" How would you go about it? Would you prefer a digital or analogue head since it can come as either one? What are the missing components and where would I source them in order to pull it all together.

With my limited knowledge and experience it seems like the best way would be to find some way to reconfigure a basic marlin/repetier firmware to drive a galvo instead of steppers on the XY axis, and swap the extruder bits for laser ttl on/off commands, and just use a ramps board to do it all. But I have no idea how to go about doing that, and I don't want to have to manually modify massive amounts of Gcode manually before doing a print. The solutions are definitely out there because people are doing it fairly inexpensively. I just keep running into brick walls, so maybe I need to take a whole new approach all together.

I would probably go with a digital head if it wasn't significantly more money and after I determined the speed would be sufficient (vs an analog head). Before I purchased anything I would make sure software existed and if not write the software.

There are two software approaches, one is to modify a slicer to export images, then draw pixel by pixel. The other means is to make a g-code to x-y point converter and use that in conjunction with either an xy2-100 head (easier and already integrated in) or lasershark board attached to an analog head (maybe faster).