So I bought a cheap 40w laser....

[jhitesma]

Great write up! Really sounds like you have a great handle on what you've gotten yourself into and the limitations of the K-40.

Yeah, I've been reading about the K40's for a few years...but stayed away due to the limitations. But, at $300 I couldn't say no and figured the limitations would make for an interesting project

I have plans for a drop in exhaust filter (using kitchen scratch pads for the filter material) I'll pass them along to you if you want.
(It does add some to the clearance behind the machine though.

Yeah, I'll check them out. Still debating where the machine will live long term and what I'll do for exhaust. For now it's in front of the back patio door and the weather is getting nice so I can just run a vent through there...but that won't be a permanent solution.

Curious how you plan to handle focus (raise lower the bed.)

I haven't decided yet. I'm debating building a motorized bed vs. just a fixed bed. I don't play on cutting a wide variety of materials so adjustable focus doesn't seem like that big of a priority. Just something I can fix at the right height for 1/8" material since that's mostly what I plan on cutting with it. But even then I'm not sure if I want to do a needle bed, honecomb, something like a bigger version of @dkj4linux's popsicle stick surface....I'll figure something out...just need to get it moving and lasing before I worry too much about it.

Congrats on joining the "Lazer" crowd!

Well, I did have the small laser on my MPCNC already But it is nice to have a "real" laser now

Didn't get a whole lot done last night. Finished printing the foot inserts and installing the new feet. Did some practice crimps with the new crimpers (so nice having real crimpers, but still take a bit of practice!) and just took it easy. My ammeter did arrive - but there's no appropriate hole to install it in. So tonight I may grab a scrap of wood and cut a new temporary control panel.

Still no word from uberclock about when my smoothieboard will ship Not very happy about that. Almost a week later and still not shipped. So...I gave in an ordered one from Amazon after I noticed that the seller on amazon is an authorized reseller. They also include all of the connectors and the 5v regulator which the official source doesn't. I figure the second board can go on my printer or MPCNC when it finally arrives.

 

[kilroy07]

Yeah, I'll check them out.

I had to clean the plans/cut files up a bit, but the 3D view should give you enough insight to assemble.
(If not, just ask.)

I designed this after I replaced the weak squirrel cage blower with one from harbor freight and the cage was caked with soot... So I came up with this prefilter, not good enough to pipe back into living spaces, but to let your bearings on your blower last a tad longer.

 

[jhitesma]

I like the filter idea, I'm still debating just how I want to deal with the exhaust plenum, and what I'll use for a fan. For now I just plan on getting this thing up and going with the stock fan and leaky chassis. But...once I've got it working I do plan on sealing up the cutting chamber much better and doing a good evacuation setup. Though as I mentioned yesterday a big part of that will depend on where I decide this will live long term. Not sure if I want to make room for it in my back room and keep it in the house, or make room for it out in my shop and keep it out there. In the house it would mean tackling the back room...which is something I really want to do. (I actually did back in April and had it really nice back there...but then a bunch of stuff collected and we trashed it further digging out a bunch of stuff my wife had in storage.) Right now the back room is a mess and it's mostly due to stuff I need to sort through from my wife and decide what to keep, what to pitch, and what to donate. But...I just haven't had the energy to deal with that yet.

Making room in my shop isn't much easier. Remember a big point of this project was to have something I could do in the house since I can't spend much time in the shop until my daughter is a bit older and can be left unattended longer. And about the only way I could fit this into the shop is if I got rid of my large format plotter. Which I was already considering getting rid of anyway now that I have the MPCNC since I haven't used it since shortly after I started on the MPCNC.

But I can't let this live on the dining room table forever. I could always turn around and sell it when it's done...but...I have a feeling I won't want to give it up


Anyway - progress. Not much to report. I didn't do anything physical on it last night. I just felt like being lazy. I started to draw up a temporary faceplate but even that felt like too much work. I did dig out a bunch of endstops I had stashed away but didn't work on wiring them up as interlocks or to replace the optical endstops. The smoothie should be here tomorrow which will hopefully kickstart me again.

I did send more time reading about the wiring...you see I'm still debating how I'll deal with laser power control.

There are two approaches. You can either tie the power level pin to 5v so it's always at full power and then PWM the on/off signal to the laser to control the power - or you can PWM the level pin and then just trigger the on/off. I'm not a fan of the idea of PWMing the on/off signal. That just feels a bit brute forceish to me. The power supply is already doing it's own PWM internally to control the laser power and stacking PWM on PWM just feels sketchy to me. But...with that method you can keep a potentiometer in the power loop to control max power. So basically you have a physical knob on the machine that you set to the highest ma reading you want to see go through the tube - then you let the controller use it's full 0-100% range. And if you keep a pot in the control loop then it's really not good to PWM the level pin since errors can add up and you won't have as fine control. So if you keep that pot in for setting max power you'll want to only PWM the trigger line to control power. Which is simple to wire.

But...I don't have a pot from the factory. I could add one...but I'm not sure I really want one. I prefer to control everything I can through software. That's not for everyone but it's what I tend to prefer. The downside here is you have to use software configs to set your max power. And as the laser tube ages you'll have to re-adjust them instead of just turning a knob to compensate. Not a huge downside for me. If I was mostly doing engraving then it may not be as idea...but I plan on mostly doing cutting and only occasionally doing engraving so super fine power control isn't as critical. So I'm leaning more towards PWMing the level pin on the power supply and just triggering the on/off and letting all of my power settings be controlled through the new controller.

If I had a fancy pot like @nebraskatrevor with the readout in it and everything then I could probably be convinced to wire it in and go with the PWM on the trigger line. But I don't have one of those on hand and the only ones I've been able to locate are rather pricey. I do have some standard pots I could wire in...but it sounds like you really want a 10k 10 turn pot for fine control and I definitely don't have anything like that on hand. I have some 5k and 10k linear pots but nothing with the resolution of a 10 turn. Still something may be better than nothing...but...do I really want another knob on my control panel.

Thankfully I have at least another day to think about it before I have to decide

 

[jhitesma]

Oh, and I keep forgetting. I did post the feet to Thingiverse if anyone wants them:
https://www.thingiverse.com/thing:3148038

 

[jhitesma]

Yesterday my "replacement" smoothie from Amazon arrived, still waiting on the one from the "official" store to ship.

I did power it up over USB and confirmed it's functional. Also downloaded the smoothieware source and and toolchain allowing me to build the firmware from source - was actually VERY easy to get setup (on mac/linux at least).

I haven't wired it yet. I'm still debating a few options for mounting my safety interlocks and digging up wire/connectors to hook everything else up.

Plus...my middleman boards arrive tomorrow and I already have the connectors for them. So instead of totally rewiring it maybe I will just hook that in to get it going quicker. I can always rewire better when I make more mechanical changes down the road.

Oh, and the dimensions and port locations on the smoothie aren't the greatest for this application. So...debating how I want to mount it.

 

[jhitesma]

Little bit of progress...but I've been lazy. The big issue is lack of space. It's getting harder and harder to work on the machine without breaking anything since I have no table space next to it. And the spot where it's sitting is where I had been doing all of my soldering - so making cables and boards has been kind of annoying since I had to setup a new soldering station.

But as predicted yesterday my middleman boards arrived:

Wonderful job on them as usual by OSHpark and showed up way sooner than expected!

Last night I finally made some room to melt a bit of solder and populated the bits I need:

Some of the features on this board don't make a lot of sense to me. Like the "LaserFire!" and "LaserPWM" pins...all they do is connect to the L/H and In pins right next to them. I guess the designers idea was that anything from the controller board should connect to this board - but that just seems silly since those lines just get passed right back out to the laser power supply - there's really no reason for them to connect to this board.

Same with the water pump connection and 12v - the only reason to pass 12v through this board is to supply the water pump (and my water pump is 110v not 12v) but the board does nothing to interface with the pump. So why not just wire it straight to 12v.

The "Temp" pins I have no clue what the point of them is. So basically all I'm using are the 4 pins that break out the stepper wiring, and the 2 pins that break out the endstop signals. I also wired up the 5v in since the stock optical endstops need 5v to operate, and the ground line is also needed to complete the endstop circuit.

So Y stepper is connected to the smoothie, now I just need to make a few jumpers to go between the power supply and smoothie and middleman. Hopefully I'll take care of that tonight and can see this thing actually move a bit!

I'm not yet wiring up the laser itself...that will come after I have it moving reliably. And I'm still debating how to do the PWM. Looking at cohesion's smoothieware config file it appears that they're only PWMing the laserfire pin and not controlling the fire pin and power pins separately:

## Laser module configuration
laser_module_enable                           true            # Whether to activate the laser module at all. All configuration is
                                                              # ignored if false.
laser_module_pin                              2.5            # this pin will be PWMed to control the laser. Only P2.0 - P2.5, P1.18, P1.20, P1.21, P1.23, P1.24, P1.26, P3.25, P3.26
                                                              # can be used since laser requires hardware PWM
laser_module_maximum_power                    0.8             # this is the maximum duty cycle that will be applied to the laser
laser_module_minimum_power                    0.0             # This is a value just below the minimum duty cycle that keeps the laser
                                                              # active without actually burning.
#laser_module_default_power                   0.8             # This is the default laser power that will be used for cuts if a power has not been specified.  The value is a scale between
                                                              # the maximum and minimum power levels specified above
laser_module_pwm_period                       200              # this sets the pwm frequency as the period in microseconds

switch.laserfire.enable false
switch.laserfire.output_pin 2.6
switch.laserfire.output_type digital
switch.laserfire.input_on_command M3
switch.laserfire.input_off_command M5

Note that the switch.laserfire.enable is set to 'false' disabling it. Instead they're just using pin 2.5 to pwm the laser. I'm still trying to figure out the smoothieware config to make sure I'm interpreting this correctly, but it sure looks to me like they're not using the power pin and are just PWMing the fire pin as recommended for machines with an analog power control.

I'm kind of leaning that way....but not having analog power control I'm also still considering doing the two pin option and PWMing the level and just on/off the fire pin.

 

[jhitesma]

Made a bit of progress last night but nothing worth taking any photos of.

Safety switch on the main cabinet is temporarily mounted with some double sided tape to test it out. I figure worst case the tape fails and the switch falls - which would trigger the safety. So while I do plan on mounting it properly this is good enough for testing.

Also started wiring the middleman board to the smoothie - and found that I only had four FF jumper wires left Ordered some more but they won't be here until tomorrow (Had to pay extra to get them quick from Amazon instead of the dirt cheap ones I usually order direct from China since I didn't want to wait for them.)

Since I now have a nice crimper I decided to just start making my own wires...but...turns out I'm pretty lousy at using the crimper. At least with the size wire and connectors I have on hand. May have been using the wrong crimp pins though so we'll take another look tonight. May just need more practice with the crimping tool. Or I may just go back to crimping by hand at this point. We'll see.

So didn't make it all that far last night....

 

[jhitesma]

Just a quickie - I forgot to mention yesterday but I finally got a shipping notice from Uberclock (the official smoothieboard mfg.) that my original order shipped. And just now I got a second e-mail from them with a different tracking code again saying it had shipped (though neither code is returning anything.)

Seriously not impressed with uberclock right now. Will be interesting to see what if anything actually arrives.

 

[jhitesma]

Well, only half of the jumper wires I ordered arrived last night. Not the end of the world, basically I ordered two sizes - 8" long and 12" long. Only the 12" arrived. So they'll work...but they're a bit longer than I need. Which isn't a big deal since at this point I'm not worrying about clean wiring just functional to make sure everythign is going to be ok.

Here's my temporary/test installation of the lid safety switch:

Pretty simple really, just a bit of tape to hold the switch in place and some more tape to hold a 1/2" square dowel in place to press against it. Seems to work so over the weekend I'll try and come up with something more permanent (as well as add one on the laser chamber in the back....I haven't decided yet if I want to add one to the electronics area....I'm leaning against it.)

I'm also toying with adding some indicator lights - just can't make up my mind about what lights I'd like and what color they should be. I was thinking one to indicate the laser is powered and another to indicate that it's armed and can fire (i.e. safeties are engaged). Won't be doing either just yet...but something I'm debating doing soon.

The smoothieboard is "wired":

Yeah, it's a rats next right now. Like I said, I'm not worrying about wiring it cleanly yet - just testing that everything works.

I didn't power it up yet...because...well...the smoothie board power wiring is confusing and I wanted to sleep on it to make sure I got it right:

It's kind of hard to see...but...the yellow wire is my +24v and red is +5v, the black wires (one is hiding behind the yellow) are grounds.

But the two connections are reversed - positive is on the outside of them both so one is +/- and the other is -/+

But....the markings on the board seem to indicate otherwise - because if you use those open pads to wire to instead of the connector...then the 24v side IS reversed and matches the 5v side. SO those markings on the board are for those extra pads not for the connector.

Here's the "helpful" explanation from smoothie:

I had remembered seeing that - but forgot the details of it. So had to dig through the docs again to make sure I have things wired correctly.

The main smoothie wiring diagram is also potentially confusing about this:

Looking at the labels it looks like they go in the same order +/- But that VBB connection makes that crossover - and on the printed card that comes with the smoothie the color choices make it VERY hard to see that the wires cross right there.

What's interesting to me is that they have a lot of warnings in the docs about how if you reverse power and fry the board it's your own fault...but the docs are not very clear or helpful in preventing that!

I haven't wired up a PWM or fire wire to the laser yet. But I should be able to power it up at this point and the smoothie should have power, should be reading the end stops, and should be able to command the steppers.

But...the steppers might be reversed...depends on if I got the wires connected in the right order or not

And I need to update the config on the smoothie board before I put power to it again. So when I reached this point last night I decided to take a break and come back today instead of pushing on and risking forgetting something or making a mistake.

Instead I decided to work on the control panel a bit since I'd like to have my controls mounted and I want that amp meter in before I try powering the laser.

So I fired up onshape and drew the control panel. The lip around the outside is just temporary and there to show me how much of the panel is actually visible vs. how much is hidden under the front of the machine when it's mounted.

And here it is without the mockup components:

The meter I modeled myself, the LCD board I found a pre-existing model someone else had done in onshape and just imported it.

I'm really not sure about this panel layout. The SD card on the LCD is blocked so won't be usable...but shouldn't matter since the smoothie has it's own SD card - and I'll probably be using the smoothie's Ethernet connection to push cut files to it anyway. But it still feels kind of cramped.

I'm also not sure how big of an opening I want for the LCD. Right now it's just the display area - so the LCD will sit flush with the bottom of the panel. But that doesn't allow much of the knob to stick above. I also haven't decided what I want to do with the estop button on the LCD (whether or not I'm even going to use it) and if I want the beeper exposed or not.

And the 3 switches...just don't feel good to me. I'll probably replace the 2 original square ones at some point just because I want them all to be similar and hate switches that need square openings. But it's probably "good enough" of a layout that I'll toss some 1/8" ply on the MPCNC tonight and cut one out just to use for now.

Which also means I'll have to work on wiring the LCD. But if I put that off a day or two more then my order from uberclock should arrive and it has an LCD adapter in it which will make that much simpler.

But...it's looking very likely that I'll be making some cuts this weekend at this rate! Once the machine is working then I can focus on cleaning up the wiring, fixing other little issues, upgrades, and using it to cut a new panel with engraved labels

 

[kilroy07]

Coming along nicely! I was just going to suggest that panel should be your first “real” cut.

I made an external one to monitor Water temp and flow rate... spray painted a sheet of acrylic black then etched out the labels for a backlit panel (came out really sharp) but not practical for your setup.

 

[jhitesma]

I don't think I'm going to put a ton of time/effort into the panel. I'm leaning more and more towards building a custom case for it. Just not sure if I want to go the MDF route or the 2020 and acrylic route and whether I want to make the cutting area larger or the machine more compact when I do it. I figure it will be easier to decide after I have it up and going and try it for a few projects. Then I can decide which way I want to go with it.

But cutting it's own panel just seems like a fitting first project - so it will get done even if I'm ok with the temporary first draft

 

[Bricks]

I always kind of envy you guys that are so techy around electronics. Give me something mechanical and I can work it out but electronics it is just a fog.

 

[Fyathyrio]

Grab a meter and verify continuity on the power/ground pads & connectors? No guesses that way.

 

[jhitesma]

I always kind of envy you guys that are so techy around electronics. Give me something mechanical and I can work it out but electronics it is just a fog.

If it makes you feel better my progress last night stalled because I decided to mount the smoothie properly before going any further. Which is a "simple" mechanical task...but which is turning into a big pain for me

The metal plate that the stock board attached to isn't big enough for the smoothie to mount to. So my original plan to just drill a few holes isn't going to work. Instead I have to either fabricate a new plate entirely, Or fabricate some kind of extension. And even then the opening in the side of the case isn't quite big enough to fit the USB, Ethernet, and SD at the same time...so I either have to decide not to deal with one of them or I have to enlarge the opening. Since I plan on mostly working over Ethernet I figure I can block the SD...but I'm not huge on that. And even with all of this the smoothie is so tall that there's a good chance I won't be able to fit the connectors on the bottom for the pins the laser is controlled from. So now I'm looking at totally different mounting options.

I will admit - the cohesion board is looking awful nice right now. Simplifies all of this. No need for the middle man and it mounts right where the stock board did. But I really didn't like the stepper drivers they used on it or that it was a bit too specalized towards the k40 and I wanted the option to also use the board for me MPCNC and/or 3D printer depending on how I like it's CNC abilities.

Electronics don't scare me, and mechanics don't scare me much either...but I'm far better at electronics than mechanics

 

[jhitesma]

Grab a meter and verify continuity on the power/ground pads & connectors? No guesses that way.

Yeah, I'll probably also use a smoke stopper the first time I put power to it. There are markings - they're just not well placed and confusing. So I'm about 98% sure I'm safe...but...still don't want to fry a $150 board before I even use it!

 

[jhitesma]

Well, weekend progress was not as easy to achieve as I had hoped. My daughters friends were all busy or out of town so I was tied up entertaining her all weekend. On the upside - I can actually see the floor in her room again now as I refused to play with her until it was cleaned. On the downside I had to do 90% of the cleaning.

I had hoped to get out and cut the new panel at least...but due to a power outage a few weeks ago the computer in my shop had to be rebooted - and that thing is so slow I wasted almost 30 minutes just getting it powered up and loading Estlcam and my artwork (for some reason estlcam suddenly stopped working on my mac...not sure what's going on there but I'm critically low on drive space there so uninstalling wine/estlcam and then reinstalling it all isn't a great option and I'll probably just have to live without estl on there for now.

Anyway, once I got the DXF loaded up and went to define my cuts...I found that I had left some out. So back to onshape, and by the time I had m gcode ready and the machine cleared off and ready...it was time to give my daughter a bath - I was out of time

Hoping I can sneak enough time to cut it while I'm home for lunch today.

I did do a bit of playing with the smoothie trying to get the GLCD working now that I have the adapter...but only running it off USB power the LCD won't power up. So I need to bite the bullet and give it full power - but I still need to finish mounting it before I tackle that and I'm still not sure how I want to approach that.

So...yeah...not much got done this weekend

 

[rockyboy]

Sounds like you got a lot done this weekend - just not on the project you expected

 

[jhitesma]

Sounds like you got a lot done this weekend - just not on the project you expected

Yeah....I also made a new chute for my coffee grinder to replace a temporary wooden one I've been using for close to 10 years:

Which is up on thingiverse in case anyone else has a Nuovo Simonelli MDX grinder and doesn't want to use the stock doser that came with it and direct dispense instead: https://www.thingiverse.com/thing:3170032 (FWIW no I didn't buy a new $1k+ commercial coffee grinder...I got it for $70 from a guy who was retiring from running a coffee trailer and who's wife listed it cheap because the bottom is broken and she didn't know how much it cost.)

I also made a smaller hopper for it:

Which I haven't uploaded yet since I optimized it a bit more and need to reprint it still.

It was kind of neat though. Sunday morning had the itch to make the smaller funnel and within 30 minutes went from having an idea to watching it take shape on my printer. Plus - got to teach my daughter the basics of onshape which was fun

Oh - and didn't have time at lunch to cut the panel...so progress will continue to be slow

 

[nebraskatrevor]

Little bit of progress...but I've been lazy. The big issue is lack of space. It's getting harder and harder to work on the machine without breaking anything since I have no table space next to it. And the spot where it's sitting is where I had been doing all of my soldering - so making cables and boards has been kind of annoying since I had to setup a new soldering station.

But as predicted yesterday my middleman boards arrived:
View attachment 116673

Wonderful job on them as usual by OSHpark and showed up way sooner than expected!

Last night I finally made some room to melt a bit of solder and populated the bits I need: View attachment 116674

Some of the features on this board don't make a lot of sense to me. Like the "LaserFire!" and "LaserPWM" pins...all they do is connect to the L/H and In pins right next to them. I guess the designers idea was that anything from the controller board should connect to this board - but that just seems silly since those lines just get passed right back out to the laser power supply - there's really no reason for them to connect to this board.

Same with the water pump connection and 12v - the only reason to pass 12v through this board is to supply the water pump (and my water pump is 110v not 12v) but the board does nothing to interface with the pump. So why not just wire it straight to 12v.

The "Temp" pins I have no clue what the point of them is. So basically all I'm using are the 4 pins that break out the stepper wiring, and the 2 pins that break out the endstop signals. I also wired up the 5v in since the stock optical endstops need 5v to operate, and the ground line is also needed to complete the endstop circuit.

View attachment 116675

So Y stepper is connected to the smoothie, now I just need to make a few jumpers to go between the power supply and smoothie and middleman. Hopefully I'll take care of that tonight and can see this thing actually move a bit!

I'm not yet wiring up the laser itself...that will come after I have it moving reliably. And I'm still debating how to do the PWM. Looking at cohesion's smoothieware config file it appears that they're only PWMing the laserfire pin and not controlling the fire pin and power pins separately:

## Laser module configuration
laser_module_enable                           true            # Whether to activate the laser module at all. All configuration is
                                                              # ignored if false.
laser_module_pin                              2.5            # this pin will be PWMed to control the laser. Only P2.0 - P2.5, P1.18, P1.20, P1.21, P1.23, P1.24, P1.26, P3.25, P3.26
                                                              # can be used since laser requires hardware PWM
laser_module_maximum_power                    0.8             # this is the maximum duty cycle that will be applied to the laser
laser_module_minimum_power                    0.0             # This is a value just below the minimum duty cycle that keeps the laser
                                                              # active without actually burning.
#laser_module_default_power                   0.8             # This is the default laser power that will be used for cuts if a power has not been specified.  The value is a scale between
                                                              # the maximum and minimum power levels specified above
laser_module_pwm_period                       200              # this sets the pwm frequency as the period in microseconds

switch.laserfire.enable false
switch.laserfire.output_pin 2.6
switch.laserfire.output_type digital
switch.laserfire.input_on_command M3
switch.laserfire.input_off_command M5

Note that the switch.laserfire.enable is set to 'false' disabling it. Instead they're just using pin 2.5 to pwm the laser. I'm still trying to figure out the smoothieware config to make sure I'm interpreting this correctly, but it sure looks to me like they're not using the power pin and are just PWMing the fire pin as recommended for machines with an analog power control.

I'm kind of leaning that way....but not having analog power control I'm also still considering doing the two pin option and PWMing the level and just on/off the fire pin.

Just a note, I had to Change my PWM period to 400 as directed by the helpful folks at the Cohesion3D page. Before that I had no variable control of the laser, after perfect digital control which I am loving BTW. Your mileage may vary but keep it in mind as you progress!

 

[jhitesma]

Just a note, I had to Change my PWM period to 400 as directed by the helpful folks at the Cohesion3D page. Before that I had no variable control of the laser, after perfect digital control which I am loving BTW. Your mileage may vary but keep it in mind as you progress!

Good to know! I've seen quite a few people suggest 200 as the PWM period for K40's - but have also seen a lot of discussion about trying to find an optimal period. So that gives me some good starting points. I imagine it could also vary depending on the power supply (how it does PWM internally) and the exact tube (and it's properties) since there is no such thing as a "standard" K40

Also, I mentioned that I've been playing with the smoothie and having a few issues. The LCD I assume is due to power issues at this point, RAMPS can power a LCD off USB but smoothie apparently can't which I assume is my current issue as I've confirmed that I have my config file setup for the GLCD. However I'm also having issues with the ethernet - so far I've been unable to get it to actually work. But without having the LCD working it's a bit trickier to debug since I have to connect to the board over USB and issue console commands - which isn't horribly difficult but is a bit more time consuming and more things to potentially go wrong.

I'm really hoping to get my panel cut at lunch today and then fire this bad boy up on real power in the machine tonight. Fingers crossed for sure!