BuildLog.Net - Document your builds!


Buildlog Title: Plasma Cutter

newest first oldest first
Builder: bdring
Member Since: 2009-11-22
reader comment Comment from: incognico on Monday, October 20th 2014 - 2:25 AM
Did this project/discussion continue anywhere? I can't seem to locate it if it did - but it's a Monday and I haven't had my coffee yet so maybe I'm missing something obvious...!?
reader comment Comment from: BenJackson on Wednesday, December 12th 2012 - 6:50 PM
There are chemicals meant to be put in the water trough for a plasma cutter to keep things "fresh". Sorry I don't know any brand names offhand, but I have read about them before.
reader comment Comment from: bobt on Wednesday, December 12th 2012 - 3:38 AM
Bart - Is it possible to get an update on the Plasma Cutter. And may be some more pictures. I just can't wait to start building one of these.


Bob Teeter
reader comment Comment from: cncmadness on Friday, December 7th 2012 - 7:35 PM
on the subject of the water tray "looking pretty nasty already" would a type sluce box work , an angled tray with a removable catch tank at one side with recycled pumped water for flow over tray push the particles into the catch tank.

i think i'm over complicating it for small scale but maybe for larger

Monday, October 22nd 2012 - 12:13 PM

The spray of sparks seems to emit a lot of the smoke. Extinguishing them as soon as possible appears to reduce the smoke in my case. I am seeing less smoke than before. The mess on the floor is gone, but the water is looking pretty nasty already.

add commentadd comment in the forum

reader comment Comment from: macona on Monday, October 22nd 2012 - 2:24 AM
A water trough itself will do almost nothing for the smoke. What does work great is having the work sit right at or just below the surface of the water, normally known as a water table. This does a pretty good number on the smoke and noise. And no, the torch wont care if the end is in water. In fact my plasma cutter has an option for using water to replace the shield gas.
reader comment Comment from: greenvandan on Sunday, October 21st 2012 - 8:50 PM
What I've done in the past is to take a 45° bit and grind the tip down a bit. I try to go for a .125" flat on the tip. That's working with 3mm Dibond which I think is about the same thing as the AluPanel. The bit will remove a little more material and allow the aluminum to flex a bit. Even long folds are pretty easy this way. You could also cut a corner reinforcement that way and stick it in with some VHB tape after you seal with RTV, I'm not sure how much stress the water will put on those seams.

Saturday, October 20th 2012 - 3:34 PM

I completed my water trough. This should make things less messy and less smoky.

I made it out of AluPanel. I thought it would be a simple way to get a water tight and corrosion resistant part.

I cut out the basic shape with a 3 inch perimeter around it. I then hand routed a V groove down to the far side metal using a 45deg bit. I thought about doing some fancy overlapping corners, but the placement of the V grooves would be very critical and I did not know the bend allowances. I ended up just cutting out the 4 corners. They were already V grooved, so they fit together really well.

I takes quite a bit of muscle to fold the longer edges, but it worked pretty well. A wider angle bit like 60deg will might have made it easier.

I then used clear RTV to seal the corners. I also ran a bead along the inside of the folded corners. The skin is pretty thin there. I did not want a little ding to cause a leak.

water_trough2.JPG

water_trough3.JPG

water_trough4.JPG

add commentadd comment in the forum

Saturday, October 13th 2012 - 11:57 AM

I have not really measured all the parameters yet. The basic machine is pretty accurate. If I cut a square, the sides are parallel and of equal length within a few thou. I return exactly where I started on complex shapes like the PS:One snowflakes, so I feel good about backlash.

I have been playing with voltage, current, and speed a lot so I have not been worried about the kerf yet. Once I settle on those values I will try to see how consistent that is. Trying to cut an exact 2" diameter circle probably needs a consistent kerf more than anything else.

The electronics are finalized now and I am finalizing the frame changes that includes adding the water through. Cutting steel makes quit a mess. It makes a nasty black dust on the floor.

add commentadd comment in the forum

reader comment Comment from: royco on Saturday, October 13th 2012 - 3:43 AM
That looks awesome Bart.

Maybe I missed it but what accuracy are you getting from that setup?

Saturday, October 13th 2012 - 1:24 AM

Here is my schematic for the control box. See the PDF for better resolution

ctrl_schm.jpg

Attachements...
plasma_box_schematic.pdf
(97.2 KiB) Downloaded 272 times

add commentadd comment in the forum

Wednesday, October 10th 2012 - 8:09 PM

That cut was run at like 80ipm at 30Amps. I have run it up to about 180ips while cutting without a problem at that same power. I can go up to 45Amps. I have million variables to play with for a while, but the results are pretty cool.

The neighbors are already lining up with projects.

add commentadd comment in the forum

reader comment Comment from: lasersafe1 on Wednesday, October 10th 2012 - 7:47 PM
Super Awesomeness! What is the wall power draw on your cutter? That sure would have taken a whole lot longer in my mill! How 'bout some Ninja throwing stars?

Wednesday, October 10th 2012 - 7:26 PM

Here is a test run of the CNC Plasma Cutter. I am cutting out a PS:One Snowflake Logo.



pl_snowflake.JPG

add commentadd comment in the forum

Wednesday, October 10th 2012 - 4:12 PM

Here is a demo of the floating Z.

The torch is on a carriage that floats on the Z axis. When the Z goes down the torch tip will hit the material, but still allow the Z to move by "floating up" on the axis. As the Z is over traveling a limit switch is triggered. The over travel distance is consistent and known so instead of setting the Z to zero with the limit switch is triggered it sets it to -0.268 in my case. If you listen you can here the limit switch click in the video.

This allows you to rezero before each pierce if you like. The code would look like this.

G28.1 Z0 (Home Z)
G0 Z0.15 (move to piece height)
...turn on torch...wait a bit to pierce...run profile...turn off torch...retract..

add commentadd comment in the forum

reader comment Comment from: macona on Thursday, October 4th 2012 - 2:59 AM
lasersafe1 wrote:
macona wrote:Plasma cutting of aluminum is usually not very pretty.


According to many posts over at CNCzone, this was true until about 5 years ago. Apparently the new inverter controls of the best systems can produce results on aluminum that are very clean with no slag using only air. Accuracies down to .003", which approaches that of laser and water cutting.

The model that Bart has purchased is among these newer types. Perhaps we will soon start to see videos of his cuts. It took me a few months to get my laser "dialed in" to be satisfactory on every type of material. I'm sure Bart is going through the same thing now as he jumps between his many projects.



It is more torch design than anything. Thats is where the innovation is, especially with the hi-def stuff.

Inverter plasma cutters have been around since the 80's with the early Linde/LTec plasma cutters and all of Hypertherm's portlable plasma cutters since the old Max42 have been inverters. Inverter is just a fancy name for a switched mode power supply.

Tuesday, October 2nd 2012 - 11:20 PM

I might try it with both size motors. The only one I really don't know about is the Z. Will the detent torque hold up the torch? I kind of think it will with the drive ratio helping add some torque.

add commentadd comment in the forum

reader comment Comment from: lasersafe1 on Tuesday, October 2nd 2012 - 8:58 PM
Cre8ivdsgn wrote:Seems to make excellent sense to me. I quick look at McMaster says that the pitch diameter of 0.625 yelds a 1.963" per revolution of the pinion. I don't recall your drive versus driven pullies, but a 2:1 ratio would keep torque reasonable for NEMA 17 motors, I think (given probable axis weights).


I'm not seeing the logic behind NEMA17. Price difference is minimal for NEMA23, and why give up the great performance of NEMA23? A few lost steps and you're screwed if you undersize the motor. That being said, have you seen the cute little NEMA 8 motors on Pololu.com for $16? They would be good for building that new "post card laser cutter".
reader comment Comment from: bobt on Monday, October 1st 2012 - 5:49 PM
Bart - Is it possible to move this thread into its own forum group like the Ordbot group as I want to build one of these units as soon as you release any documentation and such. It seems that YOU are the center of the universe with using MakerSlide and implementing it's usage for small shops.

Many Thanks for you efforts with this.

Bob Teeter
reader comment Comment from: Cre8ivdsgn on Monday, October 1st 2012 - 3:15 PM
Seems to make excellent sense to me. I quick look at McMaster says that the pitch diameter of 0.625 yelds a 1.963" per revolution of the pinion. I don't recall your drive versus driven pullies, but a 2:1 ratio would keep torque reasonable for NEMA 17 motors, I think (given probable axis weights).

More importantly the 24 pitch 15 tooth is still a 1/4" bore, making axles easy.

And its pretty easy to simply order from McMaster (I am preuming that is who you went with). Still, Moore does do nice stuff. One company I know has them make a pinion with axle as one solid piece. In quantity, and at least for industrial grade, the pricing can be reasonable for that sort of thing. It isn't reasonable for the one and two piece world of the home manufacturer though!

Still thinking out loud...

Monday, October 1st 2012 - 12:29 PM

I went with the 14.5deg 24 pitch rack and 15 tooth gear. I had a lot of trouble finding a low profile rack and gear with a reasonable shaft size for a good price. I sort of ended up with this configuration rather than scientifically chose it.

It works fine. I probably would have preferred a few more teeth on the gear, but the shaft size quickly grows. Larger gears mesh a little smoother with the rack.

add commentadd comment in the forum

reader comment Comment from: Cre8ivdsgn on Sunday, September 30th 2012 - 8:14 PM
Bart,

Did you end up going with Moore Gear and the 20 pitch rack? My calibrated eyeball is off and the rack shown mounted up against the makerslide seems smaller than 1/2" square. I was wondering if you ended up sticking with the 14.5pa, 20 pitch, 24 tooth pinion that I think you were originally planning on using. I have some helical rack and pinion stuff in my sample box, but the pinion is a 2" pd and is overkill for what I am working on.

Shamelessly picking your brain!

Thanks!
reader comment Comment from: orcinus on Sunday, September 30th 2012 - 3:27 PM
reader comment Comment from: lasersafe1 on Sunday, September 30th 2012 - 1:45 PM
macona wrote:Plasma cutting of aluminum is usually not very pretty.


According to many posts over at CNCzone, this was true until about 5 years ago. Apparently the new inverter controls of the best systems can produce results on aluminum that are very clean with no slag using only air. Accuracies down to .003", which approaches that of laser and water cutting.

The model that Bart has purchased is among these newer types. Perhaps we will soon start to see videos of his cuts. It took me a few months to get my laser "dialed in" to be satisfactory on every type of material. I'm sure Bart is going through the same thing now as he jumps between his many projects.
reader comment Comment from: macona on Friday, September 28th 2012 - 11:20 PM
Plasma cutting of aluminum is usually not very pretty. You need a lot of power to get a clean cut and even then there is a lot of cleanup. When cutting Steel or Stainless you get a reaction with the air that burns the metal away.With aluminum it melts it and the air just blows it away. There are gas mixes for aluminum but standard air plasma systems cant use them.

If you need a whole lot of a part try a local water jet place. I have found it to be very reasonable.

Friday, September 28th 2012 - 5:43 PM

I did buy the Hypertherm 45. I have heard it is the best for fine detail. When you dial in the power and feed rate it is quite amazing. I am only having a little dross where I can't get the feed rate I want yet like a fine saw tooth profile. I think I just need to play with acceleration and add in a little CV.

I saw this torch comparison at a CNC plasma supplier. It puts Hypertherm at the top and the cheap Chinese ones at the bottom.

http://www.gotorch.com/plasma-cutter-comparison/

add commentadd comment in the forum

reader comment Comment from: lasersafe1 on Friday, September 28th 2012 - 5:14 PM
Perhaps I missed it in your log. Did you end up with the Hypertherm? I routinely cut .035" aluminum plates. I normally stack them 5 at a time and use my mill with a 1/8" bit. This thread has got me thinking about Plasma. I would be curious if anyone has experienced any success with units costing about $1000 less than the Hypertherm since my thickness requirement is so low. I looked up "thin aluminum plasma" on CNCZone and most everyone is mentioning Hypertherm.
reader comment Comment from: Cre8ivdsgn on Thursday, September 27th 2012 - 4:15 PM
Bart,

My intention was only to point out that because the axis is now surrounded, it forces you to build the carriage to the X before mounting things to the Y. It was also to point out that cable carriers which would be mounted along the X would no longer work and you would have to go to a tether loop that now would move in X and Y (and maybe Z).

Its also amazing when you consider how strong 1/8" aluminum tube stock is, especially when you preserve corners when notching. I make a cover out of 1/8" aluminum that is 24" x 5", and sometimes the material has a bow to it (when cut from a larger sheet now that Kaiser Aluminum no longer supplies 6061 "flattened" or "stretched" sheet stock - or at least I can't get it). I cannot get the bow out or even change the shape when I stand on it! (Ok, its been a while since I tried that last bit and I might have been lighter then!) I'd try this now, but the covers I have are all painted and only bad things could happen if I stood on them now.

Certainly making the hole small gives maximum strength. I am only trying to point out that turning it into a slot and yet preserving the corners would give a minor reduction in strength but now return the functionality that would allow a cable carrier along the axis or encoder or whatever.

Well, and I think I got caught in the trap of thinking out loud in a blog post. Can outright senility be far away?

Thursday, September 27th 2012 - 3:44 PM

The idea can be sized to what you need. I use an overhead cable carrier. Keeping the window as small as possible helps the strength.

If you want larger rails there is a 6" x 3" x 3/16" rectangular tube at some suppliers.

add commentadd comment in the forum

reader comment Comment from: Cre8ivdsgn on Thursday, September 27th 2012 - 2:50 PM
The only downside is that now his box surrounds the X axis (or Y depending on what you call it) and precludes mounting non-traveling stuff to it (cable carriers, encoders, or anything to stiffen the axis for less deflection on longer travels).

The other possible issue is that now there is definately an order of operations to assembly.

Now that I have listed the downside, let me say that cutting parts down to one piece is more than worth all the sacrifices! And plenty of rigidity would be retained even if the hole for the X axis (or Y) were changed to a slot.

My 2 cents and folks here are kind enough to put up with this sort of thing! (Thanks again, Bart for maintaining the site!)
reader comment Comment from: Enraged on Wednesday, September 26th 2012 - 10:17 PM
Very slick solution. We did a similar idea for our big router, we ended up using 6"x4" aluminum rectangle and carved holes and such into it. Much easier and faster than cutting plate and welding/bolting.

Wednesday, September 26th 2012 - 6:01 PM

Here is the Z axis change I was talking about. As I was redesigning, I realized I was trying to create a structural box shape using plates and spacers. It occured to me to just find an existing full box shape. I decided to use this 6" x 2" x 1/8" wall aluminum rectangular tubing.

The resulting box is many times stronger than my other design. The machining took only about 1/2 hour and that included the programming.

It looks a little large in the rendering, but it the size of the box is only 6"x2"x7". The rack drives are held in place with shoulder screws and the tension is provided by a spring.

Here are some photos of the box itself and then there is a quick rendering of what the assmebly will look like. You can't see from the direction the rendering, but the Z has a floating head now to all quick automatic zeroing.

sq_tube_01.jpg

sq_tube_02.jpg

sq_tube_03.jpg


Untitled Project 94.jpg

add commentadd comment in the forum

Monday, September 24th 2012 - 6:02 PM

Thanks,

A lot of stuff was temporary as I learned what makes a good plasma cutter. I am cleaning it all up now.

I am making some changes to the frame to make it a little more portable and user friendly. I will be adding a removable water tray under the cutting area.

The X Z parts were just made from laser cut acrylic. I was planning on changing them to aluminum, but as I was optimizing the design, I had a cool new idea I want to try first. I hope to be able to show that soon.

The rack drives are working great and I am looking at cleaning up the design and releasing the source files. The pinion gear is a little expensive and a bit of a PITA to mount to the shaft. I have some ideas I want to try to solve that.

add commentadd comment in the forum

reader comment Comment from: loopingz on Monday, September 24th 2012 - 4:53 PM
That is looking super pro!
reader comment Comment from: automationtech on Monday, September 24th 2012 - 12:52 AM

Sunday, September 23rd 2012 - 6:44 PM

I repackaged the electronics today. I had the electronics all mounted to an acrylic plate, but this will make a little more robust.

I started with a Keling (ATI) basic electronics enclosure for the Gecko G540. It cost about $79 and that includes a power entry connector, a fuse (did not use), a power outlet (did not use), a lighted power switch, an e-stop button and a 9 pin D connector for the limit switches.

I only had to add a cutout for the torch height display to the front. Everything else bolted right in. The only kludgy thing in the drawer is the two 12VDC isolated power supplies. I used wall warts rather than buy two new supplies. The torch height system needs these to stay isolated from the torch.

I did not use a fan. I doubt I will need one and the power supply has one that turns on if it gets hot. with the well ventilated chassis that should stir out the heat.

DSC00283.JPG

DSC00284.JPG

DSC00285.JPG

DSC00286.JPG

add commentadd comment in the forum

Monday, September 17th 2012 - 12:41 PM

sun and lasers


Thanks for the info. I was assuming that, but was going to check with a few camera experts. I had heard that the sun would damage them, but I have seen a lot of good plasma cutter videos, so I was just being careful.

I'll do a video soon. The wiring needs a cleanup between the controller and the plasma first. I also have some design changes I want to make frame. I want to make the whole system a little more portable and make the water trough a easier to slide in and out.

add commentadd comment in the forum

reader comment Comment from: macona on Monday, September 17th 2012 - 7:06 AM
It wont hurt the camera. Only two things that can really damage a camera, the sun and lasers.
reader comment Comment from: BenJackson on Monday, September 17th 2012 - 1:00 AM
bdring wrote: I don't know if it is safe to film with my good camera.

I've seen people film TIG arcs with no apparent harm to the camera, but I know how you feel.
reader comment Comment from: gavztheouch on Sunday, September 16th 2012 - 10:00 PM
Cool!

Sunday, September 16th 2012 - 8:42 PM

I finally cut my first full CNC control plasma part.

The major hurdle was my air compressor. It was horribly underrated and could only cut for about 20 seconds before the pressure dropped. I was lurking on CraigsList for a while, but then the local home center put a decent one on sale. I did not want a diaphragm type because they are so loud. I bought this one for a little over $300. It is rated for 5.5 CFM at 90PSI.

sandborn_ac.jpg


This is a 2.0 inch square cut on the line out of 16Ga steel. It was hand written G-code and I just guessed at a bunch of values like pierce delay, pierce height, tip voltage, current etc. I need to read the plasma manual and write a post processor.

The piece is very square and the X and Y dimensions are within 0.002" of each other. The cut was "on the line" so the kerf was not compensated for. The kerf is about 0.070", but I should be able to improve that using by using all the right values.

It was quite fun to watch with sparks and smoke all over the place. I don't know if it is safe to film with my good camera.

first_cut.jpg

add commentadd comment in the forum

Wednesday, September 5th 2012 - 12:50 PM

The brackets were custom made by a friend. I only had a handful made. Attached is the preliminary drawing.
Attachements...
plasma_ms_angle_brkt.pdf
(34.01 KiB) Downloaded 318 times

add commentadd comment in the forum

reader comment Comment from: evokanivo on Tuesday, September 4th 2012 - 11:48 PM
bdring wrote:This detail shows how the MakerSlide is bolted to the 20x80 Misumi piece. These "sister" brackets are mounted above and below. They do not bother the wheels. The Makerslide is plenty strong at this length, but I plan to make it a lot longer and I like the way the 20x80 will shield the Makerslide from slag.


Would you mind sharing a source for those brackets, assuming you didn't make them? All the ones I've found so far get in the way of the v-rail.
reader comment Comment from: Digitalmagic on Sunday, September 2nd 2012 - 2:13 PM
bdring wrote:I suggest we break this off as a new topic if there is another post.

Here we go: http://www.buildlog.net/forum/viewtopic.php?f=30&t=1511

macona wrote:...It will help against wear but it is not going to do much to help if you use metal wheels. The layer will eventually crush down since the aluminum under is still soft.

Usage limits are everywhere. I wonder how it compare with vanilla Makerslide/Delrin wheel.
If this extend its life, this is already a good point, even for light applications.
Choosing the optimal metal V-wheel is perhaps a factor.
reader comment Comment from: macona on Sunday, September 2nd 2012 - 2:47 AM
Hardcoating is just a thicker anodized layer, as you said, aluminum oxide, also known as sapphire.

It will help against wear but it is not going to do much to help if you use metal wheels. The layer will eventually crush down since the aluminum under is still soft.

Where hard coating would really shine is in a slider system where you would use something like UHMW instead of wheels. More friction than wheels but more rigid.

Cutting would be best on a cold saw and get through the layer quickly so you are not rubbing on the layer. A good blade should last quite a while.

Saturday, September 1st 2012 - 1:00 PM

As I understand it hardcoating is a super hard aluminum oxide. I think that is basically what they make grinding wheels to sharpen carbide teeth with. I don't think you will find an effective blade that will last. You might find that certain orientations of the cut do less damage to the blade.

I suggest we break this off as a new topic if there is another post.

add commentadd comment in the forum

reader comment Comment from: Digitalmagic on Saturday, September 1st 2012 - 7:19 AM
Several ideas come in mind about the cutting process for these hard coated rails.

Some DIY buyers will follow exactly what you said about carbide blades, wasting them frequently, or using cheap one.

But in the Inventables shop concept:

- Carbide blades could be sold at a good price, having been validated, suitable for cutting these.
- As an extra, a cutting service at desired length could be offered.

If a kit using these hardened rails could be proposed, I think the most obvious would be a "Black ShapeOko", hardened, perhaps a little bigger, perhaps reinforced like you did on your plasma cutter rail assembly. This would offer increased cut load, and would open new horizons.
Of course, metal V-wheels, perhaps a T-slot bed, etc...

Thursday, August 30th 2012 - 12:40 PM

Zach talked to me about that. We are looking into it.

There is the issue of cutting. It pretty much wrecks a carbide blade after one or two cuts. It might be best to start with a reference design and have the material pre-cut to that length. You could always just buy a cheap $8-$10 blade and scrap it after using it for the project.

add commentadd comment in the forum

reader comment Comment from: Digitalmagic on Wednesday, August 29th 2012 - 6:12 PM
This supplier is very interesting, they talk about" Type II" and "Type III" hard-coating (what's that?), with great explanation: http://www.fortwayneanodizing.com/hardcoat/properties.htm

I exchanged with Zach to study a possible option of hard coated Makerslide and metal V-wheel on Inventables.

Wednesday, August 29th 2012 - 12:18 PM

I used http://www.automaticanodizing.com/

I don't have anyone to compare them to, but they were very responsive and turned the job around quickly. They are pretty close to me, but it was easier at the time to just use UPS.

add commentadd comment in the forum

reader comment Comment from: Digitalmagic on Tuesday, August 28th 2012 - 8:42 PM
bdring wrote:You generally just ask for hard coating and specify the buildup (thickness). Maximum buildup is about 0.003" and is what I used. The places I have found have a minimum order of about $150 which would get me about 25 pieces.

Bart, thanks! Am definitely interested in that.
Could you please tell me the company offering this? I already have a bunch of rails to submit.

Tuesday, August 28th 2012 - 12:08 PM

The process is known as hard coating. Only about 10%-20% of anodizers in my area do it. It requires a refrigerated tank and uses chemicals that have special disposal requirements.

You generally just ask for hard coating and specify the buildup (thickness). Maximum buildup is about 0.003" and is what I used. The places I have found have a minimum order of about $150 which would get me about 25 pieces.

add commentadd comment in the forum

reader comment Comment from: Digitalmagic on Tuesday, August 28th 2012 - 11:21 AM
@Bart

i would like to apply the same anodizing/hardening process to my 1200mm Makerslide.
The goal is use steel V-wheel on it.

Did you just have them processed with a "vanilla" electrolytic anodization? Or a specific surface additive process hard coated them, like adding a layer?
In other words, which exact process is needed, and how does it cost?

Thanks!

Image
reader comment Comment from: Cre8ivdsgn on Thursday, August 2nd 2012 - 2:38 PM
Bart, I was thinking of using the cam in place of the belt tensioner screws. So you could add one more hole when you initially cut the UHMW material for a screw about which the eccentric could turn. Not necessarily for rack preloading....

I wish I could make it out to Evanston! I lived there for a few years - while in school and afterwords. For a while my better half and I had an apartment above a bread bakery on Central. Mornings were pure torture!

Have a good time!

Thursday, August 2nd 2012 - 1:48 PM

I doubt the motor would slide after the bolts are tightened, but it was hard to use a screw driver and wrench while keeping the belt in tension, so I added the jacking bolts.

I really dislike secondary operations too. The holes location is not critical at all so I just hand drilled them without a press or vise. I don't really consider that a painful secondary operation. If the parts are printed, it is not an extra operation.

I have played around with some cam ideas for setting the rack engagement, but have not found anything I like. The distance is large because it is real handy to have the gear fully disengage. A cam with that range takes a lot of space. As long as I am going to all the trouble, I want to spring load it, which a cam can't do.

I finished all the home switch wiring last night. All the switches are wired in series on one input. It took a little RTFM to get the slave axis to play nice with one switch. It works and Mach3 can home it all on one circuit. I ran some G-Code too.

That is it until after Maker Faire. I have to do a little dis-assembly and flip it on it's back to get it into the van.

add commentadd comment in the forum

reader comment Comment from: Cre8ivdsgn on Thursday, August 2nd 2012 - 1:26 PM
Bart, let me preface these comments with "I am thinking out loud here simply because I am dealing with some of these design issues myself right now."

I can see where the motor could slide - especially since the motor mount is made from UHMW! The jacking screws are a good safeguard. I confess I am not a huge fan of introducing secondary operations on a part (the 90 degree tapped holes on the motor mount). You might have used an eccentric (no secondary operation), but on UHMW, even that might not be enough. The screws are likely most reliable.

I've been looking at Moore Gears and their rack supplies. Not listed is helical rack and pinion stuff, which is amazingly smooth. Because of the 30 degree helical engagement, you always have more than one tooth against the track. Its expensive though - at least as far as homebuilt stuff goes. For this machine design I am working on now, I think I will fall back to straight rack and pinion. I can go back to helical if I need to and can justify the expense.

I did have one goofy thought though. If you consider a drive mechanism for the rack (perhaps something similar to the unitized drive you have come up with, but that isn't important in this example). Now mount this to a flat plate made of thin G10 (or perhaps even the carbon fiber material I thought I saw on one of your designs) whose surface is parallel to the rack. Firmly bolt one end to the carriage and the other end to the drive mechanism and pick a position that forces the G10 to deflect a little. The rigidity of the G10 (given an approrpiate thickness) might be sufficient to keep the pinion and the rack in tight engagement without twisting and even 0.5mm deflection should be sufficent to keep the pinion engaged. At that little deflection even aluminum would work, I think. (I had part of this conversation with the folks at Moore - they said that if we could spring load the pinion's engagement with the rack, we would have the best chance of achieving the positional accuracy they could cut to.) On the other hand, this preloading might be a bit hard on the V wheels of the MakerSlide.

This is just a bit of thinking out loud.

Wednesday, August 1st 2012 - 5:32 PM

I added those bolts I use to tension the belt. The motor is attached firmly, but can still slide under pressure. The bolts are used to tension the belt.

rack_drive_tensioner.jpg


Belts are a lot more forgiving than racks. I was thinking about doing something similar with the rack using spring plungers instead of bolts. Some have good travel and can have a lot of force.

spr_plng.png
spr_plng.png (12.61 KiB) Viewed 1876 times

add commentadd comment in the forum

Wednesday, August 1st 2012 - 3:48 PM

The tensioning is not ideal, but works fine. I have some changes in mind, but I want some of the parts to be plasma cut, so I am holding off rev 2 until the machine can do it.

The tensioning of the belts took about three hands to do so I added 2 tapped holes through the sides of the UHMWPE (not shown in rendering). Bolts through the holes push the motor to tension the belt....it works great. You can't see it from the rendering, but the motor is in a pocket so the shaft can stick out enough.

I use a USB smoothstepper on my big router. It works best when you have a really high step rate (I have 28800 step/in) . I like it, but I have had some problems with it. If the computer cannot respond fast enough via USB, the smoothstepper locks up and gives a error about the USB speed. Mach3 has no idea where it stopped. You have to restart Mach3 to fix it. I bet the Ethernet version has the same issue. Get a fast computer and try not to watch YouTube while running a job :lol: Bottom Line: I like it for the router.

Obviously the parallel port is a dead end, but if you use special applications like close loop torch height control, that has to be programmed into the hardware of a device like a smoothstepper or it will not work.

I wanted to use an off the shelf solution to start with because I have never used a plasma. After that I have a group that wants to help do an open source replacement for everything.

add commentadd comment in the forum

reader comment Comment from: Cre8ivdsgn on Wednesday, August 1st 2012 - 3:18 PM
The pivot mechanism of your drive assembly - it looks as though you will adjust it so the pinion engages the rack and then screw it down. It does not look as though you are planning on spring loading the engagement. The loads are pretty light on this and so that will likely work well. I'm not sure how well this would hold up if you used this design on a router table. I just have no idea of how well the rack and pinion setup will wear.

Just to confirm: Mach3 through the parallel port? I just feel as though I am missing the boat and really everyone else seems to understand something else entirely! Along with a large table design, I am also working on retrofitting a Sharp knee mill for CNC operation and I've been wondering what control route to take.

Do you have an end goal in mind for the plasma table? Had you been thinking about making enclosure boxes or something similar that sent you down this road? I'm curious because i'm wondering how this fits in with the router you currently run!

Wednesday, August 1st 2012 - 2:22 PM

Here is a quick exploded view rendering of the rack drive. This is used on 3 out of the 4 motors and a very similar one is used on the Z. These were cut from UHMWPE, but they were designed so that they could be printed too. It can use NEMA17 or NEMA23 motors. I used smallish NEMA23's.

The ratio is 2.4:1 (20 to 48 tooth). The pinion is a 24 pitch 15 tooth spur gear.

The whole assembly rotates into the rack to set that distance.

rack_drive.jpg

add commentadd comment in the forum

Wednesday, August 1st 2012 - 1:37 PM

What are you going to use to drive the gecko steppers?


Parallel Port

I am starting with the CandCNC LCTHC Torch Height Control. It cannot use a controller like theSmoothStepper because the latency is too slow for effective closed loop feedback on torch height control with Mach3. The Smoothstepper is fast, but it's actual communications rate with Mach3 is quite slow (40 Hz I think).

add commentadd comment in the forum

reader comment Comment from: Cre8ivdsgn on Wednesday, August 1st 2012 - 1:27 PM
Bart:

Stupid question here. What are you going to use to drive the gecko steppers? I ask because I have been looking at the Smooth Stepper stuff which has higher pulse rates than some other boards. I like the idea of an ethernet interface, too.

I am guessing you plan on using Mach3?

I know these are noob questions!

Thanks

Wednesday, August 1st 2012 - 1:09 PM

I have been a bad buildlogger on this project. :oops:

I have made a lot of progress, but am not documenting it well. I have been so busy, that I just try to sneak and hour here and there to work on it, but don't have time to take pictures.

My goal has been to debut it at the Mini Maker Faire in the local suburb of Evanston, which is this weekend. I have done a big push to get it cleaned up and ready to go over the past few days.

I have had full three axis motion for a while, but not in a clean well wired way. That is done. The cable leading to the plasma is quite thick and heavy because it also carries the air hose. I did not want a huge cable carrier on a machine this size so used an overhead arm similar to what Torchmate uses. Mine rotates which works well to limit the slack required.

I have two slaved motors for the Y to drive the gantry. I did that because it is an easy way to insure squareness on such a light machine. I placed two hard stops at the back to square it. If it needs to be squared you just manually push it against those stops then engage the steppers. The 2.4:1 gear down on the rack drives has a lot of detent torque, even when off so it stays pretty square. Mach3 has a feature where you can use two limit switches to automatically square it when zeroing the axis. I may try that, but the plasma controls really eat up the inputs, so I may need a second parallel port.

The speeds are quite fast. I run 1500 ipm on the X, 800 on the Y and about 200 on the Z. It will run faster, but the acceleration is the limiting factor. The whole thing is on wheels right now. Fast accelerations tend to make try to roll around. Once I find the final home for it, I will loose the wheels.

I'll try to take some pictures tonight.

add commentadd comment in the forum

Sunday, July 22nd 2012 - 3:42 PM

I have made slow but steady progress. I have three axis motion on the machine and am now working on the final wiring. The only item I still need to buy is a bigger air compressor. I did some test cuts by hand with my little pancake compressor. It cuts but immediately cycles on and starts loosing pressure. It uses a lot of air and the air stays on for quite a while after the arc is off. This is probably to cool the tip.

Here is a shot of my interface panel. It still needs a bit of wiring.

pl_panel_front.jpg

pl_panel_rear.jpg

add commentadd comment in the forum

reader comment Comment from: bobt on Sunday, July 22nd 2012 - 1:27 AM
I know that Bart has been very busy lately but is there an update on this?

Bob Teeter

Thursday, July 5th 2012 - 10:56 PM

While I was on vacation my LCTHC (Low Cost Torch Height Control) arrived. This came from CandCNC.com. Eventually I hope to make an open source simplified version of this, but I need to get a experience using a CNC plasma first.

In addition to a bunch of cables and power supplies it came with these parts. I also ordered a Gecko G540 interface board. I had a G540 left over from the original laser cutter.

This is the interface panel. you control it via a rotary/push knob.

LCTHC_front.JPG


This is the rear view of that panel. It is really quite simple. Most of what you see is simple the back of the LCD.

lcthc_back.JPG


This is the Gecko interface. It makes the wiring a lot simpler.

gecko_adapt2.JPG


Here it is attached to the G540

gecko_adapt.JPG


This is the sensor for the plasma. At one end is a voltage dividor and the other end has the trigger and a few other interfaces.

torch_intfc.JPG

add commentadd comment in the forum

Friday, June 1st 2012 - 11:47 PM

The frame is now built. It is quite strong, but not too heavy.


plasma_frame2.JPG


plasma_frame1.JPG




This detail shows how the MakerSlide is bolted to the 20x80 Misumi piece. These "sister" brackets are mounted above and below. They do not bother the wheels. The Makerslide is plenty strong at this length, but I plan to make it a lot longer and I like the way the 20x80 will shield the Makerslide from slag.

DSC00219.JPG

add commentadd comment in the forum

Wednesday, May 30th 2012 - 1:38 AM

I am working about three projects at once, so I can only give a little time to each right now.

I finished fabricating all the gantry parts. I am pretty impressed with the rigidity. I would not try to route with it, but I think it is good enough for plasma. It has good stiffness and resistance to twist. I will get all the parts hardcoated after I am happy with the design.

I noticed I can flip the XZ carriage. One way will be stiffer and the other is farther from the slag. I'll see how it goes.

gantry_01.JPG


Here is a more recent rendering.

Untitled Project 90.jpg

add commentadd comment in the forum

Friday, May 25th 2012 - 6:01 PM

I have a couple backup plans. The rack actually adds way more stiffness than I had expected. Hopefully it will work as designed.

I'll learn a lot soon.

add commentadd comment in the forum

reader comment Comment from: Enraged on Friday, May 25th 2012 - 5:09 PM
Looks good. have you thought about running two Makerslide extrusions back to back for the gantry, basically making a 40x40? Or you could always get the machine to cut it's own gantry stiffener plate, and make a "composite" steel/aluminum gantry.

Friday, May 25th 2012 - 4:29 PM

I ordered enough parts to begin testing the basic mechanical platform for the plasma cutter. I am going to limit the size to 24" x 48" (work area) for now. The primary thing I am worried about is strength of the gantry. That is being built full size. The rest is scaled back. My biggest concern is twist of the gantry. It did some experiments with some MakerSlide of that length that look promising, but I need to get a real feel for it.

I might need to double up on some of the corner brackets, but it would be nice it would be able to cut well enough where it would be able to make its own extra strength brackets and plates. I will also add some feet or wheels to the design.

Usually the work surface needs to be replaced on a regular basis, so it is designed such that the cutter can make it's own "consumable" parts.

I will probably use cable carriers, but I want to investigate a simpler overhead arm type like you see on "Plasma CAM" type machines.

The final design will incorporate a water table or down draft system to cut down on the smoke. I talked to many people at Maker Faire about that and they all liked the water table method the best.

plas_09.jpg

plas_10.jpg

add commentadd comment in the forum

Wednesday, May 23rd 2012 - 10:48 PM

The entire stepper/gear assembly rotates into the rack. It has provisions for spring loading, but right now I am simply bolting it in position. I need to get a feel to the tension I would use.

add commentadd comment in the forum

reader comment Comment from: gavztheouch on Wednesday, May 23rd 2012 - 10:20 PM
Nice project! Is your spur gear preloaded/sprung against the rack?

Wednesday, May 23rd 2012 - 2:33 PM

I got a quote back from Moore Gear for racks. The quote was for (2) 6' foot 20 pitch racks. The racks are 0.50 square, which is wider than will fit with my eccentric spacers but it is probably only a minor modification.

The cost is $113 delivered.

add commentadd comment in the forum

Thursday, May 17th 2012 - 5:42 PM

Here is what I have for the X/Z carriage so far. I do not need much Z movement, but it is nice to have a few extra inches or it can get whacked while loading stock.

I have Moore Gear quoting me some racks. I will post the details. Thanks Diane.

rd_z.jpg

add commentadd comment in the forum

reader comment Comment from: diane-mooregear on Thursday, May 17th 2012 - 3:08 PM
We manufacture and sell gear rack commonly used in cnc routers/plasma cutters and other rack and pinion devices as a stock item and regularly sell these in small quantities to independent builders.

If you have any questions or would like a quote, I welcome inquiries at diane@mooregear.com. We accept credit cards and ship domestically and overseas.
reader comment Comment from: mikegreen on Wednesday, May 16th 2012 - 4:52 PM
It will be interesting to see what you find. Cost is an issue with the options that I listed (~$100 USD per 1,000mm for rack). By the way: I just checked, and Misumi does have 980mm M1 counterbored rack if anyone is interested. Not sure why I missed that before. It's just a few more dollars more expensive than the 1,000mm rack without counterbores from QTC gears, but the machining on the ends is different for the two products.

I have been looking at some of the discussions related to MakerSlide and flexing at longer distances. I have not done any tests like I have seen posted, but I wonder how mounting steel counterbored rack to the side of the MakerSlide extrusion affects the flex. Having two 500mm sections of rack mounted onto the t-slot in the MakerSlide certainly appears more sturdy. That's one of the reasons that I did not want to give up on an R&P design.

Wednesday, May 16th 2012 - 4:27 PM

I know a lot of the MechMate router builders use racks from Standard Steel Specialty Co.

http://www.stdsteel.com/

I am waiting on quotes, but have heard some really good prices on their 73 inch racks.

I am using a 15 tooth gear now, I would like to find a cheap 20 tooth 24 pitch 14.5 P.A.

add commentadd comment in the forum

reader comment Comment from: mikegreen on Wednesday, May 16th 2012 - 4:08 PM
On the metric side, I have had good luck with the GEABB1.0-15-10-K-6.35 pinion from Misumi. It's very sturdy. I have mated this with the RGEA1.0-500-Z rack. The nice thing about the rack is that it can be ordered counterbored and fits very nicely onto the t-slot on the MakerSlide. There is a picture in the design ideas section. Since these can be ordered with machined ends, in theory they can be joined together for longer lengths. I have two mounted on MakerSlide, but I'm not sure what I think about that configuration. I can feel it when the pinion crosses the two racks. I need to do some measurements to see if this is a real problem or not. I can't find Module 1 rack longer than 1,000 mm (QTC gears has 1,000 mm without counterbores), which is a bit short for your purpose unless joining them together works out. M1.5 comes in longer lengths. I used M1to keep the width of the rack down to try not to have the carriage too far from the MakerSlide.

Thanks for the pictures. I was about to give up on the rack and pinion idea. In looking at the torque curve for the stepper motor that I have, I understood why direct drive was not going to be the best idea, at least for a router based system. I need to gear down to around 3:1, just like CNCRouterParts has done. Even without the gearing down and only using a 282 oz. stepper from Automation Technology, I agree that it's scary strong. More than I had expected, even at high speeds. I like how you handled the gear mounting. I was thinking of something more complicated and that was what was causing me to consider just going to lead screws for the x and y axis.

It will be interesting to see what you come up with for rack and pinion on the Z axis. I'm still working out the Z axis lead screw configuration for my experiment.
reader comment Comment from: mike on Wednesday, May 16th 2012 - 2:38 PM
te_gui wrote:...how many folks were going to be at Maker Faire this weekend, and if any OrdBots were in attendance.


I'll be there with my ORDish bot -- Looking forward to meeting Bart and any buildlog folk who show up.

--Mike
reader comment Comment from: te_gui on Wednesday, May 16th 2012 - 2:12 PM
Don't want to hijack this thread, but have been wondering how many folks were going to be at Maker Faire this weekend, and if any OrdBots were in attendance. I'll try and introduce myself if I get a chance.

Brian

Wednesday, May 16th 2012 - 12:22 PM

I installed an old NEMA23 motor. I don't know the exact torque, but I think it was around 300 oz/in. The drive system has about 3.8:1 ratio. Is is a lot stronger than it needs to be so I may change the ratio to about 2:1 so I can get the speed I want.

The power is is quite impressive and a little scary. I have not been able to stall the motor by manually holding it back. I think if I mechanically locked it down, something would fail before that motor stalled. I think the weak link right now is the pinion on the shaft. It came without a set screw and I had to put my own in. The hub wall is not very thick. I think the best way to change the ratio would be to get a bigger pinion.

The setup has quite a bit of hold torque with the motor off. This will make my Z simpler. It will not free wheel down if power is lost.

I plan to drive the Y with two slaved motors. A friend runs the fab lab at the local museum. They have a Shopot with a similar setup. He says they square the Y by manually sliding the Y against hard stops with the motors off. That sounds like it would be fine for me.

PS: I am going to Maker Faire this weekend. I will bring my little sample of hardcoated MakerSlide with the rack installed. I will also bring my Quantum. I don't have a booth, so I will be hanging out with Inventables and ShapeOko.

add commentadd comment in the forum

Tuesday, May 15th 2012 - 12:37 PM

Thanks for all the info @macona.

I think I will not use the built in torch rack for the Z drive. It makes it more complicated and limits the usage to the expensive Hypertherm machine. I hope to use the same rack drive module for all three axes. They are simple parts that can be cut from sheet on a router or printed on a 3D printer. I hope to run around 400 ipm on the X and Y.

add commentadd comment in the forum

reader comment Comment from: callcbm on Tuesday, May 15th 2012 - 11:23 AM
I'm presently building a plasma and using the tch from candcnc great service and product. I built a control console to house the pc/monitor/tch my only worry is the electrical noise produced from my aged plasma making thing go funky.
Thanks Randy
reader comment Comment from: macona on Tuesday, May 15th 2012 - 5:02 AM
bdring wrote:While I want to be able to support a really cheap plasma as an option, I am thinking about going with the Hypertherm Powermax 45 with the machine torch option. It has a very good reputation and brings out all the controls necessary for a CNC system onto a connector. Most cheaper systems you have to do some serious hacking. I like the machine style torch too.

The torch has a built in rack, but I cannot find any decent dimensions or specs on it. It would be cool to use the rack, but a more universal Z might be a better idea.

Powermax45mechanizedtorch.jpg


HYP088012.jpg


I really don't see cutting anything thicker than 1/4", so it should be plenty powerful.

I have heard the cheap harbor freight type plasma have come a long way recently and have a decent pilot arc start and don't blow out all the electronics in the room every time they start anymore.


good choice on the Hypertherm. I used to be a service tech for welders and plasma equipment. Hypertherm is still made here in the US by an employee owned company. They really make the best machine out there.

A friend has a Hypertherm Max 200 for a cnc machine, though that might be too big for you! It will cut 2" steel.

The torch is made to pretty standard specs on the OD and the pitch of the rack. They are, for the most part, interchangeable with Oxy-acet torch heads for gas burning. It is mainly used for adjusting the torch when you set it to a 45 or something to cut a bevel. Also us used for gross height adjustment, say you need to cut a piece in a structural shape piece of steel.

Most lifters use a generic ball screw, nothing fancy, just the cheap black rolled screws. One other thing a lot of people forget is a collision stop. This stops the machine if the torch runs into something, which can save a lot of money in the long run. They are very easy to build. The simplest uses air. Basically two plates are sandwiched together with three steel ball in a triangle. The balls (Or cones) are attached to one plate. On the other plate there are holes with an o-ring seal on the surface where the balls or cones sit into creating valves. The holes are all paralleled together by either small hoses or the plate is drilled to create a manifold. The two plates are held together with a through bolt, spring and a nut to adjust the tension. This all comes together to create something like a touch probe for a cnc machine but pneumatic. On the line that feeds the 3 valves is a low pressure regulator and a pressure switch. The switch is just a simple on/off that opens at 1 or 2 psi and the regulator is set just above the pressure needed to close the switch.

Operation is simple. During normal operation all the valves are sealed. If the torch is hit from any direction one or more of the valves are opened and there is a pressure drop at the switch which opens and triggers a stop.

This can also be used to help set the initial height. The letter drops the torch down till it senses a touch with the collision detection and then moves it up a set distance. Some of the commercial height controllers are very, very simple. They are nothing more than a DC motor on the lifter screw. The control box after touch off just sends power to the motor for a short predetermined period lifting the torch. The actual distance is not very critical. Then the torch fires and the box just uses a comparator to control the motor up and down. There is a basic filter to smooth out the movement and some thumbwheel pots on the front to set the voltage for the comparator. Often you do need to install a voltage divider in the plasma cutter which is nothing more than a power resistor and a filter cap.

Tuesday, May 15th 2012 - 12:51 AM

The next bit of special electronics you need is a torch height control system.

As I understand it you could create a system without active torch height control, but it would not do the best job and would probably eat through the consumable parts of the torch.

Typically, the THC has a mechanical or other sensor to determine absolute height over the work piece to set the piece height. One running it goes into a voltage monitor mode. The voltage is proportional to the distance so the circuit can determine the height using the voltage and adjust as it goes.

The ones I am looking at just send a move up/move down signal to a Mach3 type program and that takes care of the actual movement.

I have a good friend who wants to make an open source design for this, but for starters I will buy one.

I like this design from Cambell

LCTHC-Detail-Text-small.png


http://www.candcnc.com/LCTHC.html

I also saw this on eBay, but could not find a lot of people using it.

ebaythc.JPG

add commentadd comment in the forum

Tuesday, May 15th 2012 - 12:27 AM

While I want to be able to support a really cheap plasma as an option, I am thinking about going with the Hypertherm Powermax 45 with the machine torch option. It has a very good reputation and brings out all the controls necessary for a CNC system onto a connector. Most cheaper systems you have to do some serious hacking. I like the machine style torch too.

The torch has a built in rack, but I cannot find any decent dimensions or specs on it. It would be cool to use the rack, but a more universal Z might be a better idea.

Powermax45mechanizedtorch.jpg


HYP088012.jpg


I really don't see cutting anything thicker than 1/4", so it should be plenty powerful.

I have heard the cheap harbor freight type plasma have come a long way recently and have a decent pilot arc start and don't blow out all the electronics in the room every time they start anymore.

add commentadd comment in the forum

reader comment Comment from: tylerv on Tuesday, May 15th 2012 - 12:02 AM
Bart is coming up with interesting projects faster than I can follow.. :)
reader comment Comment from: mattrsch on Monday, May 14th 2012 - 11:46 PM
My brother and I have been talking about building a plasma cutter for a while now, so I will be watching this thread with great interest :D

Monday, May 14th 2012 - 11:28 PM

For my next learning experience I thought I would tackle a plasma cutter.

I am going to use hardcoated Makerslide with steel V wheels. I have a bunch of angular contact wheels left over from my original laser design. I think they will do better in a harsh environment.

I am going to start with a 24" x 48" table area so that is stays small and does not take up too much room in the shop. Once I am happy with the design I will scale it up to at least 48" x 48".

I cut a piece of hardcoated Makerslide on my chop saw. Normally the saw just slices right through. This time it was a bit harder to do and it sparked. It never sparks when cutting aluminum. I was not sure what was sparking, but now I think it was some carbide from the saw teeth. The blade is quite a bit duller now than it was. I will definately be cutting and tapping the rest of the pieces before they are hard coated. That stuff is hard!

rack_drive_01.JPG


I used JB weld to glue the rack in place. It was easy to install because it registers on the inside corner of the Makerslide V. The rack I used is a 24 pitch 14.5° pressure angle 24" long piece I had laying around for a while. I used a 15 tooth pinion gear. I found that larger diameter gears tend to ride smoother on the rack. This was the largest 1/4 shaft pinion McMaster had. 1/4" shaft fits all the other pulleys I had. I uses a 20/48 timimng pulley setup from laser parts I had. The drive head swings to adjust for rack engagement. When properly engaged I cannot feel any backlash. I am sure there is some, but it is prbably below the resolution of what a plasma can do. The design will fit NEMA17 or NEMA23.

Here is a rendering of the basic rail design. The Misumi 2080 extrusion will add strength plus shield the v rails from slag spray. This is the Y axis. I will use the same drive for the X axis, but I doubt I will use the 2080 on the X.

rack_drive_01.jpg


Plasma Cutter Choice: Coming Soon

Torch Height Control Choice: Coming soon

add commentadd comment in the forum

About Us | Contact | 2010 BuildLog.Net