CNC Plasma build 1.2x1.2

[WillyInBris]

Ok guys thought I would chuck this up just to give you an idea on how I am doing things with this build.

I have given myself 4 weeks full-time to get it up and running then probably another two to get the kinks out.

Table to have a 1.2x1.2 Cut area di did look at longer but I don't really see the use with it, if I need to make it longer I ma builing this in a way that un upgrade could take place with some new rails etc.

So first off I have been quoted turn key plasma system a well known brand and would get no change out of 25k if it came in at 15k then I would have paid it but really that type of investment 25k on a startup just wont cut the mustard.

I will say though if everything does go to plans and this works out well then the above may change.

Table will be a heavy built table Tully mostly is completed from something I was building two years ago.

I am going to use turn key systems in it.

Hyper therm Plasma Cutter 45, purchased eta 1 week.
Snapcut Torch Holder purchased this week
Z axis from my first build from mothballed stock

Y Axis to be made from 8020 should have quote tomorrow
1 x 25 Linear guide from mothballed stock (needs a clean up)
Drive will be ball screw from from mothballed stock

X Linear guides 20mm Need to purchase will try with the rails first though and upgrade to guides if needed.
Drive R&P Need to purchase

Y Axis to be made from 8020 should have quote tomorrow
1 x 25 Linear guide from mothballed stock

Water tub will get that quoted when I am up and running may make one myself see how I go.

At the moment I slapped together a test unit on some Chinese steppers I had sitting around all working quite well but needs to be better more or less a test bed for some of the parts I had.

I am looking at a complete plug and play system from candcnc, or a G540 plus there DTHC there are other THC's around like the Proma that cost 1/4 of the THC unit alone but the candcnc gear looks the goods from the reading I have done so far I am guessing because I will go over the 1k mark I will have to pay related taxes as well on the import so whilst the complete system would be excellent i may just go there DHTC unit coupled with a G540 or a G200 series may have to have a chat to Homann and see what he thinks, really up in the air on this.

Mach 3 for the CAM

This is the test unit, my son loves it lol.

Slap.jpgControl.jpg


I am looking to make a gantry like the following in the attachment.

[WillyInBris]

One of the things I have seen on the Z axis is a floating Z axis, this was quite simple to do with the actuator I have, all I did was move the non driven pillow block
to the upper end of the actuator this is what the SnapCut will connect to and add a couple of springs the driven block will then lower and touch to zero out this happens a lot on a plasma table and this should remove any stress from the torch.

P6080061.jpg


Only one problem I didn't think of and that was the bloody bearings when I removed the block to put on the upper side they went everywhere lucky enough my Son thought it was a fantastic game to pic them all up
lol I reloaded it over a sink with the plug in.

The snap cut is an Aussie invention that has a magnet that will allow the torch to break away if it fouls on some metal (to replace a Hypertherm machine torch is bloody expensive) whilst I could have made my own it was quicker/cheaper just
to buy something that was made for purpose another part I don't need to worry about.

18753649_hy7b.jpg18753668_egtk.jpg18753675_v8vb.jpg

probably see if I can add a switch so that if this does happens it stops as well.

as you can see I am quite scared about damaging a torch comes from the days of I guess ramming a Router bit into the table.

[beefy]

Be careful using a leadscrew on the gantry (Y axis). Check out your maximum expected feedrate/cut speed. Look at your Hypertherm charts, see what is the thinnest metal you might want to cut, and which nozzle you'd want to use. Bigger nozzle = faster cut speed. You can then work out how fast that ballscrew would need to revolve (based on its pitch) to give you that linear speed on the Y axis. Leadscrews can whip when revolved too fast. An alternative is to keep the leadscrew stationary and revolve the nut around it. Most tables however just use R&P on the gantry too.

The Candcnc is a good system. I have an older MP3000 system from them and I wanted to upgrade to the latest ethernet Smoothstepper system they have. It has a much faster torch height control than the parallel port systems but I found one little downfall in it that made me put off buying it for now. When cutting holes it's beneficial to be able to turn of the torch near the end of cut but before motion stops. This helps prevent a divot in the circle but only if the torch switches off WITHOUT a pause in motion. Unfortunately it cannot do this (yet). I even contacted the guy at Smoothstepper to ask about this problem. He initially said why do I think it can't so I replied and said see the thread on the Candcnc forum, where they have announced this issue is more difficult to resolve than they first realised. The guy at ESS never replied after that so I took that as him saying it doesn't work. I wanted to use his system in my own plasma electronics/THC design but I'm sticking to a basic parallel port design if he can't even reply to my questions.

If you intend to do thin metals with the associated fast cut speeds then the latest ESS system from Candcnc with its fast THC could be a good choice. It also doesn't require a PC with a parallel port. But if you want to have ultimate control on hole quality with not so thin metals then you may want to consider one of the Candcnc parallel port versions. You can switch off the torch during motion without a motion pause in those systems. The THC on those just uses Mach3s basic built in THC which is slower but generally satisfactory when cut speeds aren't high. I'm periodically monitoring the Candcnc thread I started to see if they come up with a solution for the torch off motion pause problem. If they get that sorted then I'd recommend the ESS system.

Mach is not cam by the way, it's the machine controller software. The Candcnc system IS Mach3 with their own custom screen sets and software plugin, plus electronics of course.

If you use Candcnc, your cam software will be Sheetcam. It has post processors custom written for the Candcnc system. For the price, Sheetcam is quite functional and very good value.

Glad to see you're taking plasma dust into consideration (water table). I lost three computers and I blame that damn plasma dust. I finally spat the dummy and made a water tray, plus I converted a filing cabinet to a computer enclosure with a clear door, and filtered positive air pressure. No PC problems since. Protect your valuable PC and electronics from plasma dust. The Candcnc system is compact surface mount electronics so if plasma dust gets to it, all you can do is get a replacement board from the US. Better to never let that happen.

Tom from Candcnc recommended a stepper motor does 1 revolution for 1 linear inch of axis travel. I built my sychronous belt gear reductions using that "ratio", and seems to work fine, giving a good speed torque mix for my needs. My motors are 600 oz/in Nema 34.

The table is nothing to look at but is very functional. Here's a Youtube link:

https://www.youtube.com/watch?v=K3frWcfdv3E
https://www.youtube.com/watch?v=yg-s-EAqMfI

Keith

[Greolt]

Looks like you are off and running.

The only thing I would say at this point is that there is good reason why almost every serious plasma table I have ever seen, both commercial offerings and good home built, is rack and pinion on the X and Y.

Plasma requires fast speeds of travel in what is a very hostile dirty environment. R & P with correct gearing handles these requirements.

[WillyInBris]

Hi Keith and thank you for your long reply the HPM45 has a machine speed of 4.5mters on about 3mm at 45a I have actually sent out a few msgs to those with it on CNC zone and have got varied responses back for thickness so its going to be a bit of trial and error.
R&P will go on the Y as soon as I work out the right set up for it and I cut it out, the ball screw is temporary measure as not having a mill any more sucks, could use that right now but I needed the money at the time to sort some stuff out.

Thanks for the heads up on the Candcnc system will see if I can find that thread your talking about.

Already watched some of the videos on sheetcam seems the right way to go as you say.

Another question the Powermax45 uses 6 cfm @ 90 psi so I am thinking a 10cfm unit with a WP of 8bar 116psi should handle it if anything I am more worried about the moisture from the humidity we get up this way.

[WillyInBris]

Looks like you are off and running.

Yes being a home dad has it advantages lol lots of time to work on this and get it up and running.

The only thing I would say at this point is that there is good reason why almost every serious plasma table I have ever seen, both commercial offerings and good home built, is rack and pinion on the X and Y.

Plasma requires fast speeds of travel in what is a very hostile dirty environment. R & P with correct gearing handles these requirements.

Yep duly noted.

[beefy]

Hi Willy,

here's the Candcnc thread I started about that issue:
http://www.candcnc.net/supportforum/....php?f=2&t=746

Yeah, moisture is definitely a plasma torch enemy. I've got a little excavator with auger so I took a 4 foot tall LPG tank and put it in a deep hole I drilled in the ground. The in & out lines go in the top and I have a bleed line coming from the bottom and into the workshop. The ground keeps the tank cool and it acts as a great initial condenser tank. Guess I don't need to tell anyone to give an LPG tank a serious purging with air before doing any welding on one

After that the airline goes to a homemade refrigerated air dryer. I took a window aircon unit, plus a cars AC condenser and evaporator. I plumbed the compressed air to go through the condenser and evaporator. I made a chipboard closed loop air circulating box that moves the cool air from the aircon unit through the fins of the condenser and evaporator and back to the aircon unit. Got a water trap after that. Then the air goes into a dessicant air dryer (a big tube stuffed full of diaper dessicant). Finally I pass it through a small commercial filter to stop any particulate matter from the dessicant tube. People without plenty cash do desperate things LOL. You might want to look at getting rid of oil vapor too. I suppose there's a lot of commercial filters will take care of that.

My compressor is a Pilot K25 two stage running at 1100 rpm. It's motor requirement is 5hp but I had a 7.5hp motor I picked up somewhere so that's running it. With the 80 amp nozzle the compressor can more than keep up with needed airflow. I dumped the oil it came with and got some Amsoil fully synthetic compressor oil. I think the less hard a compressor has to work the less moisture they produce. Those cheapy compressors that run at light speed produce heaps of moisture I've heard.

You've definitely got the right idea about putting a cutout switch on your breakaway. I don't have a breakaway yet but I put an extra switch on my Z floating head. It operates about 1mm past the Z touch off switch. If the torch ever dives into the material the system gets an immediate E-stop signal. Prior to that I had one or two situations where the torch was being pushed into the metal and being dragged around (forgot what caused that). Some lads with breakaways have talked about how there table continues running after the breakaway / torch is lying on the table. And the torch is still running. Definitely put the switch on there.

Keith.

[WillyInBris]

IMG_44.jpg

So was thinking how are Precision Plasma getting away with this, its Direct drive they refer to Candcnc plasma kits, I am thinking say a Mod 1 with largish pinion and some Gecko 201's it obviously can be done as they are doing it thoughts everyone.

An interesting read
http://www.chacich.com/info/hdd/prin...ct%20drive.pdf
http://www.chacich.com/info/hdd/prin...ough%20HDD.pdf

It would so mush easier if they sold to Australia lol.

[WillyInBris]

lol keith sounds more like a meth lab but yep something I will have to tackle was looking at oil-less compressors that should solve one issue at least these are the units.
https://sydneytools.com.au/chicago-h...air-compressor

Guy two doors up has one in his work shop.

[WillyInBris]

Ok this so i just dropped into cncrouterparts.com seems as though some have been using their rack and pinion system for a plasma table actually
they are coming out with a build precisely for this, so have sent them a msg and see what they get back to me with.

[beefy]

Hi Willy,

I haven't delved deeply into it but these are my calculated thoughts on gearing. Now bare in mind I tend to always overengineer and am generally not happy with "it tends to work".

The PDFs mention both accuracy and torque. For plasma cutting the accuracy is good enough for me whether the Gecko has morphed to full step or is in 1/10th microstepping mode, especially when geared down. The floppy plasma flame doesn't have the accuracy to take advantage of such precision in any case. My el-cheapo X-axis rails are nothing more that 20mm angle iron with the 90 degree edge facing upwards and home made V-rollers running on top. Stretched fishing line was used to line them up accurately and my cut edges are straight and smooth (when all plasma parameters are optimal). The last two cuts I did on 3mm steel and the first thing the customers did it give nice comments on the edge quality. Not bad for angle iron rails LOL.

With everything I've read over the years, torque is your main concern with plasma. When you need a fast direction change you need both braking "torque" and acceleration torque. Steppers lose torque as their speed increases (servos don't) so you'd think that by gearing down, a stepper rotates faster (for the same linear speed), loses torque, and thereby the whole idea of gearing down is nullified. Now I haven't done any laboratory experiments on this matter but Tom from Candcnc claimed he came to the optimal 1 stepper revolution = 1 linear inch of travel ratio from experience. I guess you'd really need accurate speed/torque graph curves of the stepper motor you intend to use (these are normally in the motor data sheet), and work out from them what gearing would give you the best torque/resolution ratio. For ultimate accuracy (more applicable to milling and perhaps routing) I've heard splitting a full step up into 10 microsteps would not yield the same accuracy as full steps in a 10:1 reduction gearing. Reason is each of the ten microsteps are not exactly equal as far as how much they rotate the spindle. Microsteps are just electrically generated whereas full steps are part of the physical design of the stepper motor. Like I say though I'm only relaying what I've heard over the years. At the end of the day see who's using direct drive and how is their table doing. Maybe more powerful steppers make it possible. I've heard 900 oz/in are about the biggest you should go for plasma. Bigger than that and they can't run fast enough.

Keith.

[WillyInBris]

Ok so whilst I have plenty of time whilst I wait for the parts for my gantry to turn up I am trying to work out the following and really need some help (this is why I like ballscrews so easy to work out).

If I have a 20 tooth pinion and Mod1 rack what is the math that gives the distance travelled for one full rotation, I really should have paid more attention at school many years ago.

Just trying to get my head around this rack and pin stuff

[Ch4iS]

Linear pitch looks to be 3.14159 (pi), double checked with some M1 rack in the garage and it looks spot on. Seems the metric rack is in multiples of Pi.

So a 2.5x reduction looks to put you at ~25mm per revolution.


BTW looks like a good starting point.

[beefy]

OK if I've got this correct, a module 1 pinion with 20 teeth has a pitch circle diameter of 20mm

Pitch circle circumference (= distance travelled with one revolution) = Pi x Diameter = Pi x 20mm = 62.8mm

62.8 divided by 25.4 = 2.47 gear ratio to get one linear inch of travel for one stepper revolution.

So my calculations match Ch4iS.

Keith.

[WillyInBris]

Ok thanks guys so who here has successfully built reductions?

I am thinking it may be easier just to order the R&P kits from the states from CNC router parts may as well qet a quote will have to work out a way to mount it but shouldn't be toooo much of a hassle they are 3:1.