Hood
I am obsessed.
- Messages
- 17,456
- Location
- Carnoustie, Scotland
Because it was done that way to lull people into a false sense of security and cause them maximum bewilderment and grief
- Thread starter ned
- Start date
ned
Member
- Messages
- 954
- Location
- Eire
On the road again. The first table I built, which was too small, is becoming the legs for the second table I built.
This is year 5.
I got evicted last year and its taken me this long to get it together. Feels good.
You can see in the pic , on the floor there is a 5x10 sheet of steel, 6mm. This will be my first bench. I'm gonna roll off a bench if i can get this thing to work. Nice to have the sheet cos I think they cost mad cash these days.
My camera is sheite , till I can afford to buy a new phone, so bear with me.
This is year 5.
I got evicted last year and its taken me this long to get it together. Feels good.
You can see in the pic , on the floor there is a 5x10 sheet of steel, 6mm. This will be my first bench. I'm gonna roll off a bench if i can get this thing to work. Nice to have the sheet cos I think they cost mad cash these days.
My camera is sheite , till I can afford to buy a new phone, so bear with me.
ned
Member
- Messages
- 954
- Location
- Eire
I welded the table to the sheet, and cut it away later. The whole thing still moved, so i placed a beam across the doorway.
You can see the planks i put up against the back wall, I was worried about taking out that wall. It was pretty scary for about 15 minutes. Pretty hairy. The bed was on wheels, and the back wheels started drifting as i winched it up. Anyways, I got away with it.
Last edited:
ned
Member
- Messages
- 954
- Location
- Eire
Got somewhere with this. Spent about a month getting the gantry and gantry plates square. More time than that probably in fact.
The rails are test fitted for now. Im going to fill the tank with water and then use the machine torch to scoot along the top of the water and tell me where to set the rails.
The rails are test fitted for now. Im going to fill the tank with water and then use the machine torch to scoot along the top of the water and tell me where to set the rails.
Barking Mat
Cuddler of hedgehogs.
- Messages
- 13,134
- Location
- France, Brittany.
Fair play, takes time, but getting there.
Faster than my non existent advances.
Faster than my non existent advances.
ned
Member
- Messages
- 954
- Location
- Eire
Ah cheers Mat. I nearly lost my job last september, mass redundancy, but I ended up keeping my job. Anyways, as soon as i got the news i started working on the machine again. Been at it since actually. Superb motivation.
Some more pics. I built a lot of adjustability into the rails, to cope with the warp lol. I reckon it'll work.
I used the Fireball square to get the ganty square, was very handy for that. Did a fair but of filing on the gantry beam with a rasp, to get it all to sit right.
Some more pics. I built a lot of adjustability into the rails, to cope with the warp lol. I reckon it'll work.
I used the Fireball square to get the ganty square, was very handy for that. Did a fair but of filing on the gantry beam with a rasp, to get it all to sit right.
Last edited:
sako243
Member
- Messages
- 3,641
- Location
- My mansion in Wales
Going back to the discussion around stepper motors etc., I bought a 750W AC servo via Aliexpress for the back gauge on a big cnc hydraulic shear for £250 including driver. It was pretty trivial to setup with my custom controller. That spun up to 4000rpm but you could just tune the acceleration / deceleration in the parameters on the driver and it's been working like a dream. It has a stepper motor like interface with a "done" signal at the end. The only issue I had was that I couldn't reliably drive it at the claimed pulse speed but you can adjust the "gearing" in the parameters so it can be driven by slower pulses and these are multiplied up internally by the driver. Even with that gearing set quite high with a 2mm pitch on the lead screw it was still able to achieve 0.01mm increments which was way in excess of what was needed.
ned
Member
- Messages
- 954
- Location
- Eire
you sound like a person with a clue!
I actually did manage to figure out the wiring on a set of servos that i had, but I sold them and went with steppers for this machine, in the hopes of just getting it done.
Steppers and the myplasm controller in the hopes that I can manage to figure out that that much.
The myplasm people have since posted up detailed "how to setup" guides, which is great.
I'm gonna have to figure out the steps / microsteps / whatever the hell those parameters are soon enough, but I've a few other tasks to get through first.
Level rails
Sort out the cable chain
Sort out the belt drive
Make motors move
I actually did manage to figure out the wiring on a set of servos that i had, but I sold them and went with steppers for this machine, in the hopes of just getting it done.
Steppers and the myplasm controller in the hopes that I can manage to figure out that that much.
The myplasm people have since posted up detailed "how to setup" guides, which is great.
I'm gonna have to figure out the steps / microsteps / whatever the hell those parameters are soon enough, but I've a few other tasks to get through first.
Level rails
Sort out the cable chain
Sort out the belt drive
Make motors move
Erie Fred
Member
- Messages
- 5,251
- Location
- Erie, Pa USofA
I agree with the switch from servo to stepper, the plasma process does not need the servo's accuracy.
As far a cable chain, I wanted to go that route to mimic the professionals.
As I bought my machine as a gantry and constructed the table, the gantry came with chain already installed,
I changed my design from chain to festooning when I built the table & rails.
A simple piece of unistrut and some trolleys, and it does not need alignment like a chain does.
Easier to add/remove cables & hoses as well.
The large commercial machines use festooning in some places on the machine.
I don't have a pix currently, but at each trolley (there are 2-3 for 10' of travel) you simply make a loop
thru the trolley hook, to hang a loop of the cables. In this case a heavy ty-wrap was used, others I have seen use a shower curtain hook.
sako243
Member
- Messages
- 3,641
- Location
- My mansion in Wales
The reason behind servo's is far more than accuracy. In this scenario we don't care so much about positional accuracy (both stepper and servo can exceed what is required) but about speed as @Hood says.
Steppers are, with a good driver, typically limited to a maximum of 3000RPM. Depending on your gearing setup you may need to up the gearing to achieve the necessary speed for the plasma to cut thinner material cleanly. However upping the gearing means increasing the torque of the stepper which will likely slow down the maximum speed achievable. Servo's on the other hand can easily run at 3000rpm and often far higher meaning you don't need to up the gearing and don't have the torque requirements.
They typically have well tuned PID control loops and will allow the user to tune the acceleration, deceleration etc., relatively trivially. This might be important given the potential weight of the gantry smooth acceleration would be highly desirable.
I wouldn't push a stepper anywhere near 3000rpm on a CNC machine. I'd say that if you're needing to spin a stepper faster than 1000rpm, then you should really be looking at servos.
The big disadvantage with steppers, is the torque drop off.
Here's the torque graph for a NEMA23 stepper-
As you can see, that's a 2.3Nm rated stepper, but even with the 48VDC supply, by the time it's gets to 1000rpm, the torque has halved.
And don't fall for the bigger = better for stepper motors.
You can often get far better performance from a suitably geared NEMA23/24 size motor than you will from a larger NEMA34 motor.
NEMA34 motors really need high voltage (aka 240VAC powered) drives to get optimal performance from them, but then you're spending more than what a servo would cost.
A servo on the other hand, will maintain it's rated torque all the way to it's rated speed (typically 2500-3000rpm for the kinds of sizes you'd use on a normal sized CNC machine), and will also be capable of exceeding it's rated torque intermittently, which means they can be pushed far harder, giving you far better acceleration.
The main benefit of a stepper is simplicity. Due to their torque curve, they're more suited to direct drive, they're generally easier to wire up (servos will typically need some form of D-sub connector soldered up for control wiring), and they require less configuration.
The big disadvantage with steppers, is the torque drop off.
Here's the torque graph for a NEMA23 stepper-
As you can see, that's a 2.3Nm rated stepper, but even with the 48VDC supply, by the time it's gets to 1000rpm, the torque has halved.
And don't fall for the bigger = better for stepper motors.
You can often get far better performance from a suitably geared NEMA23/24 size motor than you will from a larger NEMA34 motor.
NEMA34 motors really need high voltage (aka 240VAC powered) drives to get optimal performance from them, but then you're spending more than what a servo would cost.
A servo on the other hand, will maintain it's rated torque all the way to it's rated speed (typically 2500-3000rpm for the kinds of sizes you'd use on a normal sized CNC machine), and will also be capable of exceeding it's rated torque intermittently, which means they can be pushed far harder, giving you far better acceleration.
The main benefit of a stepper is simplicity. Due to their torque curve, they're more suited to direct drive, they're generally easier to wire up (servos will typically need some form of D-sub connector soldered up for control wiring), and they require less configuration.
1.2metres per minute is fine for cutting, but it soon gets old waiting a minute or two for the gantry to crawl back the length of the bed.
I get fed up waiting 20 seconds for my little mill to rapid back the length of the table.
It is however still using the original old style round steppers, as I couldn't be bothered upgrading them to modern steppers, which would have let me eek a bit more performance. Ideally I'd convert it to closed loop steppers, but that would involve lots of additional wiring, and I quite like that the control cabinet can be unplugged easily as it is.
The one benefit of plasma, is you don't have to allow for cutting forces or additional weight on the bed, so you can push any system harder.
However the weight of the gantry of this machine, is going to take a fair bit of power to get moving and to slow down.
To get the best performance from a stepper system, is going to cost more than a servo system that will give you far better performance.
The big gain in time, is usually from the increased acceleration.
All those second or so gains not having to wait for the gantry to accelerate adds up over a job.
I get fed up waiting 20 seconds for my little mill to rapid back the length of the table.
It is however still using the original old style round steppers, as I couldn't be bothered upgrading them to modern steppers, which would have let me eek a bit more performance. Ideally I'd convert it to closed loop steppers, but that would involve lots of additional wiring, and I quite like that the control cabinet can be unplugged easily as it is.
The one benefit of plasma, is you don't have to allow for cutting forces or additional weight on the bed, so you can push any system harder.
However the weight of the gantry of this machine, is going to take a fair bit of power to get moving and to slow down.
To get the best performance from a stepper system, is going to cost more than a servo system that will give you far better performance.
The big gain in time, is usually from the increased acceleration.
All those second or so gains not having to wait for the gantry to accelerate adds up over a job.
henry Kadzielski
Member
- Messages
- 1,215
- Location
- Australia Wollongong
A lot of people talk about upping the speeds, increasing the 'performance'. There are down sides, one in particular. Newton can and will be heavily involved here. To every action there is an equal and opposite reaction. I have servos , now my gantry is big and thus heavy, so on the quick direction changes the table wants to walk around the workshop floor. Years ago I couldn't understand why plasma machines were so heavy, the frames strong and bulky. That's to stop it from walking around the workshop floors. My rapids are 19m/m and don't give much trouble with the table walking about, however when cutting thin material, which I don't do much of is were the problem is. 3mm steel at 45amps the letter 'S' is a big problem. I can't remember the cutting speed but it has to be higher than 4.5m/m. Just saying 
Plasmas are the kind of machine that really benefit from S-curve acceleration, however it takes a lot of calculations and processing power.
Art's Tempest trajectory planner for Mach was a good attempt at it, but he stopped developing that a while ago (I want to say even before he sold Mach to Brian), however S-curve acceleration isn't something I've really heard about since.
I know various trajectory planners do implement a quasi S-curve algorithm, but I think most probably rely on post-processors now.
One example of quasi S-curve is Dynomotion's KFlop/Kogna use jerk (essentially how quickly the acceleration is applied, and is measured in units^3) as part of rapid moves, which smooths the acceleration, but it's on a per axis basis, with no interpolation between axes. During rapids interpolation isn't needed, so it doesn't matter if each axis reaches the destination at different times.
Art's Tempest trajectory planner for Mach was a good attempt at it, but he stopped developing that a while ago (I want to say even before he sold Mach to Brian), however S-curve acceleration isn't something I've really heard about since.
I know various trajectory planners do implement a quasi S-curve algorithm, but I think most probably rely on post-processors now.
One example of quasi S-curve is Dynomotion's KFlop/Kogna use jerk (essentially how quickly the acceleration is applied, and is measured in units^3) as part of rapid moves, which smooths the acceleration, but it's on a per axis basis, with no interpolation between axes. During rapids interpolation isn't needed, so it doesn't matter if each axis reaches the destination at different times.
ned
Member
- Messages
- 954
- Location
- Eire
Ah no I appreciate the banter!
I have cable chain, gonna stick with that, ill stick up details of the components I have now in a while. I have some nice steppers and I have a guy here locally that sells openbuilds CNC components, I sold him a Gerber CNC that I bought ages ago and now he is helping me out with things like gearing. Its great!
I have cable chain, gonna stick with that, ill stick up details of the components I have now in a while. I have some nice steppers and I have a guy here locally that sells openbuilds CNC components, I sold him a Gerber CNC that I bought ages ago and now he is helping me out with things like gearing. Its great!
ned
Member
- Messages
- 954
- Location
- Eire
@Hood can I pick yer brain for a mo please? I have four steppers in total, three of them are Leadshine 57CM23 and then I have one that I can't find the model number of, so I dont know anything about it, other than its also Nema23 and should do the job just fine, I was gonna use it on my z-axis
I reckon I can figure out the dipswitch / microstep settings for the three Leadshines, but what about the fourth unknown stepper?
I dont mind buying another.
The drivers im using are Leadshine M542-05 (link to PDF)
I have a 48V 21 amp power supply to push them along, its a Meanwell RSP-1000-48
I reckon I can figure out the dipswitch / microstep settings for the three Leadshines, but what about the fourth unknown stepper?
I dont mind buying another.
The drivers im using are Leadshine M542-05 (link to PDF)
I have a 48V 21 amp power supply to push them along, its a Meanwell RSP-1000-48
