[mactec54]

Hi,

Native or retail Windows cannot do realtime, but there are ways around that, like dedicating one core as realtime scheduler, for instance Interval Zero sells such a solution.....but its far away from a hobby solution. The license cost of the realtime core costs more than Windows does.

Yes, that part you have correct, Window's OS is not a Realtime operating system, that being said it is quite easy to program so that it can operate a in Realtime though.

No, it is not something you have to buy, there are many ways to use any Windows PCs and have it perform for your application in Realtime serous gamers use their PCs in Realtime with just some simple configuration changes, this you have to be careful with as to how you set the priorities, or you can mess up your computer

All Windows PC based CNC solutions that do not use the a realtime core HAVE to be buffered.https

Any external controller FPGA Etc. requires a Buffer to be able to function, using Realtime or not.

SoftServo uses such a realtime core and therefore achieves genuine realtime performance....but at eyewatering cost.

The SoftServo control is not expensive for what you get, but is more focused on OEM manufactures

Galil DMC-40x3, the current model three axis controller is $2200USD (from Galil website) is not a hobby solution.

Galil has never been a hobby solution so why talk about it. Camsoft was the only CNC software that was designed to run with Galil, Mach3 plug-in was a failure, I don't know how Mach4 is with Galil but hopefully it works better as the software engineer that worked with Mach4 design was a software engineer from Galil

The only requirement for Realtime on Windows PCs is a processor with at least two cores. The basic operating principle is to instruct Windows to only use a limited number of CPU cores of a multicore processor.

The Realtime system is implemented as a device driver and provides its own API, which allows the user to export the time-critical code parts of a Windows application into a DLL. all of this is regular programing and easy to implement for a CNC application or any other application that needs to use Realtime processing

Note that most Windows PCs have up to 12 core or more depending on what generation CPU the computer is using, even the low end i3 has 4 cores and 8 Threads.

[JayneV]

Hi Jayne - When your home publish the new lengths and I can check for whirling.... Ignore cnczone while your on hols... Peter

Hi Peter,

We got back from NZ and now I'm nursing a nasty cold so unfortunately still no work on Carlos. I'm starting to think this machine will never become a reality........just kidding, it will happen eventually.

Here are the new lengths of the ballscrews. The X and Z axis are the same length as before, the Y is roughly half the original length at 660mm. The lengths below are the distance between the support bearings for each axis. If you get a chance, would you mind checking for whirling if I use servos?

Y - 1610, 660mm
X - 1610, 1020mm
Z - 1605, 280mm

[peteeng]

Hi Jayne - Yes Carlos is over 1 year old... time flies etc...

https://www.hiwin.us/wp-content/uplo...ew-Catalog.pdf link to Hiwin technical catalogue with all the info needed to design ballscrew drives. Happy reading.

The attached uses their formula. L is centre to centre of support bearing to centre of nut so you can reduce the length and recalc if needed. The crux is servos can go to 3000rpm (plus) a 16mm screw 1020mm long is predicted to whirl at 2333rpm and Hiwin use a 0.8x service factor so 1866rpm ~1900rpm is max by them at 1020mm. But your L will be shorter. For 3000rpm the critical length is 901mm.... But I doubt your going to max out to that speed. 10mm pitch means 3000x10=30m/min which is a scary speed in my experience (Carlos will be doing a fast tango at that rate across the garage floor) so 20m/min is very very fast and comfy... 20000/10=2000rpm near the allowable of 1900rpm (but not a fixed in concrete sort of number) so hope these numbers help. Regards Peter

by the way I have confirmed Hiwins numbers via FEA and they are as accurate as can be expected... and you can see that steppers will work fine as they top out at say 1200rpm

[JayneV]

Thank you for running the calcs Peter. There is lots of info in that HWIN document.

[JayneV]

I had 10 additional days off work after returning from NZ and was really looking forward to make some good progress on Carlos but unfortunately the project has to temporarily be put on hold yet again. The nasty cold I got turns out to be covid and after managing to avoid catching it all this time it finally got me and it’s got me pretty good. It could be much worse so I can’t complain, it’s just annoying that I am still struggling to make any headway with this machine. Hope I can get a few days work on the machine next week, in the mean time I’m keeping busy watching tv and having lots of long naps lol. I promise to finish building Carlos and make some sawdust so whoever is still following along, thank you for your patience.

[peteeng]

Hi Jayne - Don't fret over the timeline. Most first builds run well over a year if you look at the threads. There is a lot of legwork, research and doing plus life to manage. So just be persistent and a little crazy and it will happen. Peter

By the way this is the laser welder I want for Christmas if you have a hot line to Santa?

3 in 1 laser welding machine, do you like this machine?#laserwelding #... | TikTok

[JayneV]

Ooh, that is a nice welder. I want one too.

[mactec54]

Hi Jayne - Don't fret over the timeline. Most first builds run well over a year if you look at the threads. There is a lot of legwork, research and doing plus life to manage. So just be persistent and a little crazy and it will happen. Peter

By the way this is the laser welder I want for Christmas if you have a hot line to Santa?

3 in 1 laser welding machine, do you like this machine?#laserwelding #... | TikTok

Yes, I have been looking at them also they are getting better and better, still a little spendy from $10K to $40K for one like in the video

[peteeng]

Hi - some 30 years ago I bought a Lincoln welder for a titanium project and it was $26k aud so no different I think, it was the size of a medium fridge . Now same thing you can carry and plug into house power. Plus I had to upgrade the workshop power. After upgrade didn't mind as the welder made the power meter run backwards... That Lincoln was never as good as my Miller though. Lasers will become normal in commercial use I suppose and cheaper. Question is will we still have to stress relieve? One video shows how they weld under water, that would stabilise the heat affects quite a bit ?? Peter

[joeavaerage]

Hi,
yes that is a fancy welder but when it comes to welding heavy section steel you can't go past either a Lincoln DC600 or a Miller Deltaweld 600.

With 2.4mm gas sheilded flux cored wire through a good MIG torch like a Tregasskis, fully set up for high current welding with leather gloves, leather
bibbs, you name it. You can lay down a weld in one pass (350-400A) with a depth of penetration of 12-14mm....just magic. The next best thing to sub-arc.

Craig

[peteeng]

Yes Craig that maybe today but soon laser will overtake. A 15kw laser can weld 16mm thick steel no prep. A 30kW laser can cut 130mm steel so I'm sure it can weld it as well. Cheers Peter

[joeavaerage]

Hi,
do you fancy wielding a 15kW laser torch? The smallest of slips which cause you to lose control of the torch and you could cut your other hand off like a guillotine.
I can't imagine it will ever be safe to put such powerful laser beams into handheld equipment.

Craig

[peteeng]

I suppose light sabers have the same problem... we do lots of things we shouldn't. But safety is a biggy so I suppose we would use a cobot. The video is a 6kW its safety feature is via contact thru the copper tip. If it's not contacting the job its off. 6kW will take off your arm or go thru your mates abdomen I suppose. The force is strong I sense in you Craig. Peter

[JayneV]

Today I disassembled the big wooden bench I had made for Carlos to reclaim some garage space and cleared the existing steel bench where Carlos will live. The available space is now much more useable. The clear benchtop was screaming for something to be placed upon it so I cut the top and bottom plywood skins for the machine base and roughly stacked some parts together to get a feel for the new size of Carlos and it looks like a nice size for my needs. The wall bottom plates still need to be cut down from the old size so ignore the massive overhang in the attached photo.

[peteeng]

Lookin good Jayne - Peter

[Mr-MaW]

Thank you. The link for the book was just the first result I found from a local book seller. I've already found cheaper links since that post. I also asked my brother and he has the book which I can borrow in the meantime. I also found and watched that video about tabs. It's good information and it highlighted a few things I had not considered.

Jayne

Hi everyone,

My brain has gone a little mushy thinking about the CNC pretty much 24/7. I was heading towards burnout. Spending so much time in the design phase was never my intention, this project started out with a purchased, proven fit for purpose design, which I wanted to make a few minor adjustments to suit the materials I already had available. It’s so easy to get carried away with tweaking this and adjusting that, before you know it you are designing a whole new machine practically from the ground up. I would really like to place an order for the parts coming from China with Christmas just around the corner but I t’s probably already too late to expect any parts to be delivered before the new year. However, the design is not quite ready to order any parts yet until I sort out how to drive the screws (directly or with belts), so it was time for a little break to give my brain cells a rest and then return to the CNC with a fresh mind.

To help keep me motivated, I have started modelling the first guitar project for Carlos once it is built. Still needs a lot of work like figuring out how to create complex curves where the neck meets the body. I’m still relatively new to Fusion and I don’t yet know how to do that. It might also be easier to rough out the shape with simpler geometry and then carve and sand the final shape by hand, but addressing those details is still a long way away.

I’ve attached some rendered images of the guitar as it is so far if anyone is interested. I got distracted by the shiny materials in the fusion library and made the last two models from shiny metal. One is bronze with a maple fretboard and the other is platinum with carbon fibre fretboard, just because they were shiny.

I’ll get back to Carlos tomorrow.

Have a great day/ night everyone.
Jayne

Leadshine are good enough that other Chinese companies counterfeit them.

Hi Jayne - Yes I thought you wanted to build one. If you use the printed parts for structural parts you will be disappointed. Most people with printed parts change them out pretty quick. MDF or plywood is much stiffer.... Always up for doing something because we can, as long as its fun. Maybe a better plan is to use the 3D printing to make moulds for cast UHPC (grout) machine parts. Peter plastic E=3GPa grout E=40GPa al E=70GPa I like machines you can walk on...

https://youtu.be/ugwdLwid6NM

Nice to meet everyone.
If you need mineral castings.You can contact me, my email:[email protected]
I'm from China,My company has produced a lot of products, small can reach 100KG, large can do 46 tons.
My English is very average and I don't know how to upload photos. Very reason to share my product with you.

https://grabcad.com/library/cnc-lath...ral-castings-1