I've been asking questions & questions here in CNCZone, and
I think I can now show you most of my progress at this moment.
This project started quite long before ever since I have purchased a greatly WRONG machine from China.
This photo was taken at the very date it arrived.
It has all good things in it like Leadshine closed-loop steppers, PMI guide rails, 2.2kW Changsheng spindles, etc except they were built in a wrong way.
* I warn you to stay away from these guys or similar looking machines by all means.
They are selling variants of this design under many brands, not only from this company.
* One very important point : if you look at the pictures closely, you will be able to see that there is only ONE bearing support for the ball screw.
There's no support at the motor side. All variants are same, this is ridiculous and you will see indicator dancing when the ball screw is rotating.
Not to mention parallelism/slignment/bed flatness are all wrong. They are totally wrong beyond one can manage at his shop.
I tried various things like new bed and re-machining LM holes;
Far much greater than before but still not up to my standard.
OK, enough said.
Undergoing painful period with the machine, I desired to make a complete new one.
I made a wish list first;
- all servo system with encoder feedback, regardless of open or closed-loop
- gantry moving structure
- re-using already made $1,000 valued one-piece aluminium 6061 T6 30mm thick bed, machined by ultra-quality MCT
- swappable spindle between 2.2kW VFD driven high-rpm and a servo-belt-driven milling spindle
- 4th axis add-on
- 3D digitizer/Touch plate active without changing settings
- a little bit more...
And I started my drawings...
It is not shown in the picture but I put actual weight or simulated weight of every possible components into solidworks so that I can locate center-of-gravity in the middle of gantry bearing blocks as long as I can. Some minor components are missing like cables or screws, I ignored them.
Drawing the whole machine took really a lot of time. Searching for useful components, re-designing, calculating positions, hole locations, lead of ball screws, possible max rpm, etc...
And I started fabricating components.
Gantry was made out of cast aluminium, costed me total under $200 because I made wooden form myself.
Machining was outsourced, by same company who did the beautiful bed.
Y-axis dual-shaft mount was outsourced too - I couldn't make it myself.
* big one at the bottom, made out of one piece 7075 T6 aluminium
and the rest were done by me.
* Y axis motor plate
* Y axis rear-mount for ball screw
Back panel was machined using the poor CNC before I dissembled it.
Well.. originally, I dreamed of KFLOP controlling everything.
I bought one piece of KFLOP, playing it quite a long time, bothering Mr. Kerkes countless times by postings and mails and tried to make my own breakoutboard like this;
Fried one..... uh... okay.. it might happen.. I am no electronics expert....
I ordered another... fried second one.. uh...oh... my god...
I still don't know exact cause of this failure, however, no time to waste like this...
Immediate turned to possible alternatives, and decided to try ethernet smooth steppers from Warp9.
To get encoder feedback from servo drives, I still needed another board to do that and this time it was successful.
It is very boring without a video.. so I prepared many...
Testing control box/servo connection/encoder readout
Testing homing WITH index pulse
Mach3 screen mod, and made covers for Y-ball screws
XY axis test
I am putting some accessories at the moment - oil distribution, limit switches, cable managements..
More will be followed soon...
I finished all 3 axis setup and added some accessories.
I still have to go a little further but it might be good to show for now.
Max speed is reduced just a little bit, from 24,000mm/min to 18,000mm/min.. lol
You've done a very nice job on your machine, especially the aluminum table top and the castings for gantry. Very very nice. The machine seems very smooth and think you will be happy with the cuts. The ballscrew covers are the only thing that does not look professional. I wish I could find some myself, I have seen some of the metal spiral kind on some machines that are so smooth you can't even tell they are there. At least you have something to keep dirt out of the long screws. I also get stuff in my Y axis across the top and have to clean that screw repeatedly and that is a pain when cutting aluminum.
I wish my dark and painful period can be compensated with this new machine.
I know what you mean by metal spiral thing. Probably conical spring covers Conical Spring Cover manufacturer in Pune India-Bellows Manufacturer in Pune-India
In my case, compression was very important not to loose any possible travel on Y axis and I could not find any to satisfy my needs.
It sure looks non-professional.. I know... hehe.. just a desperate gesture.
I still need to make one bracket for the upper cable channel and holes on spindle bracket but that would not take long.
Most important parts were done already.
Kevin,
Honestly if the thing hangs on and does not rip as the machine moves around you have accomplished the task. Yeah I agree those spring covers are very expensive and they do not fully compress, I have tried making cardboard bellow type covers and they work pretty well and do compress a good deal if you don't select paper that is too thick.
It is very difficult to lose some that close to you, and like you said the machine is just a distraction it will not take away the pain. Everyone heals in different ways, try to keep the wonderful moments with your brother alive in you head. He is smiling down on you now.
wait a moment...
I don't understand what made you to think I lost my... who... my bro???????
there seems to be a great misunderstanding..
but I appreciate you took care of my.. um.... well.. whatever . lol !!!
That's absolutely OK!
My assembly is 99% finished.
Fine-alignment is necessary for Z-axis only, all other axis are stay-tuned and square in very good condition and I think Yaskawa's auto-tuning function is so great!
Harmonic reducing headgear for Z-axis is generating some heat inside due to full-rpm range from servo motor so I decided to change it to 5:1. (It is 10:1 for now)
That will provide enough torque and speed and less heat than 10:1.
That is a very good looking build. I love the aluminum table top you will be very happy you did that in the future as it will stay true. You might want to do some test runs with the spindle operating with the water cooling activated to ensure you have no noise issues. Seems to be moving very smoothly. Nice Job.
Thanks for the advice.
All cables are shielded properly from the beginning
I have been busy designing a new dust hood for this baby.
I think performance test is finished. After putting new dust hood, it will work as the heart of my workshop!
Mounting for alternative servo-driven milling spindle is still way to go.
regards,
Kevin
ps. Lozzy was my WoW nickname, which I haven't played alreay for years. Kevin is fine heh
Nice, the dust shoe is really a must especially when you start cutting. I ended up trying several units I designed and finally order the unit that KentCNC makes and it works great with removable lower unit with magnetic. Looks like you machine is running very smooth. Congrats.
I remember what you mentioned.
I also once looked at that one, way too much expensive for me to order one. haha
It is done now, quite late due to national holidays and several other things.
Very nice work, you really went to a huge amount of work to build that dust shoe but it should last forever. You machine is cutting aluminum very nicely.
What did you use to generate the gcode, you used 3D adaptive clearing which seemed like a fast way to remove a good deal of material.
Your finish is also very nice, wow you are doing a fantastic job.
Close-up of cut-surface is not quite good especially the slot-surface.
Lots of chatters, because of 1 path/multi layer, but... who cares surface smoothness from a dust shoe?
I suggest you to look into the term iMachining in youtube.
That's the way to utilize full-length of endmills while preventing excessive wearing of tool-tips, and it provides much faster roughing than any other method.
This logic is implemented as 2D or 3D adaptive clearing in HSMXpress - and HSMXpress is free to use, very important.
I haven't looked for MasterCam but I guess it would be there too with other name,
The last thing I am working on is a vacuum table, cast-aluminium body with MDF-top to hold thin plastic materials when I cut pickguards. (I make custom guitars)
I will make a wood-form and after it is casted, I will machine it myself.
Double-sided taping is terrible ... I hate it.
Thanks Kevin,
I will do some searching on youtube. I normally cut aluminum dry and typically with smaller end mills using about 10,000 RPM and it does a pretty good job. I have surfaced some aluminum with a 1/2" solid carbide EM and run that at about 8000 RPM and the finish was very nice. Even nicer finish if I put on some WD40, but hate the smell and it gets all over my MDF table. I have a new aluminum table top that I need to drill and tap and then that would work. Working on Guitars you probably do not want any WD40 in your shop. LOL
I tried dry cutting some years ago but results were very bad, I never cut dry afterwards and that's why my CNC bed is made of aluminium to use cutting fluid - though minimal use.
However, no liquid while making guitars lol.
The Chinese VFD spindle does not guarantee torque under 7000rpm and I will be working on that issue by adapting a servo-driven milling spindle in place of VFD spindle.
Today I have finished setting up my digitizing probe.
Scanned size is smaller than real one about 0.38mm, I suspect it due to latency issue. Then I can compensate that amount in tool-tip compensation.
Interesting thread.
Would you mind sharing the SGDA CN01 and CN02 parameters??
Can you printscreen and show a pic of your parameters?
I'm having a particularly hard time figuring out how to get smooth rotation from the SGDA-02AP drives. They only turn 90 degrees then stop when trying to autotune them.
It's got me stumped and I've tried everything under the sun in the manual.
Thanks
Probing becomes fun now with new tools!! Music? I don't remember. Youtube suggests BGM and I just pick one from the list. haha. They are free to use ones ;p
/G59
I will post servo-parameter data as a file tomorrow. You will be able to simply import them in SigmaWin+ software. Version 5.xx is for SGDA. Version 7 is for SGDV and later products I believe. Mines are SGDA and SGDS so I use version5.xx only.
But my setup is based on quadrature 4x pulse train so you will be needed to do mimimum modification after you uploaded the data.
Reply with Quote
Mach3/ESS/Yaskawa servo - under construction
