I am on the way of my 1rst real build. The idea is to be fast and sturdy. I am building it for a friend by order. Its a smaller variant of a sturdy machine i designed not long ago for my self which i would build after that build, which will serve also to figure some points of the process.
I have a lot of questions and will appreciate any help with this one, as i would like to make it really right.
From my part, once finished, i will provide the Sketchup plans for free here, so anybody can use them if wished so.
Some of the design points:
1. All steel soldered construction. Overbuild for stiffness, avoiding weak elements in the design.
2. Money is not an issue for the frame or the length of the supported rails, so i am not going to go cheap to save 10cm of rail or steel profile.
3.This machine is for continuous 3d jobs routing plastic or composite epoxy based materials so a desired stable routing speed of 300ipm is the goal
4. Usual job size would be 800x400x160, hence that commands the gantry clearance and z design
5. 3d continuous very detailed jobs, > 24h sometimes, made me choose a low power spindle like the 0.8kw water cooled one. Electricity cost is an issue here
So i am at the point of finishing the design, still waiting for some details on the profiles. Already started purchasing the components. I hope in 3 months it should be making dust
Here are some pictures of the design, still not finished:
Profiles used 180x80x4mm
Nema 23 low inductance motors , CNC routerparts Pro Rack and pinion drives 1:3, mod 1 rails, Hiwin HG 20 rails with long wide bearing blocks on all axis,
All plates are 10mm thick mild steel,
1. Now before i finish the design and order the profiles and plates , i wonder about the gantry weight. My design is 1000mm 180x80x4 profile with soldered over it at all sides 10mm steel plates. Basically the profile weight will be around 15kg ,reinforcing it will be another 50 kg+ the side plates another 10 + 2 motors and drives + Z another 12-15kg=more or less 100 kg. I am waiting for an advice from CNC routerparts about the gantry weight. If it heavy i can hollow some holes in the reinforcing plates to bring it to the desired weight , however i would like to reinforce it more than leave bare profile.
2.Another point here is that i am not sure if i can bolt to 4mm profile directly threading it or should be 10mm plates below RP rails, Hiwin rails, as i designed it
- the table would sit on 2 more rails same of similar size soldered. This should raise it for the cable guides and stiffen it, to avoid twisting . I am contemplating making a case there and all electronics mounted inside the base of the machine
What do you think about the design?
Last edited by Boyan Silyavski; 22-08-2013 at 10:53 AM.
Well the design looks strong but you have some quite major conflicts going on here.?
First the weight of the gantry really means your on the very edge of needing Servo's not steppers.!
Next because your using R&P then efficiency is very low so larger Nema34 motors will usually be needed to deal with this and the weight of the gantry.
The G540 is limited to 3.5A and 50V which isn't nearly enough for Nema 34 motors which you'll need at least 75Vdc or higher and more amps.
Nema 23 motors (even low inductance) just will just not be able to handle the Inertia this gantry's weight will put on them. Plus combined with R&P they will also struggle to give fast acceleration with that heavy gantry which you'll need for 3D work.
Even with larger Drives and higher 70Vdc voltage nema23 motors won't work with this weight gantry and high speed requirements.
Really I wouldn't recommend R&P for a machine this size and with your requirements for accuracy and 3D work, Ball-screws are the only way to go.!!. . . .You'll seriously regret using R&P.
Regards Steppers or servo's then I'd say lighten the gantry 20-30Kg and use Nema 23 motors running on 80Vdc drives with 70Vdc power supply and connected to 10mm pitch ballscrews.
If you don't lighten gantry then I'd say choose Servos has they will handle the inertia and speed much better.
Larger Nema 34 motors will handle the inertia at slower speeds but not do so well with the combination of speed and inertia has because they spin considerably slower than Nema23 motors they will struggle with reaching higher speeds anyway and the torque will be low at high speeds so with heavy gantry will struggle to handle inertia when changing direction.
So Above 80KG gantry at speed will need servos.
With Electrics in base of machine then I wouldn't recommend it has vibrations tend to loosen wires and don't mix well with sensitive electronics.!!
The 4mm steel will be fine for threads with rails.!
Last edited by JAZZCNC; 22-08-2013 at 06:54 PM.
Thanks. I am starting to realize this.
Let me explain something.
This build is a rip-off of my design for my next build which is meant to ride on 34 name Rack and Pinion.
I spend quite some time asking people around when i designed it, so it seems the overall opinion was that a 150kg gantry is quite well driven by 34 nema rack and pinion, i mean the acceleration and deceleration. There were even people on CNCzone that insisted that the gantry behaves well if is loaded more than 130kg so they made it heavy on purpose.
So everything started as i would like to test the RP on a real build, The nema 34 would raise the price, so i went for Nema 23 and started slowly underrating my original idea.
Now Just received an answer from CNCrouterparts and Ahron tells me what i suspected , that the g540 is not up to the task and i should couple their low inductance motors with better drivers, due to possible g540 overheating during long runs.
So it seems i am insisting on the RP drives . I like to test them and could not help . I am at the point of ordering the RP drives , racks and the low inductance Nema 23 . That means i must lower the gantry weight.
Ok, gantry weight.
Lets see. I can go back to 50 kg, complete gantry with Z, still keeping some reinforcement. Only eventual twisting should be considered, z is 300mm wide so no worry there in the other directions. Another point is that normally the jobs will be done with rough cuts by 1/4 router bits and fine passes 1/8 router bits, with passes equal to the width, routing wood as i said and that easy going special cast materials which route like MDB , so this is not a factor that would play in the final design.
Jazzcnc, i highly value your opinion. Why you believe the Chinese ball screws 2010 for example could give me better finish than the RP drives? All seem to be happy running RP drives, did i miss something? I know the table is not so big to really need RP.
A: They are large long machines so ball-screws are harder to use.
B: They can't afford Us made Ball-screws and the Yanks are very reluctant to Import Chinese parts.
C: The CNCRP R&P is US made so fits with above.
D: They have never used ball-screws so they don't know the difference in performance and smoothness ball-screws make.
My advise is just that "Advise" take it or leave it but I know thru experience with R&P that Nema23 motors with heavy Gantry is recipe for trouble if your wanting high feed rates.
Yes I know Nema 34 will handle heavy gantry ok and that heavy helps with cutting but Heavy and high feed rates is a completely different issue.!
Combine this with 3D work which requires Small fast direction changes and this issue becomes worse.
Now the difference between R&P and ball-screws is several.
Ball-screws are much more efficient so don't require has much power to turn, typical efficiency for Balscrew is 90%+ were R&P is 60% or lower.
Ball-screws have much lower backlash so are far more accurate and repeatable. R&P is quite high backlash and not easy or cheap to make real antiback lash using correct pinions.
R&P resolution is much lower and to get higher resolution requires at least 3:1 gearing this in turn means steppers need to spin fast and has you may or may not know torque drops off fast at higher speeds on steppers and even more so on Nema 34 motors unless running really high voltage.
R&P is noisy and needs much more maintenance and protection from chips etc.
Ball-screws give a silky smooth action where has R&P is in comparison a clunky action so this does translate back into the cutter at some level.!!. . . For general wood working then this is fine and doesn't get noticed but if your wanting high quality finish or high detail then you will see a enough difference compared to ball-screw machine.
Again just My opinion and you choose to take my advise or not.? . . . . But let me just say go try to find some one who has experience of Both and ask which they would choose for a machine this size, weight and with your requirements.?? . . . . I know the answer because I've give it you.!!
Last edited by JAZZCNC; 22-08-2013 at 08:10 PM.
I get it. I appreciate the help.
Then what would be your exact recommendation?
Drivers, breakout board, nema 23 motors , PSU combination in combination with which ball screws?
80V Digital drives and recommend Leadshine AM882
3Nm nema 23 motors like these SY60STH88-3008BF
70Vdc PSU preferably unregulated toroidal with correct amperage for all motors.
BOB recommendation is difficult because the one I'd recommend and think is by far the best is PMDX 126 but it's expensive. PMDX.COM - Products for CNC and motion control applications
The next I'd recommend and use often so know are OK boards is this one. OPPB but it's not in the same league has PMDX.
Ball-screws would be 10mm pitch and 16mm Dia for X & Y axis with 5m pitch 16mm dia. Motors would connect to them using 15mm HTD timing belts and pulleys with 1:1 ratio. This lessens resonance which helps on a steel framed machine and it's much more flexible if you ever want better resolution or higher speed has it's easy to apply a ratio with simple pulley change.
20mm Dia 10mm pitch could be used but really not required at this length and would be over kill, would also restrict performance not increase due to extra inertia of larger screw and ballnut.
All the above I've used on several steel framed machines with gantry's approaching 80Kg so know they work.!!
On my actual small machine i have trip signals from the VFD and power detection which are very handy in real life. And a second BOb af course. So that BOB wins. Actually its not so expensive for what it offers. I spend more money on my 2 boards.
1. 1610 for long axis that moves gantry and 1605 for the gantry / that moves Z/ and 1605 for the z, correct?
2. I buy belts from Belting online. Should i use the minimum sized belt pulley or go a bit bigger for the size of belt i mean? What is the typical size? I cant seem to find the sheet i used for calculating belts. I remember the AT was stronger but a kind of harder. Obviously either way it should be pitch 5, otherwise i have to go with bigger pulleys
Then 1605 for Z axis (UP/DOWN)
Don't use less than 15mm wide belts and I find 15mm is perfect size. 20 or 25mm is overkill and wasted money.
HTD is the preferred belt type for it's profile but don't think it would make a big difference here anyway has the belts are only short and not under much stress or heavy loads.
Yes 5mm pitch.
The Following User Says Thank You to JAZZCNC For This Useful Post:
After some fast brainstorm i decided on the CNC4PC breakoutboard 23 coupled with solid state relay board. This board seems more perfect for me:
1.I have my spindle control board from him and the support was great
2. All IC are mounted for easy removal without soldering / this is a big deal, i have repaired breakout boards in the past /
3. All necessary functions and really no extra ones that i dont need and would pay for choosing other boards
4. Better priced than the PMDX 126. When i started adding the spindle control/ not integrated/ ,relays, inputs, outputs, cables and so, the price jumped a lot
5. Some extra jumpers and functions i like
6. Straight out from the box 2 ports BOB with most possible inputs and outputs
Soon will start showing you packets
Thanks again to Dean/ JazzCNC/, now i know quite more than yesterday
Tomorrow will redraw the design for the ball screws.
Erm.!!! . . . Arturo's stuff is ok but it's no where near has good or well built has PMDX boards.!
BUILD LOG: Steel frame cnc router design/buildBy CraftyGeek in forum DIY Router Build LogsReplies: 110Last Post: 06-05-2015, 10:00 PM
BUILD LOG: First steel diy CNC router buildBy ivars211 in forum DIY Router Build LogsReplies: 59Last Post: 28-07-2014, 08:29 PM
By fatguyslim in forum Machine DiscussionReplies: 41Last Post: 08-02-2014, 11:37 PM
By Boyan Silyavski in forum Machine Frames & BedsReplies: 0Last Post: 06-11-2013, 05:58 PM
BUILD LOG: 7' X 4' Steel frame buildBy Ricardoco in forum DIY Router Build LogsReplies: 6Last Post: 28-10-2012, 06:02 PM