I've built this machine 100 different times in my head!!!

[JoeHarris]

Since I've decided to go with dual x axis drive, my PC case control cabinet is looking too small to squeeze another am882 driver in.

May of missed this in your thread but why did you decided on 2 drives over the long belt option??

[Iwant1]

After spending so many hours deciding between the two options, and after some emails with Jazz, I just bit the bullet and decided with dual drive. The main points in my decision were:

with AM882 there wasn't any issue with the gantry racking from one motor stalling as they have stall detection alarms, and the machine will E-stop before any damage is done.

Next, as for sync issues creeping in, Jazz says with descent drives this is drastically reduced compared to cheap Chinese drives. Obviously with one motor and belt its completely eliminated and is only a problem on very long jobs, which I doubt I'll be doing. Though you never know.

Then with 2 motors tucked away in the corners, it keeps the far end of my router free from belts and motors and tensioning arrangements, allowing for sliding long materials through.

With the long belt and one motor, I reckon a larger motor like a nema 34 would have been needed, and since I prematurely bought a nema 23, thought I'd use it on one side and buy another 23.

Also I think the 23 motor can spin faster at the top end, which will be ideal for my incorrectly ordered 5mm pitch screws, when traversing.

This is how I came to my choice, but it no big deal if I want to change change later. And I'm sure someone here will snap up my spare driver and motor, lol

Adil

[Danielroyal25]

I have been pondering over the 2 motors v 1 motor decision too. I think I will be going with the 1 motor option. You know you mention there may be issues with long jobs, what is the reason behind this.
I wanted the ability to slide large boards through the machine but may settle for turning them through 180 degrees instead.

[Iwant1]

You know you mention there may be issues with long jobs, what is the reason behind this.
I wanted the ability to slide large boards through the machine but may settle for turning them through 180 degrees instead.

From what I understand, long jobs are when you machine for 10+ hours without homing in between. Imagine one motor is losing a step here and there, and over a all those hours the missed steps could add up to a difference of a few mm's of one side of the gantry, taking it off square.

Homing would normally square it up again, but I'm not sure why on long jobs, like 3d carving, why can't you program a homing function every hour, and then continue where it left off.

Adil

[Bart]

Hi,

I don't want to hijack this post but i have a similar design and I'd like to use your experience.

I've designed the frame etc. in inventor and from the simulation i get deflections smaller than 0,015 mm with the force of 500N excreted on a tip of an end mill. Thus, the design seems to be sturdy and good enough but I'm waiting for any suggestions.

My questions...

1. Chinese spindle - which one to chose
There is plenty of different looking spindles on ebay and even more sellers so I'm not sure where to buy and more importantly which model. Which one do you have or which one you know is safe to buy.

2. how to avoid two motors.
I tried couple designs of frames to hold one ballnut for one ballscrew in the center under the table but it's hard to get small deflection without making it very heavy. Do you have some ideas how to solve it without 2 ballscrews and motors on each side?
How is the solution, with line that winds up on a rotating drum, called?

3. what's the free play on a linear rail's open bearings?
Is there any free play on them, how much do they flex?
Also I saw couple times that max speed for linear rail bearings is 5 m/s. What happens above that speed and how does it affect service life.

4. whats the advantage of linear guidways over supported rails?
They would cost me at least 2 times more. I've read the spec and according to it they don't deflect almost at all. What are the other advantages and are there any disadvantaged besides the price?


5. End supports for ballscrew.
I saw a design where instead of end supports fixed to the gantry or frame there where bearings directly fixed to gantry/frame. It seams that it might be stiffer and cheaper. What do you think of it?

Kind Regards
Bart

[JAZZCNC]

#1: This guy is ok. Most of them are the same spindle and probably out the same factory.?
2.2KW WATER-COOLED MILLING AND GRINDING SPINDLE MOTOR WITH 2.2KW INVERTER VFD q3 | eBay

#2: Honestly there is no solution to match driving from both sides whether twin ballscrews, R&P or belt drive it's by far the best and easiest way to get accurate stiff machine. The wire/drum is just a bodge and best doing it correctly first time.?. . You'll only end back there in the end.!!
Unless your machine is narrow format, which it doesn't look, then single screw just won't work good. Making gantry wide helps but there's a limit before it gets unusable or feasible.

#3 Can't answer that question without knowing which make and what size linear rails. What I can tell you is even the lower quality units are capable of handling forces far beyond what you throw at them or what your machine design will handle, so it will fail long before they do.!!

#4: No contest here linear rails win hands down in every department by a very big margin. They can't be compared really in accuracy and performance terms.
The other advantages are longevity, they will easily last several years longer than supported rails, even in very harsh conditions. Hassle free usage, fit grease occasionally and forget.
Disadvantages over linear rails is they don't tolerate poor workmanship and misalignment.

#5: Show us and will comment but until then not sure what you mean.?

Now your design.?? . . . You have a problem Houston (several actually).!!! . . . . but this one Is a common mistake easily missed.!
The linear bearings on the Y axis can't possibly be fastened like you have them drawn.? Chicken and egg problem.!! Fasten plate on one set and you can't access the bolts for the other bearings.!!

I'd also consider using different thickness material for the Z axis.? Those thin plates, even with the bracing will resonate while cutting which will affect quality of finish. Don't rely to heavily on the Cad simulation, without inputting all the variables and cutting forces then it won't show this. I know from experience that if you plan on cutting hard woods or hard material like aluminium then your Z axis just isn't strong enough.
Just remember no matter how well built or how strong or accurate the rest of the machine is it's the Z axis that takes all the cutting forces
and if this flex's or resonates the rest don't matter a jot.!! . . . The machine is only has good has it's weakest link, this Z design is weak.!
( Also Don't rely on the Linear rail adding strength they don't really add much.!)

[GTJim]

Hi Adil, this is a very interesting build and a great deal of care has been give to getting the design right.
One question regarding the electronics, do you have a wiring diagram/schematic that you are using? If you do, could post it up as I'm currently struggling with this.

Jim

[Bart]

Hi,
Thank you for your interest in my design

I guess I should have started with some spec of the design.


Frame dimensions
W x L: 1400mm x 1800mm

Red and Blue is steel
Yellow is aluminium 10mm thick

Linear rails
X: 1800mm x 25mm
Y: 1350mm x 20mm
Z: 380mm x 20mm

Motors:Four Sanyo Denki 103H7823-1730@$80 each - 4A, 2,2 mH, 0,65R, ~2 Nm up to 600 RPM
http://www.plccenter.com/en-US/Buy/
any suggestion of better value for the money?

Driver :TB6600

Gantry
Steel square profile 80x200x (4 or 5)mm

Z axis aluminium plate is all 10 mm

#1: This guy is ok. Most of them are the same spindle and probably out the same factory.?...

That's what I'm hoping for but almost every seller has a picture of slightly different design so I was wondering if any of them is better or more reliable then other.

#2: Honestly there is no solution to match driving from both sides whether twin ballscrews, R&P or belt drive it's by far the best and easiest way to get accurate stiff machine. The wire/drum is just a bodge and best doing it correctly first time.?. . You'll only end back there in the end.!!
Unless your machine is narrow format, which it doesn't look, then single screw just won't work good. Making gantry wide helps but there's a limit before it gets unusable or feasible.

I've just google wire/drum and it's called cable drive and it's not as easy and cheap as it seams - if it is to be precise.
Belt drive is ok but it has a disadvantage of not reducing the moment of inertia of ballscrews. Right now, one ballscrew
has an equivalent mass of 60 kg, that is as mach as moving mass for X axis and same moment of inertia as motor.

#3 Can't answer that question without knowing which make and what size linear rails. What I can tell you is even the lower quality units are capable of handling forces far beyond what you throw at them or what your machine design will handle, so it will fail long before they do.!!

Yes, I've seen the spec for the linear rails and linear guideways and bot of them are pretty strong for the size.
So for the X axis, moving mass is 60 kg and max force when cutting would probably be around 500N which gives (simplifying) 125N per linear bearing.
However, when i asked about flexing of open bearing, what I had in mind was an order of magnitude. Is it 1 um, 10 um or 100 um of flex.

#5: Show us and will comment but until then not sure what you mean.?

Homemade PCB milling machine - YouTube
start watching at around 1:35

Now your design.?? . . . You have a problem Houston (several actually).!!! . . . . but this one Is a common mistake easily missed.!
The linear bearings on the Y axis can't possibly be fastened like you have them drawn.? Chicken and egg problem.!! Fasten plate on one set and you can't access the bolts for the other bearings.!!

Absolutely right :) - I've noticed it some time ago but i was too lazy to change all the constraints :)
here is an updated photo

I'd also consider using different thickness material for the Z axis.? Those thin plates, even with the bracing will resonate while cutting which will affect quality of finish.

It's 10 mm aluminium plate - what do you think it should be?
I was thinking about making a sandwich of wood and aluminium to dampen vibrations. What do you think?

Sorry for slow replay but i was too busy to write a post

Best regards
Bart