Reducing the space in my garage.

[Rich]

Hi all,

I thought I’d better introduce my build and thank all those that have contributed to this forum, it’s helped my a lot already.

I should own up to the fact that I have started my build. I was having trouble getting my head around the challenge, so started mocking up an idea with bits of sheet material I had hanging around.

The design started in sketchup, I will post it up on the site once I have found a host.

I’m quite a practical person, but I know my limits, especially when poking a stick into an electrical circuit. If I could borrow from your collective experience, that would be great.

Here’s the outline, photo’s to follow.

The machine will be used for profile cutting sheet materials using a drag knife or a small router. It’s for personal use, so no more than 10 operating hours a week. The cutting area is 1600mm x 800mm, with 100mm clearance between the gantry and the cutting table. The main material to be cut is natural wood nothing too hard – mostly European species, circa 1600 on the janka hardness scale. I will be aiming at a feedrate of 8m/min.

My target resolutions are 0.1mm the X and Y axis and 0.05mm on the Z-axis. The Z-axis is the priority. I appreciate they are not considered particularly tight by the standards found in this forum.

Although I had read and viewed as much as I could before starting, the technical stuff takes a while to sink in, so the machine was designed with http://www.cncroutersource.com/ as the main reference point. It’s a great resource for beginners like me, although I have made a mistake with the position of the gantry over the X-axis carriages. If the machine works, I will correct this.

Frame
The frame has been built using dexion bought from the local scrap yard. The frame sits over and will attach to a woodworking bench while the machine operates, borrowing mass and stability. The table is made from mdf, its a torsion box. Dexion is being used internally and externally to create the structure. The mdf used isn’t the moisture resistant type.

To hold work down, I’m thinking t-nuts for two thirds of the table, one third vacuum.

Gantry
The gantry has been constructed, it is 10kg and when I give it a shove, it runs three quarters the length of the X axis. Its dimensions are 1200 x 50 x 150mm. Made from two lengths of aluminium box section 50 x 50 x 3mm, they have been spaced by and bolted to an extruded aluminium I- beam, 50 x 25 x 2mm. The design flowed from what was available from the scrapyard at the time, bearing in mind the designs I’d studied online; 3mm is a little light for creating stable joints in aluminium, steps have been taken within the design and will be taken within the final assembly.

The ends of the gantry are connected to aluminium plate, which make up one face of a torsion box that connects to the carriage running on the X-axis.

Z-Axis
The Z-axis is a box wrapped around and over the Y-axis, think ‘Joe’s CNC’. Drive is at the rear, which I appreciate is sub optimal. I’m assuming this will introduce additional wear and could lead to binding. The front and rear plates of the box are made from 6mm aluminium plate. The box is 170 x 89 x 230mm.

A Tr 12 x 3 trapezoidal screw made from steel C45, with a derilin antl-backlash device will provide motion to the Z-axis carriage. Tr 12 is much more than I need, but is defined by the backlash device. The screw will be fixed and supported. The spindle will be supported at each end by 6mm aluminium angle. I've looked a micro-leadscrews, but dismissed this option.

The Z-axis will be belt driven, no reduction. I will be using 2 x 15 tooth T5 pinions, providing the 6 tooth engagement required and a polyurethane timing belt, 280mm long. The requiring a centre distance will be102.5mm.

Linear Rails
At the time, and without the benefit of the information contained within this forum, cheap linear motion was the main stumbling block. I went with Makerslide, which might be considered unfortunate:

http://www.mycncuk.com/threads/6569-...ght=makerslide

The Makerslide has been mounted horizontally on two dexion rails that make up each X-axis. The rail has been mounted in parallel on the each vertical face of the aluminium box section that make up the Y-axis. The Z-axis consists of four sections of Makerslide, two sections working as rail attached to a carriage made from 6mm aluminium plate, two sections acting as structure for the router/knife.

The X and Y axis solidly resist the forces and moments of inertia described by cncroutersource when tested firmly by hand:

http://www.cncroutersource.com/do-it...NC-router.html
http://www.cncroutersource.com/build...nc-router.html

The Z-axis needs a little more work to make it completely solid. I will create a pattern, against which I can tighten and lock off the v-wheels on the carriage.

If I keep the rails clean, then hopefully the main issue will be wear.

Visiting the scrapyard the other day I pulled out a section of V-track (US Patent), 2 meters or so long, made of carbon steel. They may well become part of the design, dependent on the rate of any wear.

Motion
The intention is to drive the X and Y axis using reinforced timing belts, per Mike Everman

http://bell-everman.com/products/lin...elt-linear-sbl
http://www.cnczone.com/forums/linear...-ever-say.html

The belt to be used is Brecoflex T5 10mm steel reinforced belt. The system requires reduction, circa 3:1, over above. The gearing for the X-axis will be boxed in within the ends of the gantry protecting the exposed belt from debris, I have some work to do to find the parts at ‘beltingonline’, to make the gearing fit.

http://www.cnczone.com/forums/linear...er-say-14.html
http://www.cnczone.com/forums/linear...er-say-20.html

Per the above thread, the gap between the upper belt and the rollers must be held at 0.05 -0.1mm. and the belt must be seated within a channel preventing it track laterally. Resolving these two points has not been too problematic. I have a few more hours on the X-axis before it is finished, I will be using four UHMWPE pads to reduce the reliance on bearing.

In the real world the end result can looks something like this:
http://www.youtube.com/watch?v=SpIuQ...KmL6gQvVQ-GvWw
http://www.youtube.com/watch?v=T9QU3YQe-Sc

Does anyone here have experience of the Bell - Everman approach?

Drive
This feels tricky. The recommended approach is 3Nm, Nema 23, low inductance motors, ideally run at 65-70Vdc on 75-80vdc drives.

The motors I have in mind are: www.cnc4you.co.uk/index.php?route=product/product&path=83_84&product_id=364 . For what I have in mind, I suspect it is preferable to set up the wiring in series i.e. high torque/low speed.

I’m still considering what steppers to use. I understand, 1/8^th micro-stepping should be sufficient for my needs. I still have a little time in hand before I need to raise the funds.

Power supplies, the options are between regulated and unregulated supplies. It looks like unregulated supplies are better suited to deal with fluctuating loads. More reading required on this so I can gather the right parts.

Control
BOB. I’m currently looking at the OPPB V3. A four axis controller offering the required 25KHz step frequency and 10A (80% total) load requirements, if wired in parrallel. Running 60Vdc means I would lose some functionality. It is built with MACH 3 in mind. More reading required.

Mach 3, I’ve used the product briefly in the past. It has a forum for support not provided by the manual.

So, even without the addition of photographs, it might sounds like there are a lot of weaknesses in the choices I have made to –date. At this stage, I’m not overly concerned, as it takes time to build the skills and insight required. The hope being, when it come to building v.2, I will have all the parts and skills required to build a much stronger machine.

Thanks in advance for any comments!
Rich

[Rich]

Table design using dexion and images of the build so far.

The front plate of the z-axis is currently made from 2 x 3mm aluminium plate. The final build will separate these two plates, increasing stiffness and absorbing vibration., I will use sheet ply and a vibration dampening material.

[EddyCurrent]

If this version 1 will suite your requirements why build version 2 ?
If you think it might not suite your requirements then go straight to verison 2 now. After reading your post I would go straight to version 2.

[Rich]

Thanks Eddy. I did give that issue some thought. I decided to factor in my own in-experience and concluded that I would carry on building this slightly fragile machine first. I may need to cut plate aluminium in the future, hence the need for v.2.

Spent last evening reading the support offered by Gecko: http://www.geckodrive.com/support/st...or-basics.html . It offered a very helpful understanding of why unregulated supplies are preferrable i.e. motors behaving as alternators. On a three axis machine, using 4 x 3.1Nm motors (CNC4YOU) in parrellel, the numbers appear to be 11.088 amps and 57.25 volts. The numbers on power requirements will need double checking, they seemed low.

Anti-backlah Nuts
Is this a good buy: http://www.damencnc.com/en/component...--spindles/101

I was originally looking for something that worked with a Tr10 x 3, but I have been unable to find a workable solution, other than the MDL-2-TR 10x2D Special Nut, sold by Merchant Dice:

http://www.ebay.co.uk/itm/CNC-Z-Axis...item483f9663e7

[EddyCurrent]

Igus do these, I expect them to be top quality coming from them

http://www.igus.co.uk/wpck/2371/dryl...zgewindemutter

On a four axis machine, using 3.1Nm motors (CNC4YOU) in parrellel, . . .

Did you mean a 3 axis machine with 4 motors ?

[Rich]

Lol - yes, I'll edit. Thanks for the link!

[irving2008]

If your drivers can take it go for the higher voltage as this translates directly to upper speed limit. Current requirement is typically 60% of motor rating, so.60% x 3 x 4.2 is 8A approx. This means you can get away with smaller capacitors .1 x 8/70 is approx 12000uF @ 100v, ripple rating can be lower at around 4A.

Re BOB, correct only 5v for signalling and 24v for limit/home switches needed. All i/o should be opto-isolated at the BOB for inputs and at drivers for motion control step/direction or relay outputs for other controls.

[JAZZCNC]

Could someone please let me know what the answer is here - I just need to have these thing clear in my mind as all the calculations can be a bit confusing!

I've mentioned the ideal voltage to run these motors at and why in 2 post's now but you seem not be listening.! . . . One more time then with little more WHY.!!

The Gecko site suggestion is working on the safe side but in practice the modern hybrid stepper motors we use and in particular the low inductance 3/4Nm available to us easily handle 70Vdc. Combine this with Modern Digital drives which handle current, motor heating and resonance much more efficiently then performance is greatly increased in several ways. More speed, more torque at higher speeds and smoother running motors.

Regards the transformer then all you really need is to size it so it outputs the voltage required and some extra amps for overhead. The volts are the main thing to concentrate on because if you go to high here you can blow the drives if tolerence peaks higher than there max rating. This is why 80V drives running 68Volts is safe.
You can get away with 500Va and 2x50V wired parallel or 2x25v wired in series with 3 x 4.2a motors without any troubles and I've run these with 4 motors in past without any issues but the 625va is the better option with 4 motors as it gives higher safety margin for say 4th axis.

Reason you can do this because all 4 motors are never pulling full amps at same time and the way the unregulated transformer/drives work they only pull amps on half the PWM cycle so don't draw the full rating of the motor. Combine this together and you can get away with Transformer Amps 60% of total motor amps but you still need the voltage to get the speed/torque from steppers.

Like irving says you can get away with much less caps and using 3 x 4700uf wired in parallel is my prefered way as they are cheaper, easier to find and lower ripple I believe.!

Regards the Bob then not sure what your meaning because voltage is very important to the BOB.? . . Feed it more than it's I/O can handle and you'll fry it every time, infact on the cheaper Bob's like the Oppb v3 you'll drasticly shorten it's life if you run any where near it's limit.

99.9% of BOB's will have voltage requirement or voltage range to operate and run things like Opto's and relays etc. The Motor output side of a Controller is often 5v logic and all the BOB does is pass it thru or in some cases will filter and boost the signal to 5V if lower.
Only the standard I/O side can vary regards voltage/current and even then your BOB will have a limit it can handle. The cheaper BOB's will restrict you to 5V others will be 24 or 30Vdc.
Again in practice your better running 24V thru I/O and the safety system limits Etc to help resist any noise interference.

DONT buy a cheap BOB they are very important and seriously under rated in how they can negatively affect the machine performance and shorten the life expectancy of your Hair.!!

Hope this helps.!

[Rich]

Thanks Irving. I have a parts list together now )

Thanks Jazz, I won't be adding a fourth axis, so the 500va transformer is all I need.

The specs on the BOB haven't changed - OPPB v.3. I have a 12 volt supply to hand, although I may well need to supply 24volts for the alarm circuit, quoted as 30v @ 100mA.

Does anyone have a recommended source for the Leadshine AM882 stepper, someone on Aliexpress perhaps?

[Rich]

Been a while.

The table is finished - top need a final coat of urethane. The internal section of the table was coated with an urethane, grout like in its consistency, and has dampened the vibration. I haven't measured the frequency response - but visually I'm happy. The table was built with three internal vacuum boxes. They work, but I messed up the table top. When it came to the final construction, it tuned out that I'd not thought through or simply measured properly the height of the sections I was using. At the last minute I changed my approach and made matter worse than they need to be. I can fix the residual issue - its just annoying to have to fix the issue.

The table slides over a wooden bench and locks tightly into place. Its bloody heavy and a little clumsy - so a couple of height adjustable wheels are required to make this a little more practical. They will aid the lifting of the table when the front end is moving over the side ramps. Its set up so when table drops into place it meets its mark and all the feet then touch the floor. The addition of the extra weight of the gantry has just made the lifting of the table tricky - no place to grip at the right height for lifting.

Once its over the bench its becomes a solid unit - aided by a couple of clips and jams.

The machine has come together. Everything in place except the leadscrew - just need to get that on a lathe. The belts and pulleys have been fitted - the belt has yet to be bonded into place. Various threaded holes in plate aluminium have been re-enforced with backing nuts. I've yet to add any thread lock. Flange bearing, Thrust bearing and bright bar, all purchased from Technobots: http://www.technobotsonline.com/ - fast delivery! I just need to find a couple of crank handles - just in case I need to back off a motor or two in the future. I have two, but they have the wrong ID.

The parts for the PSU should start arriving tomorrow. I will be following the plans set out within this thread:
http://www.mycncuk.com/threads/5075-...ght=Confirming
A Great thread.

Having one motor to hand, will help with the setting out of the cabling.

Started looking for a motion control cards - this seems to be a fair summary of all the basic points. I suspect it will be the last thing I buy - those 12 months will soon tick by:
http://www.automation.com/pmd/Motion...hineDesign.pdf

Dust extraction is mostly finished - up and over the table, as per the spindle cable.