800X800 Build - Little to no knowledge of what I am doing…

Hello, I am new to the CNC world, I am a graphic designer who enjoys building stuff and tinkering with the oily bits of old cars…

I am looking to build a machine that is suitable for hardwood and occasional light use with soft metals. I started looking at off the shelf solutions such as the Workbee and Shapeoko machines I am seeing mixed reviews and they are not exactly cheap for something that I will likely outgrow. I also saw many suggestions of building a CNC from scratch, I am pretty handy with a welder and steel is cheap and strong so I think this is the best route for me.

I am planning on building the below machine out of 60x40mm box section 4mm thick, with the gantry and z axis support being 40x40mm 4mm thick steel using epoxy to bring everything into tolerance. The plate for the Z axis is planned to be 5mm thick.

All rails are planned to be 15mm from BST Motion on Ali Express (marked red on the screenshot below). The ballscrews (also from BST) are marked green, I was planning on using 1610 for the X and Y axis and a 1204 for the Z axis.

For drive motors I plan on using 265oz 2.8amp Nema 23 for the X and Y axis and a 175oz on the Z axis. Then run it through an Openbuilds BlackBox for control (all from Ooznest). Then use either a 1.5kw or 2.2kw liquid cooled spindle from BST.

This screenshot is a work in progress, I am also new to Fusion 360 but it isn't so different from the adobe stuff I have used previously. I still need to add supports/brackets for motors/ballscrews, carriages for the rails and work on the Z axis further.

Does this look like some sort of workable plan or do I need to have a rethink? Any comments, advice or constructive criticism welcomed…

Attachment 29728

Nearly slipped thru the net this one.!

Ok well first the base frame is pretty much ok, just a little bracing on the corner uprights would call it finished.

The Gantry needs a little work and to be honest, I think you are making it harder than it needs to be with this design. If it was me then I'd just stick with a conventional setup and stump up the extra cost of buying a piece of profile. Look for designs that use the "L" shape layout I came up with years ago, it's tried and tested on 100's of machines large and small. It also makes fastening rails and ball-screws much easier.

Next, the Z-axis is a complete right off.!!
I'm assuming you meant steel 5mm plate.? If so it just isn't strong enough and neither would it be flat enough. If going to use steel you'll need at least 10mm and even then it would need to be machined flat. So again keep it simple and go with a 15 or 20mm aluminum tooling plate, it costs a little more than steel but it's easy to work with and just gets the job done.

Lastly component selection.!
The X and Y ball-screws are the correct choice but I wouldn't go with 1204 for the Z-axis, either 1605 or 1610 would be my choice. 1610 allows you the freedom to connect the motor to the z-axis screw with a short HTD belt and apply a ratio, the belt drive helps with resonance and the ratio means you can tweak the Z-axis to do whatever you want and handle a heavy spindle if needed because the torque increases with a ratio.

Thank you for the pointers, I had wondered if my gantry was making life difficult, I was trying support the weight rather than hang it however I suspect I was trying to run before I could walk… The Z axis is the bit that I have found hardest to get my head around, I did indeed mean 5mm steel plate, your tooling plate suggestion seems like a much better solution, I will look at the Z completely afresh. Good to know that I am not too far off with the electronics, the idea of a belt driven 1610 to give me different gear ratios makes a lot of sense.

I really appreciate the help, building from scratch is a pretty daunting process but I do enjoy a challenge!

Quote:

---

Originally Posted by Tick

Good to know that I am not too far off with the electronics.

I really appreciate the help, building from scratch is a pretty daunting process but I do enjoy a challenge!

---

Ah, I didn't look at the electronics but you didn't really give much info other than motor size which doesn't really help as there are other factors that are important, like inductance and psu voltage, etc.Then you have the drives to consider which you don't mention.

Also, I don't know anything about the Oznest control, but if it's one of those Gbrl things which runs on arduino then I wouldn't bother, spend a little more and put a decent ethernet motion controller which runs on good software.

But your long way from needing any of that stuff and my advice would be don't buy any of the electronics until it's needed or close to being needed. Do the research and you'll see why Gbrl and all those tinpot control systems are a bad idea.

The building is easy enough provided you break it down into bite-size chunks, treat it like eating an elephant, one small bite at a time.

Sounds like I am getting ahead of myself. I will get reading take my time…

Progress! I rejigged the plan according to the feedback, ordered a bunch of supplies and got busy with the welder and drill press. I don't have a cnc or mill so this has all been done with basic tools and I am pretty happy with it so far.

While I wait for a few deliveries I am starting to think about motors and controllers. I do not fancy trying to sort out an Arduino, I have looked at the Openbuilds Black Box controller, Duet 2/3 but the Spark Concepts xPRO V5 looks like a good choice for plug and play.

How do I figure out what size stepper motors to use? I have allowed for Nema 23 brackets but I don't know what torque I need, from what I understand bigger is not always better… I am using 1610 ballscrews all around and the gantry weights around 35kg without the spindle.

Photos below show where I am at currently.

Attachment 30314Attachment 30316Attachment 30315

Quote:

---

Originally Posted by Tick

While I wait for a few deliveries I am starting to think about motors and controllers. I do not fancy trying to sort out an Arduino, I have looked at the Openbuilds Black Box controller, Duet 2/3 but the Spark Concepts xPRO V5 looks like a good choice for plug and play.

---

None of the above is my replay. All of them are a bad choice for a good router, ok for a 3D printer may be but for a router then I wouldn't touch them for a number of reasons.

#1 They use USB which is flaky and unreliable. Ethernet is the only way to go if you want a reliable router.
#2 They use GRBL control software which again is ok for 3D printers etc but for a router, there are much better and easier options.
#3 The drives are inbuilt and weak, if one goes down you lose the whole board.
#4 The all-in-one solutions often suffer from heat issues which shortens life spans
#5 The all-in-one solutions can be limiting on upgradability and they are not much easier in respect to wiring than separate components so the Plug n play aspect is a bit of a gimmick.

Quote:

---

Originally Posted by Tick

How do I figure out what size stepper motors to use? I have allowed for Nema 23 brackets but I don't know what torque I need, from what I understand bigger is not always better… I am using 1610 ballscrews all around and the gantry weights around 35kg without the spindle.

---

Don't fry your brain trying to work out motor sizing etc, you don't need to reinvent the wheel as it's all been done before by 100's on the forum so just go with what works for a router this size, which is 3Nm to 4Nm NEMA 23 or 24, they are both the same bolt spacing only the outside dimensions are different, 23=57mm 24 = 60mm.

However, the key to a good machine is not just the Nm rating of the motor but the inductance of the motor and PSU voltage plus the quality and type of drives you use. This is exactly why the above All-in-one setups are less than ideal for a router as you'll see when read further down.

The drives and voltage used are key to performance and reliability. Stepper Drives come in 2 main types Analog or Digital.

Analog drives are old technology and best avoided for several reasons which mostly revolve around how they handle resonance which can cripple the performance of a stepper motor.
Digital drives are far superior in how they handle resonance and current, this results in much smoother running motors and more power across the speed curve. This really is a night and day difference and can't stress how important it is to avoid analog drives.

Next up is PSU and drive Voltage, Voltage is the key to speed and torque from a stepper motor and for best performance from a typical 3nm or 4nm stepper motor, this is around 60Vdc, however, because of how steppers work they effectively become dynamo's when de-accelerating so when sizing stepper drives you need to leave a safety margin of about 10% on the Max voltage the drives can handle.
So if running 60V PSU then you need drives with a max voltage of 70V.

This is exactly why Bigger isn't always better and why often a Nema23/4 motor will outperform the same size or larger Nema 34 motor when running the same voltage. Large motors have much higher inductance and in a nutshell, this means they need much higher voltage to get the same performance as a smaller motor. The mistake new users nearly always make is running too low voltage with large Nm high inductance motors.

This is also why the All-in-one controllers with inbuilt drives are not good for routers running NEMA 23 motors in the 2nm to 4nm torque range, even if they say can handle 6A motors.
Because they only run on 24Vdc they can't possibly give the performance because without the volts they can't push the current through the motor's windings unless the motors are very very low inductance. Like I say for 3d printers which are using Nema17 size or less which have very low inductance they are ok, but for a good router, they are useless.

Now, stepper technology in the last few years has transformed, and now the BEST by a long way and not much more expensive than a standard stepper system is the Closed-loop stepper system.
These use digital drives with advanced electronics that connect to a stepper motor which is fitted with an encoder on the back which monitors the position and feeds this back to the drive which corrects for any positional error and if the error goes out of a set range sends out a fault signal.

IMO this is the only way to go these days because the difference in price between a standard stepper motor/drive setup and a Closed-loop system is not very much and they are light years better.

Controller wise then Ethernet-based controller is the only way to go if you want high reliability and best performance. USB is flaky and easily affected by electrical noise and I wouldn't touch it for a router that runs high-frequency spindles etc.
Likewise, GRBL based control software which requires G-code senders, etc are a faff about and are often not optimized in how they calculate cutting trajectories, etc which gives smooth motion and performance.
Compared to dedicated control software and hardware combinations like UCCNC software with an AXBB-E controller they are positively still in the stone age.

All these things work together to give a balanced machine, get one wrong and the machine will be compromised in some way, get a few wrong and you can easily cripple the performance.

Often (99% time) it's a false economy trying to save money with All-in-one or cheap kits as they are just not very good and the upshot is eventually you end up ripping them out and starting again. Bite the bullet and do it right the first time and you'll save money and a lot-a-lot of stress.

The setup I would recommend and actually fit on machines I build is this below. ( I can also supply these if you decide to take this route)

AXBB-E ethernet controller with UCCNC software.
4.5NM Lichuan Closed loop stepper system with 70Vac drives.
60Vac torodial transformer

This setup combined with your 10mm pitch screws will give you first-class performance and reliability/repeatability. With velocity well in excess of 10mtr/min and provided the machine is constructed correctly with a suitable spindle will easily allow you to match correct cutting parameters for most tools and materials, which wouldn't be the case with a lesser setup.
Not cutting correctly results in poor tool life, broken tools, poor finish quality, and lost material when the shit hits the fan not to mention the all-important loss of time doing it all again. Proper hair pulling stuff is avoided which all add's to the savings and hair long term.

Trust me on this don't rush out and buy the first thing that looks easy or cheap because it won't be it never works out well. Even If you can't afford the correct components, wait and save up until you can, don't be tempted to buy rubbish just to get it up and running.!. . . You will regret it, I've seen it time and time again.

As a first machine I think the structure you have built looks fine, and will get you used to what CNC is all about. You will probably tinker with it and stiffen the gantry but the main thing is you will be learning and listening to the machine.

My first machine in around 2008(?) was made of plywood, with a Bosch router, 20mm un-supported round rails, 16mm ballscrews, 3Nm Nema 23 steppers, and a 24V all in one board (overdriven at 32V). I was not sure if this was a hobby for me, or if it was a passing phase, so went in with a minimum approach. I quickly found the shortcomings and went about upgrading.

Fast forward to 2021 and 4 machines later, plus lots of trigger's brooms in-between, I have a machine I'm finally happy with. The only parts I still have are the stepper motors and ballscrews !

Thank you for such a well thought out response JAZZCNC, all your points make a lot of sense and I really appreciate the help! The controller side of things is where I was getting really lost, there are so many options… You say that you can supply components do you have a website that I can look at?

Quote:

---

Originally Posted by routercnc

As a first machine I think the structure you have built looks fine, and will get you used to what CNC is all about. You will probably tinker with it and stiffen the gantry but the main thing is you will be learning and listening to the machine.

---

Thanks, I am already looking at the gantry thinking it could do with some gussets in places and triangulation to improve rigidity. I imagine this will be a never ending cycle of building and improving…

Quote:

---

Originally Posted by Tick

Thank you for such a well thought out response JAZZCNC, all your points make a lot of sense and I really appreciate the help! The controller side of things is where I was getting really lost, there are so many options… You say that you can supply components do you have a website that I can look at?

---

At the moment we don't have a Website, we are working on one at the moment but it's not ready to go live just yet.
If you contact me at the address below I can send you any info you require and answer any questions you may have. I will also provide you with my phone number so can give me a ring if prefer or need advice or questions answering on products or anything CNC-related really.

[email protected]

Quote:

---

Originally Posted by Tick

Thanks, I am already looking at the gantry thinking it could do with some gussets in places and triangulation to improve rigidity. I imagine this will be a never ending cycle of building and improving…

---

I'm sure others will agree but don't go too far with adding bits, sometimes it's easier to just use the machine find its weak spots learn from it what you would like in a machine then build Mk2. It Will save you money long run as you are not plowing money into a machine that you will only end up hitting a brick wall with how good you can make it sooner or later and you'll get a better machine more quickly. Hence why some like Routercnc are on Mk4, it's the best way.

Just logged in to update this and realised it has been just over a year since I started the build… Work got in the way for a good while but I am back on the case now. I purchased the motors, drivers and an AXBB-E from Jazz CNC. So far the wiring has been fairly straight forward, I am hoping I have got it all right thus far. I have been trying to solve a few things but the more I read, the more I realise how little I understand…

How does the alarm system work on the LCDA86H drivers? From what I can work out if wired correctly I can have it work much like a limit switch if a step is missed.

Does anybody have a clear wiring diagram for a xyz touch probe with an AXBB?

Finally does anybody have reliable sources of information for setting up a Huanyang VFD with a Huanyang spindle? The VFD is 2.2KW and the spindle is 1.5KW. I have read the instructions and various bits online but feel like setting it up is going to be guess work…

I am hoping to get the machine running in the next few months

Attachment 31173Attachment 31175Attachment 31176Attachment 31174