Ready to pull the trigger and could really use some input.

Hey all,

I've been wanting to build a CNC for quite some time now, and I am trying to get all of my ducks in a row. The machine will have about a 12"x12" work area with roughly 6" of travel for the Z. I mainly want to machine plastics and some soft metals (brass, aluminum) and maybe some wood engraving.
My biggest problem at the moment is the electronics.
These are the motors and drivers that are being considered:
https://www.aliexpress.com/item/2255...c00Y9kuki&mp=1 (x4)

And this is the Spindle that I've been eyeballing:
https://www.aliexpress.com/item/2251...c00Y9kuki&mp=1 (x1)

I was planning on using a DDCSV3.1 controller with MPG ( https://www.aliexpress.com/item/3256...c00Y9kuki&mp=1 ) to keep some manual control available for various reasons.

I am absolutely stumped on selecting a PSU for this configuration. I have read so many posts on selecting a power supply and I keep falling further into the rabbit hole than I need to go and end up getting frustrated once the conversation begins to be beyond layman's terms. I could really use some help with this and would love it if you all could help me understand how to come to a resolution.

Also, is it correct that if I am using the DDCS for an offline controller, I would not have the need for a BOB or Mach3/4 software?
Would I only need software for modeling and creating the g-code?

All of your advice/input would be greatly appreciated!
Thank you!!

Hi Clef,
Firstly the 2.2kW spindle is way overkill for what you want and its too big for a 12" machine. An 800W unit will be compact and do the job. I use a 600W DC spindle (see images) for that sort of thing and its fine. Your design will have to have a structural bias. A) either towards aluminium which means it will need to be very very stiff B) towards timber and plastic which means it can be less stiff and simpler.

Now a 12" machine will use 4 or 3 motors. The size you have picked is very grunty and 4A. If you use 3 motors then the absolute max current is 3x4=12A but this is unlikely to occur so a 10A supply is fine. Your controller is 24V so you need a 24Vx10A supply so that's easy. Now if you use 4 motors 1) The same applies as the twin motion system you only need a smaller motor on the twin drive say a 2A motor. There is no need to have 6Nm (2x3Nm) on one axis and only 3Nm on the other.

I just noticed you say 4x motors so this comment is applicable. So pick smaller motors for the twin drive. I run this machine on a single 24V 10A supply and its quite comfortable even though it has 4x3Nm motors... On my current build I have downsized the motors as with ballscrews the forces available are very big 100's of kgf with smaller motors. Once you try to get power supplies over 10A they become very $$$. You have also picked hybrid motors and unless you are trying to make a very fast machine I don't believe the extra $$$ for these are worthwhile. The $$$ can be put into better bearings or spindle which is better spent. If you want extra speed then a 24V system is too low a voltage. Use 48V. With the $$$ saved on hybrids go higher again 60V and the motors will spin fast nicely. Can't comment on the DDCS , I use UCCNC and I'm very happy with it...hope that helps. Peter

Hi,
those stepper motors will give you plenty of power. I see they are 2mH inductance which is fair-to-good and so should be pretty fast, but you will need a high voltage supply to get
those speeds, I'd recommend 60VDC. You'll buy a power supply once, make it the right one and you'll never have to go back and do it again.

A 2.2kW spindle is a beast, that's close to 3hp! If you buy well then you won't ever regret it and you'll never have to go back and buy another, so I don't see any harm in going
for 2.2kW. The price of the kit is very attractive, but also means that the components will be of low quality to match the price. A higher quality 800W or 1.5kW unit might actually cost
less. Most cheap VFDs work well enough but they do go bang eventually. If you want better a quality VFD (at higher price) then consider Delta or Hitachi.

I have no idea what DDCS is.

I use Mach4 ($200USD) and have done for seven years, its good. I use an Ethernet SmoothStepper ($190) and have done for seven years. Its a very good idea to use a breakout
board as it provides buffering and a reliable way to connect wires. Without a BoB you risk damaging the motion controller with the smallest of slips. I have used cheap and cheerful
bi-directional breakout boards such as a C10 or C25, but recently I made my own breakout board. It has three ports (51 inputs and outputs) with six differential Ste/Dir pairs for use
with Delta servos.

I have used Fusion 360, and have a subscription, $500usd/year, and overall its pretty good and compares favorably pricewise with its genuine competitors. Ther are cheaper and\
even free CAD/CAM software for generating Gcode, I can't comment on them.

Craig

Thank you guys so much for the input and information!

Please pardon my lack of knowledge here. I have just read too many posts of people either underpowering their motors or frying drivers by not selecting the correct PSU.
So, When selecting the PS I want to not exceed the voltage rating of the drivers(70v) and also have efficient enough amp rating to run all motors(4ax3 or 4ax4)? Maybe something in the ballpark of 60v 600w?
I don't know why I just can't wrap my head around this, but the wattage is just the product of Volts x Amps correct? or am I missing something? ....Man, electricity (among many other things) makes me feel stupid.

There are a couple of pictures of the frame that I plan on using. It is very robust, but I'm sure could be reinforced in a few spots, such as the Z axis.
My main goal at the moment is to at least get something moving to where I can get acclimated. I'm sure there will be many adjustments and modifications to come along the way, but I am so new to this, that just getting motion out of this thing will make me feel like a 6 year old at Christmas!


Thanks again for all of your help!

.

Hi,
that looks like a very substantial machine and I think will make a very good conversion to CNC.

I would definitely stick with the largish steppers you have selected, you'll want the torque to accelerate all that mass. If I were you I would consider 80 or 100VDC drivers,
the higher the voltage the faster the machine.

60VDC 600W would be adequate, but 80VDC and 1000W would be better. These large power supplies are expensive and most people buy switchmode supplies on the basis
of cost....but they are not the best choice. Linear supplies are far more rugged and forgiving of overload.

https://www.antekinc.com/1000w/

These are all 1000W, with output voltages from 35VDC to 80VDC.

If you want to make improvements to your machine I would start with the lead screws. Lead screws always have backlash, and backlash, even the smallest amount really
screws CNC toolpaths. I would suggest you replace them with at least 16mm diameter, 20mm diameter preferred, ballscrews. Rolled ballscrews (C7 grade) are quite good and
well priced, but if you want good, then ground ballscrews (C5 or even better C3 grade) are where you want to go.

Craig

Hi Clef - Yes watts = VxA. Yes the voltage of the PS has to be less then the voltage recommended for the drivers. The PS does not have to have the full capacity of the motors eg 4x4A as all motors will not have the same demand at the same time. Some say 0.5x full some say 0.7xfull. Either way if you underspec the PS its easy to fix use 2x10A supplies. Your main issue at the moment is picking the operating voltage of the system. Being first build you should keep it simple. Being your first build I suggest you get one of these:

https://www.oyostepper.com/goods-120...per-Motor.html Use one 24Vx10A PS

Will work fine, I've used three and they are simple. Its easy to upgrade electronics once you get your head around everything else if you need to. Peter

Hi peteng,
did you look at the pics OP posted?. This machine is VERY substantial, at least 200kg by the pics, those tiny little motor and drives you've linked to are a joke, he wants/needs torque in the region of 400oz.in or the machine will be
much LESS than it could and should be.

Craig

Hi Craig - Yes the imaged machine is substantial but its not the machine he is building. If he uses those leadscrews his speed will be down but have lots of force. Clef is in the fuzzy front end of the project where everything seems to be complex. The OYO driver will run an N23 motor at 3A and if the ballscrew is picked correctly it will produce all the torque and power and speed he needs. Plus being a one box solution its easy to wire and understand. Once he gets down the road say 6 months its easy to upgrade if needed. Clef has a long way to go yet before he gets a complete picture of what he wants to unpack at Christmas. He needs to fix some solutions and move on to all the other considerations. Most people just pick the biggest motors, if you do the math they are usually way overdone for the job. Then the whole project can be reviewed in context. Keep at it Clef. Peter

Hi,

Quote:

---

Yes the imaged machine is substantial but its not the machine he is building

---

Ah??.....that's not what I read:

Quote:

---

There are a couple of pictures of the frame that I plan on using.

---

I took that to mean this is the machine frame he actually has?

Quote:

---

Most people just pick the biggest motors, if you do the math they are usually way overdone for the job. Then the whole project can be reviewed in context. Keep at it Clef. Peter

---

Yes I have seen that, people select the biggest 34 size steppers they can find, only to find they go like crap because of the high inductance. That's not what Clef is proposing here, he has linked to
24 size 425 oz.in and very acceptable inductance, 2mH. I think those would be near perfect. If anything I would up the drivers just a bit, but even the 70VDC drivers hes has proposed should do nicely.
I would suggest that getting the 'right' components for the job first time is the cheapest solution always, you don't have to go back and re-purchase because you know now what you did not know then.

My advice would be get the steppers/stepper drives and the biggest and best power supply he can afford. Then he can get started.

In my previous post I suggested that ballscrews would be advantageous, and they would, but they could be done 'down the road' without any cost penalty, he has the lead screws already. Just hook the motors up
and make some chips.

He has suggested that the Z axis needs beefing up....and it may do, but that too can be done 'down the road'. I see little point in trying to beef something up until you have real world evidence
that it is needed. Looking at the pics suggest to me that whomever designed and built that machine has done a good job, and it may prove that no stiffening of the Z axis is required, we can only hope!

I would, at least for a while, defer the purchase of the spindle. The spindle and spindle drive is usually the most expensive single part of any machine. The chances of getting the 'right' spindle
straight up is slim. The notion of 'right' is highly debateable also. My suggestion would instead be to get the motion controller/BoB/CNC software sorted. Those are items that can be researched
and purchased with good certainty that they will be acceptable in the long run, ie no going back and spending yet more money.

With the steppers/drives, CNC software, motion controller and BoB the machine will be able to move on a CNC toolpath. Then a 'cheap as chips' spindle, or even a router spindle could be added
and now start making parts. Then Clef would learn first hand what constitutes the 'right' spindle, with a view to saving up and buying one. Typically the 'right' spindle always totally blows your
budget! I am not in favor of buying something as a temporary measure, but I think in this instance its probably the right way to go.

Craig

Hi Clef - Usually the motors are sized once the structures are defined so you know the weights you want to move. Your machine at 12" travel is small and won't need 20m/min rapids so a high voltage system is probably not required. Nice to have if you are trying to make a very fast high G machine to maximise output but I don't get that's what you need. The images you have are a mill not a router. As I said you will need to bias the structure to metal cutting or soft material cutting. But a cnc can never be too stiff. Say you use 10mm pitch ballscrews then at 500rpm the machine is moving at 5m/min that's fast for a 12" travel doing hobby work... In timber you may cut at 3 to 4m/min thats fast too. A lot of cutting will be at less then 2m/min. A 10mm pitch screw at 3Nm will produce 170kgf that's more then enough to cut aluminium....Peter

Hi Clef - to clarify, the images posted are the actual machine you are going to convert to CNC? I say this as you state it is a 4 motor machine but the images are a 3 motor machine? If the imaged machine is the one you are going to use what is the pitch of the leadscrew? :) Peter

Peter, Craig, thank you both so much for loads of information and suggestions.
I think my explaination of what I would like to have in the long run has been a bit murky. My apologies for that.
The machine that I have posted IS the base that I physically have. My previous employer had purchased it with optical measuring gear attached to it. Basically it was a glorified comparator.
I aquired this with no monetary cost as they were not using it and had relocated it to a storage attic within 4 months of their purchase. They happily gave it to me at the cost of just driving it off of the property.
I don't have any details of the specs, besides the linear rails.

Like most newbies in the DIY CNC venture, cutting steel would be the best outcome. However I understand reality, and if I can at least get this thing cutting aluminum, I would be more than extatic.
The base probably weighs close to 300-350lbs, so I could only assume its going to take a decent bit force to move the ways with any considerable speed.

For starters, I'm just looking to give this thing adequate enough electronics to not be disappointed once its up and moving. I'm not exactly looking for the best of the best at the moment. Yes, down the road I'm sure I'll be finding things to be replaced such as screws and rails, along with any structure modifications that may be required.

I had mentioned 4 motors with the intent that down the road I could add a 4th axis. This doesn't need to happen right from the start. So I apologize for any confusion there.
As for a budget, I don't necessarily have one. I'm willing to spend a few bucks to have a substantial machine. Honestly, if the screws and rails will work for the time being, then I'm not going to fuss with them unless it's absolutely necessary.
I am at the very beginning of building my first machine, so I still have a BUNCH to learn.
I can't thank you guys enough for all of your input and information.

By the way, my name is Caleb. I am extremely excited for this project and am sure it will be a Rollercoaster of "Eureka's " and "Oh S#!t's". But I'm buckled in and ready to go!

Thanks again.
-Caleb

Sent from my SM-G991U using Tapatalk

Hi Caleb,
as I posted before that machine looks to be very substantial. There's a lot of water to pass under the bridge before you get a 4th axis.

For the moment get three steppers and drives, and a big kick-arse 60-70-80VDC power supply.

Then you need to decide on CNC software and motion controller.

I use Mach4, and like it, and its performed well for me for seven years. Peter uses UCCNC and he reports good results. These are not the only CNC software that runs on
Windows PCs but they are two leading contenders.

Both UCCNC and Mach4 require a motion control board, I use an Ethernet SmoothStepper while UCCNC requires a CNCDrive (manufacturer of UCCNC) controller
like a UC300. Both control boards benefit from having a breakout board. One of the more sophisticated brands is CNCRoom. They make an MB3 for the Ethernet SmoothStepper
which has all three ports developed for 51 IO's or the UB1 specifically for the UC300 board, not sure how many IOs, a few more than the MB3 is my guess.

There are cheaper options for breakout boards for both boards but they miss all the features of MB3 or UB1 respectively As a ballpark figure a MB3 or a UB1 cost about $200USD.

For instance a Mach4Hobby installation would cost:
Mach4Hobby $200USD (perpetual license)
MB3 breakout board $200
Ethernet Smoothstepper $190
TOTAL $590

A UCCNC installation would cost:
UCCNC software license (perpetual) $70
UC300ETH motion controller $150
UB1 breakout board $200
TOTAL $420

Note the Mach4 license allows for up to five machines for one license purchase whereas UCCNC requires a new license for each new machine. In truth there is little
difference between the two in terms of overall cost.

If you want or are happy with Linux then LinuxCNC is s good choice. LinuxCNC is open source and free. Mesa motion control card are reasonably priced, say in the region of $200-$250.

Any of these solutions would work well. There are some other commercial solutions that are very nearly a turn-key arrangement, albeit with minor loss of flexibility but very
competitively priced , Centroid Acorn and Masso are two that come to mind. I have not used either but they gather favorable comments on the forum and well worth your time to
research.

Craig

I am going to go ahead and order the original stepper/driver combo(x3) that I had posted in the beginning of this thread, along with the controller... Holding out on the spindle for now, until I do some more research.
I like the idea of not having to have my laptop connected directly to the machine for the g-code and DRO to do its job.
The mills that I have run in the past( Makino F5 and Various HAAS machines) have always had a "Cabinet" of sorts that allow offline control and location display, so I guess its just a comfort thing to me.

Now, back to the PS. These drivers are rated 16-70VDC. I understand that I don't want to exceed the 70V, but how close can I get? I'm looking at the units that Craig had suggested and am wondering if I can go with 70V 1000w? or should I leave some margin and go with something between 60V and 70V for any reason?
From what I understand, the drivers will regulate the necessary amperage to the motors, so 1000w should be safe?

If and when I want to add a 4th axis I could still run with this same supply?

This controller comes with (x2) 24v 75w PSU. So can my main 220 input be wired in parallel to all 3 PSU? (220V-------power switch-----70V-----24V-----24v)?

My next question would be wire size/ style. I have read of people using anything from 22awg to 18awg shielded wire, but couldn't really find a definitive reason of why they chose one way or another.
Also, would it matter if it is stranded wire or solid?

I truly am sorry for all of the "newby" questions, but I knew that I could reach out to this community with confidence that you guys could provide real world experience and knowledge.

Keep being AWESOME you guys!
and THANK YOU so much for all of the kind help.

-Caleb

Hi Caleb,
if the driver is rated at 16-70V technically its 70V max. But its up to the quality of the manufacturer to determine its reliability at 70V. Plus the motors create back emf and this can create spikes that can damage the driver. Ask the forum if anyone has used them at 70V... The 4th axis depends on its current so jump that issue then. Its unwise to put switched power supplies in parallel I think. But you can run 2 motors from one and the other two from another no problem. Some people use one small PS per motor. The larger the wire the better. There are tables with amperages vs run length vs wire size if you do some searching. No difference solid vs strand. I use heavy duty 4 core speaker cable I forget the size... I know its thicker then the motor wires. Peter

Hi,

Quote:

---

Thank you for your response. How will Mach operate directly into the servo motor amplifier?

---

If they're any bloody good 69.95VDC, if they're not good enough they'll pop, so buy better ones next time.

Quote:

---

From what I understand, the drivers will regulate the necessary amperage to the motors, so 1000w should be safe?

---

Yes, the drives use PWM to regulate the current to a safe level, so use as big a power supply as you can afford.

Quote:

---

Also, would it matter if it is stranded wire or solid?

---

No difference electrically but there is a lot of movement/vibration in a working machine, solid wire will break eventually.
As a rule-of-thumb use wire size such that the current is 5A per square millimeter. Thus a 5A motor use wire of area 1mm².
For intermittent currents 10A/mm² is adequate. Note the Chinese often and usually use wires which are too small.
The 5A/mm² was an old RAF rule, or another phrase is: 'Ship Shape in Bristol Fashion', ie good.

Craig

Hi Caleb - The speaker wire I use is 0.824mm2 (18awg) so its good for 4A using the RAF guide. Solid is good for static runs not for run that flex, but I don't think you will use solid unless you have a roll in the shed that you want to use up. Keep at it. Peter

I have a friend who was a postie in Bristol and he's always in ship shape...

Hi Craig,

[QUOTE=joeavaerage;2512844]Hi,



If they're any bloody good 69.95VDC, if they're not good enough they'll pop, so buy better ones next time.


so I should be pretty safe with the 63v or even the 67v?

I'm not sure if this comment was directed towards this post or not... I don't know where the statement you are quoting here came from.

Thanks
-Caleb

Hi,
my apologies, the quote came from the clipboard of my PC, and its unrelated to this thread.

Just to clarify, if the drivers are rated at 70V, then why not use 70V?. If you intentionlally use a lesser voltage then you are tacitly saying:
'these are cheap Chinese drivers and I'm going to de-rate them because I don't believe the rating'. If you are of that opinion don't buy them!
Buy Geckos and be done with it.

Craig

Good point Craig.
I personally have a very un-bias opinion when it comes to the electronics, as I have ZERO experience with any of it. Going for the 70V. We'll see how it all comes together!

Thanks for all the help!

-Caleb