I finally ordered a benchtop milling machine that I plan to convert to a CNC mill. If you could peek inside my mind, you would be completely shocked by this decision.
In 2010, I wrote a post about how it had become easier to build a DIY CNC machine. At the time, I had researched the idea a couple of times, and there were two paths I could take.
Back then, hobbyists could undertake huge projects to convert common benchtop mills for CNC control, or they could build their own CNC routers using off-the-shelf parts.
Advertisement
Several companies, such as CNC Router Parts – now called AVID CNC and known for their bolt-together CNC routers – sold components that made DIY CNC builds seem easy.
Every time I looked into the hobbyist CNC world, I arrived at the same conclusion, that conversions were hard, routers were easy. This idea stuck with me.
Since then, my plan has been to buy a larger benchtop milling machine for manual use. After doing much research over the years, the relatively new Precision Matthews model PM-728VT benchtop mill remained my top choice.
My other plan has been to save up for an Avid benchtop Pro CNC.
I have been keeping up with turnkey developments, but haven’t found any appealing alternatives. I bought smaller tools, and while they produce some parts for me, they’re not very close to what I’m looking for.
Yes, you do a lot with belt-driven CNC routers and miniature mills that zip around with 1/8″ end mills, but there’s a limit to their functionality when talking about cutting metal.
Advertisement
Some new designs seem promising, such as the Langmuir MR-1, which is described as “an affordable CNC gantry mill for machining metal.” With the Langmuir MR-1, the bed is filled with concrete during user assembly for added rigidity. It’s indeed a unique design, but I’m not sure I’d want to be an early adopter.
Conversions still seemed messy.
While not as ideal as a full-fledged CNC capable of quickly turning aluminum billet into functional parts, a manual mill would help with at least some parts I want to make. Some expansion of what I could build myself is better than nothing, right?
It’s not like any machinery is getting more affordable.
But could I justify the space and expense when the Avid Benchtop Pro CNC router (or something like it) still remains a future goal if/when finances ever permit it? That question had me hesitant to buy anything.
When researching benchtop mills and leaning towards the PM-728VT for the umpteenth time, I saw a post on CNC conversion, and the idea started to take hold.
I went down that rabbit hole a bit, and in doing so I realized how little I knew about the control side of larger CNC machines.
In thinking about it, one of the reasons I never got too close to buying an Avid, aside from its high price, is because its controller is designed specifically for Mach 4 software. Mach 3 was hugely popular, but I’ve heard enough bad things about Mach 4 to want to avoid it.
I didn’t know the first thing about building or even selecting a different controller to use with Avid’s hardware, or any brand’s CNC framing or linear motion kits.
Over the years, I thought that you connect a driver board, such as the 4-axis Gecko G540, some stepper motors, and… run some control software? I assumed this was 90% of what I needed to know about the electronics and control aspect, and that the mechanics were the hard part.
Pictures of Avid’s control enclosure looks more complicated than that.
The idea of a manual to CNC mill conversion was never an option, and at one point over the years I gave up on the idea of building a gantry-style router myself.
There are plenty of kit and read-to-go solutions today, but nothing seemed perfectly suited.
So, I kept waiting, maybe some day I can budget the money and space for an Avid or Tormach, something new and suitable would come along, or another solution would present itself.
And then came the light bulb moment recently.
I convinced myself that converting a benchtop mill for CNC control, and designing and building a controller to operate it with, would be a good idea. At the least, it sounds fun. Plus, if I change my mind, the practically plug-and-play components are universal and could be used in future projects.
Details about the Precision Matthews PM-728VT specifically helped to fuel my decision.
If I changed my mind or it didn’t work out perfectly, most of the same controller parts could be used for a future Avid CNC build, or a CNC-converted lathe.
The PM-728VT weighs around 370 pounds, and offers a sizable increase in stoutness and rigidity compared to my existing micro-sized equipment.
On a scale from 1 to 10 in the manual machining world, this might be a 2.5 in terms of size and rigidity, only ahead of micro and mini mills. In no uncertain terms, it’s a small mill.
I saved up the budget for a bigger machine, but this one seems well-suited for my current and foreseeable needs.
Precision Matthews offers some good-to-have accessories, such as DRO options and a power feed. I intended to buy those at first, when seeing the 728VT as a manual mill upgrade, but was reminded that they offer a ballscrew upgrade kit specifically for CNC conversion.
The company advertises the ballscrew kit as being easily installed with no modifications necessary. Its installation or use also doesn’t affect the milling machine’s 5-year warranty.
Precision Matthews also offers a plug-in spindle control board, although I’ve read mixed reviews about it. That’s something I can figure out with more research.
Third party conversion kits are available for many other benchtop mills from Precision Matthews and other brands, but they tend to require modifications. That’s where mill conversions get messy and require much more attention that I ever wanted to give them.
It seems that maybe the mill was designed at least partially with CNC conversion in mind, and this largely factored into my decision.
All of a sudden, this looks like an easy CNC conversion project, at least mechanically speaking.
The spindle is designed and spec’ed for manual use, but I’m not worrying about that now.
On the controller side of things, yikes! I’m a bit in over my head, but sorted through enough to see where decisions will need to be made.
Do I go with stepper motors, or servos? What controller package? Software? Homing sensors? Do I need limit sensors?
There are now all-in-one controller options like the Masso G3, which works without needing a separate computer, more sophisticated but still affordable solutions like the Centroid Acorn control board, and traditional software such as Mach 3 (or 4) that work with different control boards. There are also Arduino-powered CNC controllers, Ethernet Smooth Stepper boards, and lots more to look into.
So far, I think I’m going with the Centroid Acorn board, which comes with free software and an affordably priced “Pro” upgrade option.
The Acorn is described as a “do-it-yourself 4 axis CNC control board for use with mills, lathes, routers, plasma, and a wide variety of machine tools and special applications.”
My goal here is yes, to convert a manual mill for CNC, but also learn all of the little things that I need to know if I ever want to upgrade to something different or bigger.
I have the ominous feeling, however, that this is not going to be an inexpensive venture. The bright side is that, when all is said and done, I’ll either have a CNC-controlled benchtop mill, or a manual mill and bunch of CNC components I can still use towards other automated project needs.
This is going to be quite the adventure.
Harrison
Some serious content on the way!
Looking forwards to hearing about it.
Rafe
Mach 4 is perfectly fine FYI. If you’re doing it right, you’re just loading and executing code with it. It’ll often lose axis points when turning the machine on and off, but if you’ve got an extra 30 seconds to spare it’s not a big deal. Get a good code production and design export software and you’ll be fine should you ever decide to get an AVID machine.
MM
Most important thing, by far, on any machine tool be it CNC or manual is weight and rigidity. When you first get the machine go through it and make sure the gibs are tight, there’s no backlash in any of the leadscrew nuts or their thrust bearings, etc.
Mills need all the weight they can get, that is why one of the solutions you saw called for being filled with concrete. Proper industrial ones are set in special deep foundations and are secured to them with cement. As far as small mills go, often times it makes more sense to just get a floorstanding mill, since heavier mill beats a lighter mill plus a heavy table. As I once read in a book about building race cars: if you have to add weight to the car, do you want that as useless deadweight ballast? Or do want that as strength in your chassis? Anyway, I digress: heavy table is a necessity. It would not be inappropriate for your table to weigh a thousand pounds. Or double that. But do whatever is practical for you and make sure the machine is securely bolted to it. Best would be a heavy fabricated steel table filled with the hollow space filled with shot ballast or sand, and use thinset mortar between the machine’s base and the table top before you tighten the bolts down. The heavier you can get it the less vibration, chatter, and noise. The machine will be more pleasant to work around and you will have better tool life and superior surface finish.
Stuart
Yes.
My goal is for a sturdy enclosure, but perhaps not 1000 pounds just yet. I like the idea of the 728VT for its smaller size and mobility if needed.
I’ve also been reading about the option for “epoxy granite” mixes where gravel and sand is used with epoxy to fill gaps in cast base sections.
MM
I built a lot of tables for vibration-sensitive lab equipment when I worked in a university lab before I had my machining business, it is remarkably cost-effective compared to purchasing them. Weld up a frame using square steel tubing. The thicker the wall the better of course but even something like 1/8″ is very good. Use diagonal bracing on the legs, and cut holes where the braces weld to the main tubes so the fill can enter the braces as well.
I prefer to use a loose fill rather than concrete or epoxy materials. Leaving the fill loose allows it to absorb vibration. And the fill can be removed to lighten the table to make it easier to move in the future. Sand or gravel is cheap and easy. Better but more costly is steel ballast pellets or steel bead-blasting media. And further up the cost scale is fine lead shot, but that’s very costly. A cheat trick I’ve used several times is to put as much scrap steel as possible inside the frame as I’m welding it together, and then at the end pour in the sand. That’s a good combination of cheap, dense, and vibration-absorbing. It’s easy to get the sand to fill all the nooks and crannies by tapping the stand with a hammer as the sand goes in.
fred
PS – don’t use beach sand for ballast as one of my neighbors did for some outdoor structures. A few years later he learned that the sand was probably not perfectly dry and had enough sea salt content in it to promote corrosion, Had he been spied removing the sand from the beach – he also might have been subject to a fine.
MM
Yeah, you don’t want any moisture in there. Not only would it promote corrosion but moist sand doesn’t flow well so it will be hard to fill. I always used sacks of play sand from the home improvement store and never had any issues with that.
Nathan
I was going to ask – how does the benchtop mill connect and sturdy it self to the bench. and I assume here the most important thing is to have it level and square to the bench. IE if you throw a level on it and a plumb against the mill. then the mill collet/head is dead vertical and 90 to the floor and the base plate is dead level to the floor.
And I guess it bolts to the bench – would you bolt the bench to the floor or maybe wall?
Meanwhile read up a bit on that acorn thing – I like the initial premise. I think it would be wise to have limit switches though as a precaution – if the software fully supports that. IE don’t trust the encoder to know well I can go ___ inch from 0 to the left and not run over….. but have a finite limit switch to ensure – oh no stop encoder says there’s a another 0.54 inch but stop is met. If I read it right it does allow for that.
what software will to draft with to send to the CNC software?
seems interesting.
Stuart
The stand or future enclosure base should be level.
The Acorn has 8 inputs. More can be added, but at considerable expense ($299).
With limit switches, one input can be for all home, and one for all limit.
With proximity sensors, one input can each be for x, y, and z.
A rotary table might also require an input.
E stop requires an input.
If I go with servos, one input could be a fault sensor.
A probe and height sensor would be 2 more sensors.
With servos, the ones I’m looking at have their own hard limit travel settings.
With Acorn, proximity sensors can be used for x, y, z home, and then software for travel limits.
So that could potentially be 3 home inputs, 1 drive OK, 3 for probes and height probe, 1 for E-stop.
Physical over travel limits could potentially be tied into the E-stop just in case, although it’s less than ideal.
The Acorn seems to prefer all physical switches or just home proximity sensors. They have schematics examples for both, but not one where there’s 6 proximity sensors.
It’d be easier with normally open sensors, which is what Avid requires for their machines, but that’s a bad practice compared to normally closed sensors.
If you have a NO limit sensor and there’s a failure, it cause lead to an over travel crash, whereas a failed NC switch or sensor, or broken wire, would cause a limit trigger.
Nathan
I could see where a servo with built in over travel limit would be a good compromise
MKY
Stuart –
The Acorn is described ***a*** “do-it-yourself…
Chris
Did you ever look into openbuilds?
How do they compare with your findings?
Stuart
The OpenBuild is a very different class of machine. I haven’t looked into it too closely because it overlaps what my existing small machines can do.
Scott
I did a lot of research a few years ago and ended up going with the Centroid Acorn. I’ve been very happy with it! They’ve now come out with a 6-axis controller and I’m considering upgrading to it.
Robert
Sorry, but reading all the above, this seems like and opened money pit. That you will find yourself not satisfied with the rigidity and controls of your set up and continually want to upgrade. But it doesn’t seem like there are graceful upgrade paths. For example, if a 1000 lb table isn’t enough, I’m guessing you can’t really make it heavier, you have to start again with a bigger stouter table and pour concrete in it. There seems a reason the industrial ones are so expensive.
Stuart
Yep. I’m aware of that going into this. Will see.
Like I said, at the end I’ll either have a converted CNC mill, or a manual mill and parts I can still use in other potential builds.
It’s still a necessary step. Going to a more capable machine would require a sizable increase in budget, space utilization, and electrical upgrades. I cannot justify that when this might be enough for my needs.
If I need more, this is a necessary step. I like to think I have reasonable expectations.
Speed and throughout is an important factor in industrial machines. I’m also not going to be working with tougher materials yet.
There are some cheats for increasing rigidity , but you’re right that there are no graceful upgrades.
I’m not ready for a pricier machine, although I know this is lying to myself as the cost of the control system and tooling will add to everything substantially.
I’ve considered what can be both gained and compromised by going with a half step forward here, and it’s better than just standing still and wishing I had an Avid, Tormach, or substantially more robust manual mill.
Greg
Can’t wait!
Dave
Early adopter of the Langmuir MR-1 here and liking it so far. You can really chew up aluminum and steel with great accuracy for a prosumer mill. Biggest drawback is no automatic tool changer, but there is a tool setter and probe. Cutting envelope is quite large. Customer support has been great to us early guys.
Ron
Can’t wait to see how it goes. I’ve thought about the same thing but had all the same reservations. I remember Hoss converting his G0704 and all I kept thinking is that’s just way more work than I wanna do.
Lyle
Sounds interesting. I’m looking to buy a CNC. I’m not going to be cutting much metal I think, maybe some aluminum. I’d mostly want this to help me make cabinets for a kitchen remodel. I don’t have room for a nice big table saw with proper in/out feed tables and good fence.
I’m thinking of something like the Stepcraft M1000. I’d love to hear some feedback from people.
Eric
If you don’t mind sharing, how much do you plan on spending for the parts you see as your final unit?
I’m sure you’re already familiar, but the Clough42 an Blondihacks YouTube channels are both terrific sources of info for home machine shop folks.
Stuart
I haven’t determined that yet. I originally factored in the cost of a stepper-based system, but am leaning more towards servos, which will increase the cost.
The cost of the mill by itself is considerably higher than it was when I priced it out 2 years ago in an email chat with a reader. The cost of controller package components is considerably higher for servos or steppers than when I priced it out originally. Because of that, I’ve been very afraid to add everything to a single spreadsheet yet.
Treating this as a learning – and coverage – opportunity is allowing me to justify spending more than I anticipated paying based on the figure I came up with the last time I ran the numbers maybe 2 years ago.
If I put a full estimate on paper, ToolGuyd’s CFO will likely force me to rethink things.
I only recently learned about Blondihacks when looking into lathe options, and Clough42 for CNC content. Thanks – they’re both great recommendations!
Eric J
Gotcha. Thanks for the reply.
Good luck with the project.
William Adams
(ob. discl., I work for Carbide 3D)
If I did more metalwork, I probably would have made a push to get a Shapeoko HDM (though it would have had to be tipped on edge and lowered down my basement stairs on a hand-truck).
Curious what size steel stock (and what alloy) and what sort of parts you are planning on.
Stuart
Your team has done very interesting things with the HDM, but I ruled it out for my current needs.
Steel – likely only small parts (modifying hex stock for shafts, slots for keyways, clamping blocks, other smallish things). Aluminum – robot wheel hubs, mounting brackets, mechanical frame pieces, and so forth.
At one point I’d like a machine that can handle larger 2D aluminum parts out of say 1/2″ 6061. Right now, I need a volume of 4″ to 14″ x 6″ x 2″. It’ll be nice if I can work the ends of longer pieces.
I wanted to be able to do hybrid manual cuts, such as simple edge rounding with a rotary table, or facing blocks for easier referencing on other machines without having to fire up design software.
A manual mill provides different workholding and tooling capabilities than any router-based CNC I’ve seen.
Maybe an HDM is in my future in lieu of an Avid. But for now, I wanted a machine I could potentially move from the garage to the basement for supplementary use if I eventually upgrade it.
John Banister
Golmatic makes a sturdy bench top mill that can handle thicker materials. They’re available both as manual and cnc, so presumably they’re susceptible to conversion. They are somewhat pricey, though.
Daniel V
Dr Dflo on YouTube did a series on upgrading a PM 833T to CNC. Don’t know if you’re interested, but it might be worth checking out to see what all he went with. It looks like the upfront cost of a 833t is about 1K more than the 728VT.
https://m.youtube.com/playlist?list=PLZ9ac-f1U5V-DexnK048D8IMYERToVxFI
Stuart
Thanks! I did watch some of his videos, and debated about the 833T/833VT, but it requires more space and power all around. In one of his videos I believe he mentions a 1300W servo or stepper motor for the z-axis. I’ve been reading that a <300W servo motor gave others good results with the 728VT.
Nathan
I was going to ask. What about buying a used manual full size mill and converting it? I ask mostly because I dream.
MM
In my opinion that is a wise move. Used older machines can offer tremendous value and are often much better built than what is on the market today. However, there is always a risk with buying used machinery, same as a car maybe it’s a lemon or there’s some expensive hidden problem.
I bought many used machines for my business over the years. I never had any significant issues, just little minor things like belts needing replacement or a dodgy electrical switch. However, I also went out of my way to look for relatively high-end used machines.
One potential disadvantage of older machines is that many of them were designed at a time before carbide tooling and therefore their spindle speeds may be fairly low compared to a more modern machine. However, this is usually easily addressed by replacing the motor with a modern one and a VFD control, something you’d probably be doing anyway as part of a CNC conversion.
Stuart
That’s a far bigger investment, and more machine than I can handle right now.
It’s a good idea for some, but not right for me.
There’s also more hassle that I’m not up for right now. The “oops, this was a mistake” scenario would be a bigger hassle remedy.
TonyT
Check on the mechanical side for things like backlash on the ways, because those kinds of problems will be difficult to fix. (At work we have a Bridgeport 2.5 axis semi-CNC with significant backlash issues, which our machinist can work around, but it’s a major problem).
I know many people do have Brideport/Bridgeport-style mills in their garage, but do note you’ll need 3 phase 208V power.
Addicted2Red
Mach 4 is fine now, especially if your control board/breakout board is properly wired. ESS smoothstepper with a good BOB makes a big difference.
UCCNC is a bit better and cheaper. It doesn’t have the GUI customization that Mach4 has. I like CNC4PC’s boards. Especially the one with the daughterboard that uses Cat5 cables for the connections. Makes everything so much easier and you get a ton more input/outputs than Acorn for less money.
I personally never liked Acorns hardware.
Nkazier
When I did mine, I emailed over, technician was great. I only have one power supply, and nema 34 equivalent servos, I’ll try to find the part number tomorrow and post it. Bolt on replacement for the HASS conversion I did, bought the mill, made my own brackets, then converted the steppers to servos. They also have a really simple setup and how you tune the motors is even easier. I might have made a video on my YT channel, I can’t remember, been a few years since I last made one. Check it out, kazierkreations
Sanjay Singh
How to work with you sir is there any channel to contact u. As i am already in CNC MACHINE manufacturing industry and i want to learn..
Stuart
I will not be providing instructional content, but I do plan to share any progress of my efforts.
JoeM
I’m very jealous of you now, Stuart!
Of many dreams I’ve had, an autonomous workshop is one of them. Where I just design some stuff, and send the various parts to the appropriate CNC or 3D Print setup it required. Waiting a few hours, I end up with my creation, and can start testing and rapidly prototyping.
Very Jealous Indeed! (In a positive way though.)
Stuart
If you want a CNC machine, you can still put together something small and slow. I saved up for a very long time for this, and have a lot of work ahead of me.
Even once I’m up to speed with the machine, and I build the control panel, and everything is working hopefully perfectly, I then have a whole lot of learning to do. I can design for 2D, but my 3D skills got rusty.
Cnckeith
Looking forward to more articles about your new project.
Pm727m CNC conversion video
https://youtu.be/pN1iDhBmo2I
Marty
Centroid CNC Acorn is a great choice. Hardware and software from the same manufacturer has specialized in CNC machine motion control for well over 30 years. Centroid CNC Users support forum and search YouTube for Centroid Acorn and you’ll get many hits. I have posted many videos on Centroid Acorn. You can search martyscncgarage to see them.
Stuart
Thanks!!
That’s part of what convinced me about this path; I’ve seen some of your videos before, and others’, and found them to be informative and reassuring.
I’ve been looking through Centroid’s Acorn documents, and they seem to provide a bit of hand-holding and guard rails where needed.
Marty
I helped write the Acorn installation manual. I tried to approach it in a systematic way to guide the new user along. The Centroid Users Forum has become a great resource. There are power users and Centroid staff monitoring things. No question is a dumb one. Most everyone wants each other to succeed. Best of luck in your journey no matter the route you choose. I hope you find it enjoyable and you learn much from it. I’m sure there will be times you get frustrated. When you do, don’t be afraid to reach out for help.
TonyT
Some comments:
— I’d highly recommend going with servo motors, especially if you want higher speeds. Servos have extra torque when you need it (peak torque) and a pretty flat torque curve, while steppers quickly loose torque as speed increases.
— I always use optical limit sensors for the positive and negative limits of each axis.
— Make sure all your components are going to fit together. For example, make sure the output of your CNC controller matches the input of the stepper or servo drives. Most likely you’ll be using step/direction.
There are a number of places specializing in desktop/hobby CNC components. On a more professional but still somewhat affordable level for servo motors, don’t forget places like Automation Direct, Teknic, and Anaheim Automation. Looking through my bookmarks, a couple companies worth considering (but I have no experience with) are Masso (masso.com.au) and CS Lab (en.cs-lab.eu, a Polish company, looks like they resell Delta, which is one of the more affordable professional brands)
Marty
Great advice. Choosing motors and drives is important. Things that work with each other and are properly sized for the application.
Your budget will dictate things. Fortunately for your build, I think you have many possible options. I would try and stay away from open loop steppers. Hybrid closed loop steppers at the very least. That is, motors with encoders that provide feedback to the drive ensuring the motor is where it was told to be. If not a fault signal is sent to the control to stop
Nathan
Look at servos for your drive train, clearpath is what I eventually got too due to the Stepper motors not being as good as I hoped. It’s a huge cost up front, but well worth it at the end. If I did the conversion again, I would have bought a tormac to be completely honest.
Stuart
Thanks – that’s what I’m considering.
What I like about Clearpath is that i) I found specific motor recommendations for this machine online already, where a tech responded to questions in a public forum, and ii) their documentation makes it seem fairly straightforward.
My only question is about whether I’ll need one power supply or two, based on motor wattage.
I haven’t heard the greatest things about Tormach machines over the years. Even if that wasn’t cause for hesitation, they’re well above my budget.
Marty
If you are going with Teknic Clear path, then call them. Part of their pricing is the support you get. Don’t be afraid to call them, provide details about your build, and ask them specific questions. They are very good to deal with.
TonyT
Again, I highly recommend using servo motors over steppers. Almost every time I’ve used steppers I’ve regretted it (only exception was a small stage where steppers where the only choice).
To do power supply sizing, it’s not simply a matter of adding up all the max continuous or peak currents, and then getting a power supply to match. Instead, consider, how many motors will be operating at a time. (Yeah, I know servos will always servo, but a servo holding position will be using a lot less power than a fast moving servo).
Also remember that normally you don’t have to use a power supply that supports the motor’s rated voltage, but a lower voltage will limit the maximum speed (and available torque at higher speeds). I’ve run some 300V rated motors at 48V because the velocity at 48V was fast enough.
Teknic does sell interesting power supplies, but you don’t have to use them. At work, I’ve mostly been using Delta 48V DIN rail power supplies with 50% overcurrent for ~3 seconds. Mean Well makes similar power supplies (SDR series IIRC).
You should also look at total cost for motor + drives + power supply; it might be cheaper to use motors with AC drives (probably step/dir input), no power supply required. At a glance, looks like ~$700 for 400W servo motor with AC powered drive, so probably not cheaper than ClearPath.
Stuart
The ClearPath servos can operate at 24V-75V, with 90V max.
The Teknic power supply operates at 75V. Meanwell’s models deliver 48V.
The Meanwell SDR-480-48 or SDR-480P-48 (with parallel function) might suffice. It’s $166 to $175.
Teknic’s IPC-5 is $248. It can cool passively up to 350W and with a fan up to 500W.
Will the added speed or torque of 75V for the IPC-5 vs 48V for the Meanwell make a difference? I don’t know. But, the price difference isn’t large enough compared to how much the entire control panel will cost to build.
I was also looking at DMM, but Clearpath seems a little more DIY/beginner friendly with respect to selection and documentation.
TonyT
(Note: WP is being annoying, so I’m trying to split this comment)
The simplest way is to simply use the suggested Teknic servos.
To complicate matters, note that servos are typically available in a variety of windings, depending on the desired torque vs speed tradeoffs for a given motor voltage. For a given winding, speed is proportional to voltage and torque is proportional to current. For a given size power supply (e.g. 480W, 10A at 48V or 6.4A at 75V), the optimal voltage will depend on your required speed and torque, and matching with the motor’s winding.
Teknic makes Clearpath servo motors with windings designed for different input voltages (note that the integrated drive can always handle 24V-90V). To get an idea of the differences, look at the very nice pages for the https://teknic.com/model-info/CPM-SDSK-2346P-RLN/?model_voltage=48VDC (designed for 48VDC) and the https://teknic.com/model-info/CPM-SDHP-2331S-ELN/?model_voltage=75VDC (designed for 75VDC), and select different input voltages to see how the torque curve changes.
TonyT
Note that the Clearpaths I looked at have a pretty high max to continuous torque ratio (e.g. 620 oz-in peak, 124 oz-in continuous). A ratio of 3:1 is most common, and I’ve seen up to 10:1. Typically, time above max continuous power is limited to a few seconds, so make sure the motor’s continuous torque is adequate, don’t just look at peak torque.
Also remember that you need to keep the voltage below the absolute maximum (including any voltage coming from regenerative energy when the motor decelerates) or you’ll kill the driver,. Note that DMM recommends 60V input for their 75V rated drives. Also note that the IPC-5 and DMM’s power supply allow normal amounts of regenerative energy to go back into the power supply (a plus versus a typical DIN rail power supply).
It’s very hard to find a high volume switching power supply with an output voltage over 48V – the IPC-5 and a Taiwanese power supply resold by several vendors are the only ones I’ve found (and I’ve looked a lot).
Final note: both DMM and Teknic have been around a long time, I can remember noting both companies over two decades ago, although I haven’t used products from either. I’d say that Teknic is much better known in industrial automation.
Stuart
Thanks!
I’ve read the same, about the regenerative power dump.
As for torque, I’m going by what a Teknic rep said in an online forum specifically about this model mill. I’m assuming that if I take measurements and consult with a tech, the current recommendations will be the same.
And if I’m wrong, Teknic looks to have a good “oops these motors don’t work very well for me” return and exchange policy.
Leonard
I think I commented before about this previously.
I bought a bench top CNC from a local to me manufacturer, CNC Masters. It’s weighs about 800lbs. Too heavy for my bench so I opted for their stand as well. It was about $7K
I thought about doing a conversion but I knew that was just going to be a lot of time I didn’t want to spend. I also didn’t want the CNC mill to be the hobby. I wanted to use for my hobbies. It’s the same reasoning I used when getting my 3D printers.
I didn’t buy a Tormach initially because of shipping. I’m really happy with what I got. The customer service from CNC Masters has been top notch.
I posted about my setup on Garage Journal. If your interested, have a look.
https://www.garagejournal.com/forum/threads/uncle-lennys-house-of-toys-and-tools.103998/post-9512679
Stuart
The CNC Masters Baron looks a lot like the Precision Matthews PM-932M .
I’ve heard from a couple of different companies over the years who import mills, add conversion kits and controllers, and sell them as CNC solutions.
There’s definitely benefit in buying a turn-key CNC-converted mill from a company that knows their machines inside and out and already ironed out all of the problems.
But there’s also benefit in going the DIY conversion route. In this case, with Precision Matthews offering a bolt-on ballscrew upgrade kit, there’s still some learning involved, but far less than used to be true.
Leonard
I definitely thought about it doing the CNC conversion to mill. I would have learned a lot. I still like the idea of building a CNC machine and may make a router version. But I’m going to have to figure out to fit one in my current shop.
Part of my decision was because I would do one off designs or thought they were going to be a one and done. Then a client would say, “This is great. Can you make 10 more?”
The CNC Masters machine comes with its own operating software. So I didn’t have to buy that.
I did have to buy CAM programming software. I bought VisualCAM. While I can program basic g-code, I really wanted to get to making parts.
Owning a CNC mill has been quite a journey. From understanding the machine, learning feeds and speeds, designing workholding, and figuring out the order of operations.
Paul C
Brushless servo motors have all but eclipsed everything in their category. See the following:
https://www.motioncontroltips.com/what-is-a-stepper-motors-velocity-limit/
The problem with steppers is that there is a STEEP dropoff in the middle of the speed-torque curve. In fact the ideal motor would have more of a “circle” shape to it where it peaks at the peak power point, dropping off gradually under either maximum torque or speed. But instead stepper motors have the opposite characteristic so except at the extremes (which you really don’t need) you are dealing with the worst part of the speed-torque curve.
The problem with servo motors in the past is the need for cleaning and brush maintenance, and cost. Brushes have all but disappeared and costs have come down. Plus a stepper motor requires a complicated pulsed output which means transistors, and their associated cooling issues such as the driver circuits that have to generate very sharp, crisp 120-140 V firing pulses, and the various filters and controls. A brushless servo motor drive is really just a DC drive. Most consist of just 2 diodes and 2 SCR’s (for output adjustment) for a single phase version. In fact it is actually possible to build an entirely analog brushless servo drive. You can test them with just a car battery. You just can’t get much more simple than that compared to the drivers for stepper motors.
Franco
I really know next to nothing about CNC mills.
Just before I moved to my current home, I purchased this in early 2018..
https://photos.google.com/photo/AF1QipPujTfbMWLeADYpgUKDid6fJ8KtDC3iFHr9PSM
Never had a chance to set it up at my old place. Since moving to my current home, about a month before all the COVID lockdowns, I have not had an opportunity to set it up, and probably won’t be able to touch it for 3-5 years as I am overloaded with renovations.
I purchase 2nd hand, pretty much as you see it in the photo, it was $3000 CDN. It is a PM-932M (Precision Matthews, which I believe your is also), which I never purchased with the intent to use as a CNC. But with the timeline of when I get to start setting it up and learning how to use it, and your ongoing CNC conversion, maybe I might try the same thing.
Good luck with the conversion and I am curious to see how it turns out.