[JayneV]

Hi everyone,

It has been a long while since my last post here. Not a lot of progress with Carlos. Work had me on a crappy roster which didn’t leave much time to play with Carlos. Thankfully that has finally been sorted out and am now back on a good roster giving me time to enjoy my projects. Not having much free time turned out to be a blessing in disguise. The Y axis on Carlos was designed to be about 1500mm, much more than I need. After building the bench Carlos will live on and using it as a temporary stand for the mitre saw, it quickly became evident that this CNC machine will consume a large percentage of the available space in the garage. A couple of weeks ago I took possession of a nice lathe my brother no longer needed and after making space for the lathe I decided that Carlos needs to shrink. I have a very heavy and sturdy steel bench that is 1400x1000mm so I have resized Carlos to fit that bench. The wooden bench I build for him will be dismantled to free up some space. I was not really happy with that wooden bench anyway. The wood was second hand construction timber I had lying around and it wasn’t very good quality. Carlos’s Y axis is now roughly half the length it used to be. The X axis will remain as it was. This was the easiest option to shrink the footprint without major redesign. Thankfully the build was not too far along making the change a simple decision. Attached are before and after screen shots of Carlos being resized.

The only real issue with cutting down the Y axis is that I have already purchased the ballscrews and linear rails. I believe the rails can just be cut without creating any problems, is that right? The ballscrews are a little more complex because after cutting them the ends need to be re machined. If that proves too complex I could buy two new screws and either sell or find another use for the ones I have.

Hi Chris,
I’m glad you are enjoying the thread. Hope to see a build thread for your machine too. The wax works great with the aluminium. There is a noticeable difference when the wax starts to wear off. I periodically lightly touch the blade to the wax stick while it is running which seems to work well.

Hi Pete and Mactec,
Thank you both for the info. I’m still undecided about which way to go with the motion controller. I do like the Masso controller. I have not yet compared pricing between various options. If I go the PC based route, I will need to purchase a PC plus the controller. I know there are lots of options out there and the PC doesn’t need to be all that powerful so an older second hand machine will likely be sufficient. The Masso is more expensive but it is also an Australian company and if possible I like to support local businesses even if they are more expensive than imported options. Still plenty of time to consider my options before it’s time to make a final decision.

[mactec54]

Hi everyone,

It has been a long while since my last post here. Not a lot of progress with Carlos. Work had me on a crappy roster which didn’t leave much time to play with Carlos. Thankfully that has finally been sorted out and am now back on a good roster giving me time to enjoy my projects. Not having much free time turned out to be a blessing in disguise. The Y axis on Carlos was designed to be about 1500mm, much more than I need. After building the bench Carlos will live on and using it as a temporary stand for the mitre saw, it quickly became evident that this CNC machine will consume a large percentage of the available space in the garage. A couple of weeks ago I took possession of a nice lathe my brother no longer needed and after making space for the lathe I decided that Carlos needs to shrink. I have a very heavy and sturdy steel bench that is 1400x1000mm so I have resized Carlos to fit that bench. The wooden bench I build for him will be dismantled to free up some space. I was not really happy with that wooden bench anyway. The wood was second hand construction timber I had lying around and it wasn’t very good quality. Carlos’s Y axis is now roughly half the length it used to be. The X axis will remain as it was. This was the easiest option to shrink the footprint without major redesign. Thankfully the build was not too far along making the change a simple decision. Attached are before and after screen shots of Carlos being resized.

The only real issue with cutting down the Y axis is that I have already purchased the ballscrews and linear rails. I believe the rails can just be cut without creating any problems, is that right? The ballscrews are a little more complex because after cutting them the ends need to be re machined. If that proves too complex I could buy two new screws and either sell or find another use for the ones I have.

Hi Chris,
I’m glad you are enjoying the thread. Hope to see a build thread for your machine too. The wax works great with the aluminium. There is a noticeable difference when the wax starts to wear off. I periodically lightly touch the blade to the wax stick while it is running which seems to work well.

Hi Pete and Mactec,
Thank you both for the info. I’m still undecided about which way to go with the motion controller. I do like the Masso controller. I have not yet compared pricing between various options. If I go the PC based route, I will need to purchase a PC plus the controller. I know there are lots of options out there and the PC doesn’t need to be all that powerful so an older second hand machine will likely be sufficient. The Masso is more expensive but it is also an Australian company and if possible I like to support local businesses even if they are more expensive than imported options. Still plenty of time to consider my options before it’s time to make a final decision.

You can always just play with Mach3 for free it will run 500 lines of code just to get things running, Masso is a good package but is like you say expensive and you are stuck with what you get as it is all proprietary, service is everything though

[JayneV]

You can always just play with Mach3 for free it will run 500 lines of code just to get things running, Masso is a good package but is like you say expensive and you are stuck with what you get as it is all proprietary, service is everything though

I currently only have Mac computers. Also wouldn’t I need some hardware for Mach3 to interface with or are you talking about running the software in some kind of simulation mode without any hardware for it to control? If it is simulation mode then I could run windows through boot camp on my Mac and install Mach3 to play with.

You are right about the value of service. I have read good things about the support Masso offers plus they also have an active support forum on their website. The support and being a local Australian company (their head office is only about 30 minutes drive from my home) both have a monetary value to me which in a way partially offsets the higher price when comparing products.

[mactec54]

I currently only have Mac computers. Also wouldn’t I need some hardware for Mach3 to interface with or are you talking about running the software in some kind of simulation mode without any hardware for it to control? If it is simulation mode then I could run windows through boot camp on my Mac and install Mach3 to play with.

You are right about the value of service. I have read good things about the support Masso offers plus they also have an active support forum on their website. The support and being a local Australian company (their head office is only about 30 minutes drive from my home) both have a monetary value to me which in a way partially offsets the higher price when comparing products.

If you had or could find a good used computer with windows XP on it, this would most likely have a Parallel Port and all you would need then is a cable and a $12 Breakout Board Power supplies and you could run all 3 Axis, If you were to get a UC100 then you could use a newer computer as it only needs a USB port

The only problem I see with the Masso is there low Khz speed it would be ok for Stepper motors use, not so good for Servo motors if you compared to other new controls, they are at the low end, it looks like a good control but lacks where it is most needed, if you compare to say UCCNC they have up to 400Khz the Masso is only 110Khz which is only 10Khz better than a UC100 which you can have for under $100 and you can use their UCCNC control or Mach3 with the same Hardware, support for any of these controls today, is only an email or a post like here on the Zone

[mactec54]

Here is something you may be interested in if you have not seen this, it seems to be for Mac computers https://carbide3d.com/carbidecreate/

PlanetCNC will also run on some Mac computers
https://planet-cnc.com/

[JayneV]

Thanks for the links, I wasn’t aware of those products. I’ll have a look when I return home in a week. My better half and I are currently having a much needed holiday in New Zealand. We had this crazy idea to run the Queenstown marathon yesterday and are feeling every step of it today. Lol. It was a great day though and the scenery was absolutely beautiful.

What is the importance of the kHz speed in motion controllers and why is the lower speed ok for stepper motors but not so good for servos? I am guessing it has something to do with how fast the controller can send commands to the motor drivers? Btw, I was planning on using steppers. It just occurred to me, by using a cheaper motion controller setup, perhaps I can put some of the savings towards using servos instead. Would that be a worthwhile upgrade for my type of machine or is it overkill?

[TMToronto]

I found a few references helpful when I started setting up my hobby CNC. Here are two which may help answer some of your questions, at least with respect to steppers.
https://www.geckodrive.com/support/a...and-resolution
https://www.machinedesign.com/mechan...motion-control
I live in Canada, and purchased a smaller CNC made locally - Onefinity Woodworker. I chose the Masso G3 for my controller because I wanted to have a 5 axis option without the use of a PC. I purchased a compatible touch screen (vs buying the Masso Touch) and am very happy with it. For my machine and settings, I come no where near taxing the 110kHz limit, but I am not interested in extremely fast rapids or overly high micro-stepping.

I would have chosen the Centroid Acorn, but it is only a 4 axis controller and requires a PC, but I really like that it is a well established company and the products are highly configurable. Having said that I am very much pleased with the Masso. The controller has plenty of inputs and outputs, supports the G and M codes that I need (these are continually being added to), and has great support. The forum community is active and helpful, and the company is constantly adding new functionality, both on the hardware and software side, often in response to user requests for new features. It was an investment for sure, but one I remain happy to have made.

[mactec54]

I found a few references helpful when I started setting up my hobby CNC. Here are two which may help answer some of your questions, at least with respect to steppers.
https://www.geckodrive.com/support/a...and-resolution
https://www.machinedesign.com/mechan...motion-control
I live in Canada, and purchased a smaller CNC made locally - Onefinity Woodworker. I chose the Masso G3 for my controller because I wanted to have a 5 axis option without the use of a PC. I purchased a compatible touch screen (vs buying the Masso Touch) and am very happy with it. For my machine and settings, I come no where near taxing the 110kHz limit, but I am not interested in extremely fast rapids or overly high micro-stepping.

I would have chosen the Centroid Acorn, but it is only a 4 axis controller and requires a PC, but I really like that it is a well established company and the products are highly configurable. Having said that I am very much pleased with the Masso. The controller has plenty of inputs and outputs, supports the G and M codes that I need (these are continually being added to), and has great support. The forum community is active and helpful, and the company is constantly adding new functionality, both on the hardware and software side, often in response to user requests for new features. It was an investment for sure, but one I remain happy to have made.

The higher it is the smoother it will run your machine this is low for the price

There are lots of packages just like theirs with much more power and a 1/3 of the price they started badly with that 110Khz control, even a stepper machine has a big benefit of more power just look at the Smooth Stepper that is a good example

Why was Acorn not enough for you the 4 Axis Board can run any 4-axis machine, Acorn also has an Avanced 6 axis control

[TMToronto]

From what I gathered from various discussions over the year on the Masso forum, I think the 110kHz may have been a trade off - deciding how best to put finite processing power to best use. Perhaps in the future there will be a Masso G4 with this limitation (for some users/builds) improved. I believe the Acorn sets the default at 200kHz which I imagine they think is fine for most users. I was reading that some drives require a maximum pulse width to operate reliably, so some people might set pulse per second even lower to meet this suggestion in the drive manual.

As for the Acorn, my CNC uses two steppers for the Y axis and I use two separate drivers to take advantage of Masso's auto squaring. I also wanted the other axis as I am designing an oscillating tangential knife that I plan to build and use. Adding up the Acorn kit and extra 'pro' version software I needed for the functionality I wanted (5 WCS, ATC features,...) brought the cost to within about $350-400 of the Masso, and I would still need a computer which I did not want. The Acorn 6 axis was not out at the time I was building my up electronics and hardware, and it would have been a lot more expensive.

For my next build (when time and funding permit), which will only require 4 axes, I would like to try a computer based set-up for the higher levels of customization, but also if for no other reason than to allow a comparison with the Masso controller.

[peteeng]

Hi Jayne - Good to have you back. Since you now have a lathe you can shorten the ballscrews easily. Re servos. They are a budget jump but if you want more speed they are the way to go. Ultimately once you settle into your machine you will want more speed. Steppers will top out torque at say 600-800rpm then servos will go thru to 3000rpm so you will get a job done 3x faster. Why take 45mins when you can do it in 15mins? So comes down to budget... explore servos worthwhile if you have the $$$. Peter

[mactec54]

From what I gathered from various discussions over the year on the Masso forum, I think the 110kHz may have been a trade off - deciding how best to put finite processing power to best use. Perhaps in the future there will be a Masso G4 with this limitation (for some users/builds) improved. I believe the Acorn sets the default at 200kHz which I imagine they think is fine for most users. I was reading that some drives require a maximum pulse width to operate reliably, so some people might set pulse per second even lower to meet this suggestion in the drive manual.

As for the Acorn, my CNC uses two steppers for the Y axis and I use two separate drivers to take advantage of Masso's auto squaring. I also wanted the other axis as I am designing an oscillating tangential knife that I plan to build and use. Adding up the Acorn kit and extra 'pro' version software I needed for the functionality I wanted (5 WCS, ATC features,...) brought the cost to within about $350-400 of the Masso, and I would still need a computer which I did not want. The Acorn 6 axis was not out at the time I was building my up electronics and hardware, and it would have been a lot more expensive.

For my next build (when time and funding permit), which will only require 4 axes, I would like to try a computer based set-up for the higher levels of customization, but also if for no other reason than to allow a comparison with the Masso controller.

When you have a 4 axis Board like the Acorn, and you have 2 motors Drives per axis you use a splitter and drive both axis motors from the same port, you then have perfect timing for each Drive / Motor also, the Acorn would have way more I/0 than you need for ATC as well, and yes, the Acorn is not cheap either it is 400Khz though which is on par with most other controls

[mactec54]

Thanks for the links, I wasn’t aware of those products. I’ll have a look when I return home in a week. My better half and I are currently having a much needed holiday in New Zealand. We had this crazy idea to run the Queenstown marathon yesterday and are feeling every step of it today. Lol. It was a great day though and the scenery was absolutely beautiful.

What is the importance of the kHz speed in motion controllers and why is the lower speed ok for stepper motors but not so good for servos? I am guessing it has something to do with how fast the controller can send commands to the motor drivers? Btw, I was planning on using steppers. It just occurred to me, by using a cheaper motion controller setup, perhaps I can put some of the savings towards using servos instead. Would that be a worthwhile upgrade for my type of machine or is it overkill?

Enjoy your NZ break away

Servos are smooth and quiet, run much faster than a stepper motor and will never loose position, so if you have any thoughts at all of using servos the extra cost is worth it, to get the max speed of a Servo you need a good control that has a pulse rate from 300Khz and up, 200Khz you can scrap by for the lower RPM Servos most new controls offer 400Khz this will get you up to 5,000 RPM which is what most small Servo Motors will run up to, 110Khz will not do it for a normal Servo you won't be able to reach the max RPM of the motor even using the Electronic Gearing in the Servo Drive, this is the only reason I have not brought and tried the Masso because of the low 110Khz output I'm on their mailing list, so watch for any updates with their control

They are never over kill, unless you are not serious about your machine, Steppers are for the low-end performance if you are looking for the long-term, trouble-free machining then Servos are the best package

[TMToronto]

Early on I had considered the solution you mention - two drives sharing the step/dir signals from one controller axis - but I was not sure how I would effectively handle/correct any missed steps between the two steppers should that occur. I also thought that I would not be able to take advantage of any auto squaring features if available.

With the help of an electrical engineer in the Masso forums, I worked to design and build a Y axis splitter board. It was a great project and I learned a lot about the electronics involved, and how to make a PCB - see image. It was designed to work with the Masso and still allow the auto squaring functionality. I ended up not using it (for now), as moving from the bread board prototype to the completed soldered PCB left it with missed connections that I just didn't have the time to trace.

[JayneV]

I will have a look at servos to assess whether I can justify the extra cost. If they are too much of a price jump over steppers maybe they can be a future upgrade so I’ll keep that in mind when selecting a controller. Could you recommend any servos for my machine or point me in the right direction to begin my search? Carlos is designed to use 3Nm Nema23 steppers. The ballscrews, if I remember right, are 1610 for the X and Y axes and 1605 for the Z, or are the X and Y 2010?? Hmmm, I will have to double check when I get back home, it’s been a year since I bought them. Anyway, will the critical speed of the ballscrews limit the benefit of the higher speed servos?

[peteeng]

Hi Jayne - When your home publish the new lengths and I can check for whirling.... Ignore cnczone while your on hols... Peter

[joeavaerage]

Hi,
I use Delta 750W B2 series servos and they kick arse....better than any stepper ever made, hands down. You are correct of course, they command a premium and is
that premium justified over the cost of steppers?

The principle advantage of servos over steppers is speed and resolution. For instance my servos are rated to 3000rpm, but in fact I use them up to 5000rpm. Most steppers won't even spin
at 1000rpm without load, let alone with a load.

if I remember right, are 1610 for the X and Y axes and 1605 for the Z,

Imagine if you used servos and that you direct coupled them to the screws then at 3000 rpm the X,Y axes would move at 0.010 X 3000 =30m/min and the Z axis at 0.005 x 3000=15m/min.
That is very very fast for a hobby machine. You are also correct that it may well be that you would have to limit the speed of the servos to prevent the ballscrews from whipping, so some
of that really top speed would be lost.

I use 3205 screws in my new mill, and they, being 32mm in diameter don't whip at any reasonable speed, and so I push them to 5000rpm. The axis movement is 0.005 x 5000=25m/min.
25m/min is scary fast, and I seldom use it, in fact 10m/min to 15m/min is entirely fast enough for me.

Servos have this programmable feature called Electronic Gearing. It a rather more flexible but not dissimilar idea to microstepping in steppers. I have programmed my servos to 5000 steps per
revolution, and being direct coupled to the screws results in a linear resolution of 1um per step. This have proven to be very adequate. At 5000 rpm however the pulse rate to the servo
drive is 5000 x 5000 /60=416kHz. So you can see that having high speeds AND fine resolutions requires a very high pulse rate, much higher than the Masso for instance.

I use Mach4 and an Ethernet SmoothStepper. The SmoothStepper can run at 4Mhz....so can handle 416kHz at a canter. I have a servo driven spindle that when I drive at its rated 35000rpm with
an 8000 count/rev encoder requires an even faster pulse rate of 3500 x 8000/60=466.7kHz. Again the SmoothStepper can do this no trouble.

It is entirely possible for me to use the Electronic Gearing feature to reduce the required pulse rates to 100kHz or less, but only at the expense of resolution, and I am loathe to give
up all that resolution that I've paid so dearly for.

This is the reason that Matec has suggested that the Masso controller may be a little slow....it may limit either your speed OR resolution should you ever use servos.
In a related issue; most stepper drives and servo drives are limited to about 200kHz input with the so called 'single ended' signalling, sometimes also called 'open collector'.
To go beyond that speed requires 'differential signalling', which requires a twisted wire pair for each signal (Step/Dir) to the drive. This adds some complexity to the breakout board,
not insurmountable. I suspect Masso have elected to stick with 'single ended signalling' in order to avoid the extra complication. That compromise would be fine for steppers
of modest resolution and speed but might be found to be limiting for servos being faster and higher resolution.

Craig

[mactec54]

Early on I had considered the solution you mention - two drives sharing the step/dir signals from one controller axis - but I was not sure how I would effectively handle/correct any missed steps between the two steppers should that occur. I also thought that I would not be able to take advantage of any auto squaring features if available.

With the help of an electrical engineer in the Masso forums, I worked to design and build a Y axis splitter board. It was a great project and I learned a lot about the electronics involved, and how to make a PCB - see image. It was designed to work with the Masso and still allow the auto squaring functionality. I ended up not using it (for now), as moving from the bread board prototype to the completed soldered PCB left it with missed connections that I just didn't have the time to trace.

Looks like that was a good project, you can get this one from CNC4PC also which is very well designed, if you look at the DMM Breakout Board it already has this built in, it was the first ever Breakout Board to have this 2:1 axis conversion built into the Breakout Board

[TMToronto]

Thank you, I had not seen that before.

[JayneV]

Hi Jayne - When your home publish the new lengths and I can check for whirling.... Ignore cnczone while your on hols... Peter

Thank you, I will post the screw sizes when I return home. I check in on CNCZone mostly at night when I’m lying in bed as a distraction from the post marathon soreness until I fall asleep. Lol

[JayneV]

Hi,
I use Delta 750W B2 series servos and they kick arse....better than any stepper ever made, hands down. You are correct of course, they command a premium and is
that premium justified over the cost of steppers?

The principle advantage of servos over steppers is speed and resolution. For instance my servos are rated to 3000rpm, but in fact I use them up to 5000rpm. Most steppers won't even spin
at 1000rpm without load, let alone with a load.



Imagine if you used servos and that you direct coupled them to the screws then at 3000 rpm the X,Y axes would move at 0.010 X 3000 =30m/min and the Z axis at 0.005 x 3000=15m/min.
That is very very fast for a hobby machine. You are also correct that it may well be that you would have to limit the speed of the servos to prevent the ballscrews from whipping, so some
of that really top speed would be lost.

I use 3205 screws in my new mill, and they, being 32mm in diameter don't whip at any reasonable speed, and so I push them to 5000rpm. The axis movement is 0.005 x 5000=25m/min.
25m/min is scary fast, and I seldom use it, in fact 10m/min to 15m/min is entirely fast enough for me.

Servos have this programmable feature called Electronic Gearing. It a rather more flexible but not dissimilar idea to microstepping in steppers. I have programmed my servos to 5000 steps per
revolution, and being direct coupled to the screws results in a linear resolution of 1um per step. This have proven to be very adequate. At 5000 rpm however the pulse rate to the servo
drive is 5000 x 5000 /60=416kHz. So you can see that having high speeds AND fine resolutions requires a very high pulse rate, much higher than the Masso for instance.

I use Mach4 and an Ethernet SmoothStepper. The SmoothStepper can run at 4Mhz....so can handle 416kHz at a canter. I have a servo driven spindle that when I drive at its rated 35000rpm with
an 8000 count/rev encoder requires an even faster pulse rate of 3500 x 8000/60=466.7kHz. Again the SmoothStepper can do this no trouble.

It is entirely possible for me to use the Electronic Gearing feature to reduce the required pulse rates to 100kHz or less, but only at the expense of resolution, and I am loathe to give
up all that resolution that I've paid so dearly for.

This is the reason that Matec has suggested that the Masso controller may be a little slow....it may limit either your speed OR resolution should you ever use servos.
In a related issue; most stepper drives and servo drives are limited to about 200kHz input with the so called 'single ended' signalling, sometimes also called 'open collector'.
To go beyond that speed requires 'differential signalling', which requires a twisted wire pair for each signal (Step/Dir) to the drive. This adds some complexity to the breakout board,
not insurmountable. I suspect Masso have elected to stick with 'single ended signalling' in order to avoid the extra complication. That compromise would be fine for steppers
of modest resolution and speed but might be found to be limiting for servos being faster and higher resolution.

Craig

Thanks Craig, great reply. I’ll re read it more carefully after returning home in about a week. Servos sound like they are well worth considering even with their higher cost.