-Socials-

Subscribe: https://bit.ly/Man-Without-Shed

Discord:   / discord  

Summary

In this video, I share my experience with tuning servos, particularly JMC and older DELTA ASD-M models, which proved to be less straightforward than anticipated.

However, during my DIY CNC router build, the MightyMill project, I encountered a Delta B3 servo for the Z axis. Initially apprehensive due to past experiences, I was pleasantly surprised to find that newer servo models like the Delta B3 and A3 offer a streamlined, one-button solution for auto-tuning.

Join me in this demonstration as I showcase just how effortless and efficient this process has become with modern servo technology. Whether you’re a seasoned CNC enthusiast or a beginner, mastering servo tuning has never been easier.

Don’t miss out on optimizing your CNC machine’s performance – watch now and discover the simplicity of auto-tuning with Delta B3 and A3 servos!

My build has finally begun and in this post I will take you on the building journey. Make sure you have read part 1 here , part 2 here , part 3 here

Lifting it up in the shed

It’s been some time since my last update. The machine is complete, but it’s worth noting that I had to construct a shed specifically to accommodate it; the machine proved too large for the existing brick structure.

I bought a garage “engine block” crane to lift the approximately 350kg structure in the shed. You really notice that using the right tools in a situation likes this makes huge amount of difference.

Replication method at work

After moving and leveling the machine I had to make the surface of the aluminium extrusion flat in order to mount the rails to it.

It is a bit of a messy situation but hey! it works.

After making the surface flat I cut out the slot with a hot-knife in order to mount my rails to it.

Adjusting the gantry level

I used the metal epoxy “gap filling” method to level the gantry vs the Y1 and Y2 gantry rails:

I used Mirror glaze wax release on the gantry to make sure it just fills the gap and does not stick to the gantry extrusion itself. Not this works REALLY well!

After making sure it is level, I let it cure.

in a Dan Gilbart video I recently learned to make figure 8 motions when doing this. I tried it out and noticed it worked well. I did not actually check it on the surface plate but you notice that the surface is quite flat. Within >=0.0x mm I would estimate. Good enough for this part.

Mirror Glaze

Mirror Glaze Wax release; 16$

Aliexpress.com

Fixing the electronics

I decided to rewire the electronics from scratch. It worked out quite well I must say!

After I wired most of it on the desk I hung it on the wall and started connecting it to the CNC.

Finishing the mechanics

I kinda dropped and messed up the C5 1605 nuts I had on my spindles. As they were already too short anyway I decided to pick up some new ones from Aliexpress:

I was seriously impressed with how riged the DFU version of the ballscrew feels. There is no play at all. I later on measured it on the CNC itself with load and you dont see it budge at all. I can really recommend these C5 1605 DFU screws from Yixindianqi Store on Aliexpress.

So after completing most of the electronics, I went off to mount the spindle and servo motors. Initially, I had put the MAKITA RT0700 router on there; but that died very quickly after a little accident when I crashed it into the toolsetter.

Because the design of my frame is more of an afterthought … I had to add some way to mount the fixture table/wasteboard on the CNC. While it is not ideal to weld on a relatively precise machine frame, it was still the easiest method to do it.

After completing this stage I mounted 36mm of MDF plate to it and went on to the next step

Many small steps in between

So as it goes in a project. you lose focus on documenting a bit and do all kind of small steps. During this time I got probe basic working well on the CNC, tuned the servos and did some final adjustments on the mechanics.

Tramming and aliging the axis

So as in any precision machine build where your mechanics are to perfect, you are going to have to adjust the tram and yaw of the spindle.

I started first with fixing the “nod” of the spindle by putting in a 10mm endmill upside down and using an indicator to make sure the measured distance stayed the same while moving the Z-axis up and down. This was relatively quick to be fixed.

For the trimming (rotation of Z spindle ) I did the following:

So I 3d printed a spacer so I could mount my indicator on the spindle. Turned it round to make sure the distance of the indicated stayed the same. As my z-axis is quite heavy, I found the use of a strap quite useful to slowly get it in the right spot.

Making chips!

After spending some more time with the machine I made another short video where I did some more drilling and milling

Can it mill hardened steel?

So a question that comes up quite a lot is if the Lusintun BT30 ATC spindle can mill steel? The answer is yes, the spindle itself is not the limit. The stiffness of the machine is.

I did some testing on some hardened steel to make sure it could handle normal steel with ease.

Conclusion – Mightymill build Part #4

Since part #3 quite a lot has happened. The machine is together, is working and performing pretty well. As this is the first full CNC build that I have done there are plenty of areas that need improvement. One of the main things is to make the Z-Axis stiffer and make a better connection from the gantry to the ballscrew.

Other than that I’m happy with how it performs!

Make sure to subscribe to the newsletter to make sure you don’t miss any updates.

Feel free to reach out to me on discord

The replication method is a widely used technique in manufacturing and metrology to obtain accurate copies of a surface. The method involves replicating the surface of a reference material onto a workpiece using a medium such as metalized epoxy. This article provides an overview of the replication method, including its advantages, limitations, and applications.

Replication Method:

The replication method involves using a medium such as metalized epoxy to make a copy of the surface of a reference material, such as a granite surface plate to a workpiece. The surface to be replicated is first cleaned and coated with a thin layer of release agent, typically some kind of wax. The release agent is allowed to dry and is buffed to a shiny surface before the epoxy is applied. The epoxy is applied to the setup by pouring or injecting it into the setup and then allowed to cure. Once the epoxy is cured, it is peeled (or hammered) of the reference material, leaving behind an exact replica of the surface.

Advantages of the Replication Method:

One of the main advantages of the replication method is its ability to produce accurate copies of a surface. The method is also relatively easy to perform and requires only basic equipment and materials. Another advantage of the replication method is that it can be used to replicate surfaces of different shapes and sizes, making it suitable for a wide range of applications.

Limitations of the Replication Method:

While the replication method is generally accurate, it is important to note that it is subject to certain limitations. One limitation is that the replication method can only produce a replica of the surface that is in contact with the reference material. If the reference material has any defects or irregularities, these will be replicated onto the workpiece.

Applications of Replication Method:

The replication method is widely used in metrology to produce accurate copies of surfaces for measurement purposes. It is also used in manufacturing to replicate surfaces for quality control and inspection. Additionally, the replication method can be used in the restoration of damaged surfaces, such as those found on antique furniture or architectural features.

Resources:

Academia:

There is a book from the 80’s or 90’s by a professor of a US university who writes about the surfaces replication method, and how they used it to make a giant 5 axis grinder… I just cant find it at this moment…

Ah, here it is!

Principles of Rapid Machine Design – MIT – Eberhard Bamberg:

https://my.mech.utah.edu/~bamberg/research/PrinciplesOfRapidMachineDesign/Principles%20of%20Rapid%20Machine%20Design.pdf

At-scale use of the replication method for big cnc machines – Rampf group:

https://www.rampf-group.com/en/aktuelles/blog/2021/how-to-2-replicating-mineral-casting-components-with-maximum-accuracy/

Replication techniques and portable machining in machine tool rebuilding – Molglice:

http://www.moglice.com/articles/replication_techniques/wrotethebook.html

PRECISION the easy way, making CNC parts too big for a hobby mill – @floweringelbow:

DIY experiment:

With most DIY builds the most difficult part is creating flat mounting surfaces for your carriages and rails. In my Mightymill build, I have the same challenges.

Creating precision surface copies using metalized epoxy

Total Time: 9 hours

Prepare your workpiece

Make sure it is free of oil, and has a decent amount of surface roughness for the epoxy to adhere to.

Prepare the reference surface

I used Miracle wax no.8 as a release agent. Follow the manufacturer’s instructions (I did 3 coats) with buffing in between. I used a 40$ <2um error small granite surface plate.

Prepare the metalized epoxy

I used West epoxy 105 + 206 hardner, although I’d think 209 might even work better. I did a 50/50 mix of epoxy + iron powder (<63um particle size).

Tape off your part

Note: I used a 3-4mm thick parallel to regulate the thickness of my epoxy layer.

Pour the epoxy

Make sure you remove all air and have proper coverage. This is actually the most difficult part

Let the epoxy cure

I think I let mine set for about 12 hours

Remove from the reference surface and inspect

Any overflowing epoxy you can just hit off with a hammer or grind it away with a die grinder. The initial surface is quite rough due to the remaining release agent on there and any defects of your reference material.

Optimize the surface roughness and flatness

You can use either something like a 123 block + abrasive, a “kitchen” flattening stone, or any other relatively flat material. I used some dishwashing soap + the above-mentioned iron powder to basically do a lapping operation to the surface.

After about 3 – 5 minutes of lapping the surface should be smooth with most of the high sports removed.

Inspect and repeat

There are various ways to inspect surface flatness. I used a Mitutoyo indicator, a 123 block, and a high-precision HGW20CC block as my reference.

The initial flatness before the optimization stage was about 8um over the whole surface. After a cycle of 1-3 times of optimizing and inspection, it is <= 2um. Which equals my reference material. Lapping becomes challenging because the materials just want to stick together. The main issue at this point is that this is extremely difficult to measure.

As you are at or nearing the requirement of lab-based CMMS or optical measurement methods.

You could make this perfectly flat by doing this to 3 surfaces and using the Withworth 3 plate method.

Does it pull a vacuum?

yes: https://youtube.com/shorts/egtODwy4WiE

Estimated Cost: 5 USD

Supply:

• Slow cure no shrink epoxy such as West 105 + 206
• Iron powder <63um particle size
• Tape

Tools:

• 123 block

Materials: Workpiece Reference material Mold release wax

Conclusion:

The replication method is a simple yet effective technique for producing accurate copies of surfaces. While it has certain limitations, it is a valuable tool in metrology, manufacturing, and restoration. It allows a hobbyist or a professional to make micron-to-sub-micron accurate copies of reference surfaces. And thereby able to create flat surfaces for DIY CNC machines or metrology equipment at a relatively low cost and low effort.

My build has finally begun and in this post I will take you on the building journey. Make sure you have read part 1 here and part 2 here .

Assembling the machine frame

It took a while, but finally I was able to move the machine frame in my recently cleaned shed and get it assembled. For me, the concrete filled machine frame was one of the most time consuming aspects of this build.

The leg assembly was the most difficult to move. While I had put GD-60S wheels under it, it is till a top heavy assembly which weighs 120kg. With a bit of sweat and hurt, I was able to position it and get the assembly going.

I was able to assemble it without too much hassle. To get it all square I had to use the Farmer’s jack to pull the legs inward as there was a slight misalignment. After that it is square (diagonal measurement) within 1-2 mm over a span of 1.6 meters. Which is fine for this application.

It all seems to fit pretty well. Time to go on to the next step.

Lets “rise” up!

It was time to make the riser plates. I used my old and not so trusted HBM BF25 manual mill for this. It actually had lots of troubles during this process. Wobbliness … (DC) motor brushed dying, and other fun electronics issues… But i got it going again and was able to finish the job.

As you can see in the image, the side finish of my milling does not look very good. As I mentioned before my HBM bf25 is quite wobbly and not in a good state. There is still quite some play in the column and in the gibs even after tuning.

As the side surface is not a critical one for my application it is not too bad.

in a Dan Gilbart video I recently learned to make figure 8 motions when doing this. I tried it out and noticed it worked well. I did not actually check it on the surface plate but you notice that the surface is quite flat. Within >=0.0x mm I would estimate. Good enough for this part.

Whetstone set

This set of whetstones was affordable at 25$ and I was able to get it locally quickly.

Aliexpress.com

It all fit well without too many issues. With such precision applications be sure to torque bolts and screws correctly. It is essential that you do in precision applications to ensure equal load and proper contact between surfaces.

After this I also ran the whetstones over it quickly to remove all the burrs.

In addition to this I had to buy the following items:

I used some 3mm thick ground steel parallels for equal spacing vs the alu block.

I used this method to ensure that the side of the extrusions meet the adapter plate perfectly at the same reference plane.

Finally! it is starting to look like a CNC!

Conclusion – Mightymill build Part #3

Moving and re-assembling the machine frame was one of my biggest hurdles. But once I got going it went relatively smooth. The Farmer Jack’s helped a lot with this.

I ordered some new goodies such as Delta ASDA-M controller, 400W ECMA 20b optical encoder ac servo motors. A 1 meter 000 grade granite parallel (all surfaces) and some nice steel AC servo motor mounts. It is now all on the train from China. Hoping for it to arrive at the beginning of December.

The next step will be making the adapters for the ball screws and motors, mounting the extrusions to the table and making the z-axis adapter so it can all be mounted together.

Stay Tuned!

Make sure to subscribe to the newsletter to make sure you don’t miss any updates.

Feel free to reach out to me on discord

CNC controllers require understanding of the exact length of tools in order to compensate for any tool length differences between milling cutters.

In this article, I go over the various style of tool length sensors available. Evaluating aspects such as accuracy, versatility, and cost.

Lower cost mechanical tool setter

The lower-cost mechanical tool setter such as the HLTNC CNC tool length sensor which I reviewed here is the most popular method of tool length measurement for DIY hobby or semi-pro CNC machines.

There are various models available on Aliexpress and at other sources of these lower cost mechanical tool setters. Internally they are all constructed similar. There is a plunger where the end mill presses down and internally there is a switch that gets contacted when the plunger is moved down by a certain distance.

The repeatability of this style of tool setter has been shown and tested to be within +-0.01mm – 10um. For most DIY, hobby, and semi-pro CNC machining applications this is suitable.

However you will face some challenges with irregular shaped milling tools such is index-facemills or bigger tools that exceed the size of the plunger.

In addition this style of mechanical tool setter does not allow you to measure the tool width.

The elecrtical connection of this type of tool length sensor is quite easy. It with either have an NC or NO type switch internally which allows you to connect it to your cnc controller board such as an UCCNC or Mach 3 breakout board.

Non-contact tool setter

Non contact tool setting system often recognizable the U shape is one of the most common methods for automatic tool length measurement for industrial machines.

These non-contact tool setter work by sending a small laser beam from the sending fork to the receiver fork on the other side.

Due to its non-mechanical nature of it, the accuracy and repeatability of it tool length measurement is generally around 1 – 2 um for the more affordable models such as the Sick WF2. On the higher range of the spectrum, the Renishaw NC4+ achieves a repeatability specification of +-0.5um for smaller tools and +-0.75um for larger tools

An additional benefit of this type of non-contact tool length sensor is that it allows irregularly shaped tools such as big insert-based face-mills to be accurately measured. This can often be a challenge when using a contact-based tool setter.

In addition to that you can do width measurement of an milling tool. This is especially beneficial to capture catastrophic damage to a milling tool or to verify that the physical tool matches the tool size as set in the cnc controller such as linuxcnc or UCCNC.

This width measurement is usually less accurate than length measurement. One can generally expect about 20 – 50 um measurement accuracy using a non-contact

Endmills and other milling tools are often not completely flat on the bottom face due to cutting edges etc. To measure the longest point of the endmill accurately this kind of non-contact tool length sensor requires the spindle to turn (~1000 rpm) while measuring the lowest point of the milling tool.

On the second hand market such non-contact tool setters are available from just 10$.

Popular term for this on european ebay is “Gabellichtschranke” which translates to “Fork light barrier” – one example of this is the Balluff BGL000A which you can often find new or used for 20 – 40$

Premium – mechanical tool setter

In comparison to lower-cost mechanical tool setters that are often used in the DIY CNC community, there is also a premium style available that can achieve repeatability numbers comparable to or better than non-contact tool setters.

Renishaw OTS 3D touch trigger tool setter is one of such premium mechanical tool setters.

While listed as a mechanical tool setter in this article, internally it is just like the non-contact tool setter optically based to achieve better stability and accuracy numbers.

Brand new the Renishaw OTS 3D tool setter costs approximately 2000$ but used these can be found for about 500 – 1000 $ on places like ebay.

The Renishaw OTS 3D tool setter lits a repeatability specification of 1.00 um.

There are a few copies of this design on aliexpress but they are still quite pricey at 1000+ $. At that price point, I would advise to look at second-hand options in the west.

Conclusion

In this article, the various kinds of tool length setters are described with each having their pro and cons. For most hobbies and DIY use, a lower cost mechanical tool setter will be sufficient.

Looking for more versatility and better accuracy? have a look at (second hand) non-contact tool setters.

Are you a job-shop that is running VMC all day long, with coolant etc – for you the premium 3D mechanical tool setters for a brand such as Renishaw might be the right option.

Introduction

While looking for a Integrated AC servo drive solution for my MightyMill build, I was tipped by the machine builder DamenCNC to look at the Delta ASDA-M series of AC servo drives for my gantry based cnc router.

In this article I’ll share some of the highlights that this product has to offer and my opinions regarding it.

What is the ASD-M AC servo drive?

The ASDA-M series goes beyond the typical servo drive. It includes 3-axis servo drives in one frame to support real-time data exchange among three axes without any time delay and provides 3-axis synchronous motion control with performance better than comparable servo drives. Such as a diy solution with ASDA-A3 and ASDA-A2 drives.

This synchronous features set allows a new gantry control function for synchronous control that is precisely adapted to pure rigid mechanical systems.

What features does the ASD-M have?

The main features of the Ac servo drive ASDA-M are as follows:

• 3 servo drives integrated in one unit
• 3 external encoder inputs for full closed loop operation (i.e. linear glass scales)
• Multi axis synchronised motion
• Expansion capabilities
• Step/dir input
• See Datasheet for all the information
• …

Benefits of such a drive for a (DIY) CNC router application?

One of the most challenging items I came across with my Mightymill build was Y-axis (gantry) synchronisation during high speed machining (HSM). In more DIY style of routers this is often something that is overlooked.

But when your goal is to develop a semi-professional system these become aspects that need to be mitigated.

When you are doing high speed machining (HSM) with feed rates of 10000mm/s and up AND require accurate parts gantry synchronisation becomes a serious problem.

The ASDA-M servo drives solves this solution by integrating multi axis synchronisation of Position, Speed and torque with a 16kHz loop speed. In addition it is basically 3 ASD-A2-0721-M drives integrated into one. This makes the overall installation much easier and compact

Alternative solution? ASD-A2 or ASD A3 drive

Delta offers other servo drives which have the gantry synchronisation function.

The options that I found comes down to basically tying two ASD-A2 drives together or two ASD-A3 drives.

This makes overall wiring and control more difficult. In addition to the lower slower sync rate between axis versus the 16 kHz that the ASD-M offers.

And in case with the A3 drives could require additional communication or hosting over a protocol such as Ethercat. Which makes it more difficult to integrate with DIY style cnc controllers such as UCCNC, Eding CNC or Acorn CNC.

How much does it cost?

Pricing of these ASDA drives varies a lot by country and from which dealer you get it from. I have seen western webshop prices of around 1500 USD while the Taiwan-based prices start from 700 USD for the ASD-M-0721-M model

There are various (second hand?) drives on ebay hovering around 600 USD mark. For the ASDA-M drive that I bought I paid on the lower end of the above range (Taiwan direct) ex taxes ex shipping ofcourse.

For basically three ASD-A2 drives in one I find that a good deal. Especially when looking at the extra features it offers.

How does it work-out in practice?

I’m currently waiting on my ASD-0721-M drives and ECMA-C10604RS AC servo motors to ship from Taiwan. When they arrive I will start mounting them on my machine and share the process with you. Make sure you subscribe to our e-mail newsletter. So you don’t miss these updates.

In the past few months its little brother the PrintNC Mini which extends the PrintNC capabilities further but in a much more compact and cost effective platform.

Features of the PrintNC Mini

The PrintNC mini is the baby that came out of all the findings and learnings of the (opensource) PrintNC project. The main improvements are as follows:

• Dual HGH15 block bearings on the Y axis – Fixes the main stiffness issue which was often one of the first upgrades done for the PrintNC V3.0

• Triple HGH15 bearing block configuration for the X axis – This setup drastically increases the stiffness of the X axis and Z axis assembly

• Re-design of the Z-axis – Allows for a smaller overall package while retaining simplicity to build and overall stiffness

• In-chassis Nema17 stepper motors allow for a much more compact build

PrintNC mini specifications

The standard build has the following dimensions:

• Wasteboard 395 x 500 mm
• Y axis beams: 75×50 mm steel tube
• X axis beam: 50×50 mm steel tube
• Z travel: 110mm
• Y travel: 300mm
• X travel: 300mm

The PrintNC Mini is designed with parametric modelling. You can scale-up and scale-down the design within reason.

What makes the PrintNC so special vs other DIY CNC router offering?

The PrintNC Mini’s design makes use of rectangular steel beams which are interconnected with bolts to keep it DIY friendly.

It avoided all the shortcomings from entry level desktop CNC-routers that are on the market.

Such as the Indymill and Carbide 3D’s Nomad which are far less capable then the PrintNc

To name a few:

• Ballscrews on all axis
• Dual and triple ” Hiwin ” bearing block design
• rigid steel frame
• Supports watercooled 65mm spindle
• 3D-printed parts on the parts you can get away with less tensile strength
• Designed to mill non-ferrous materials

PrintNC Mini current status

The PrintNC mini is still in beta mode. Meaning it is only available to a select few which have purchased the 3D files and get access to the member zone. At this point approximately 2 or 3 completed builds were spotted on the Discord.

It is too early to know the full extents and capabilities of the PrintNC mini. But looking at how it was designed and the already excellent performance of the PrintNC DIY router, I have full confidence in it.

Looking for something completely different?

Have a look at my build the Mightymill :

Conclusion

The PrintNC mini is currently in Beta status and a few builds have shown up on the discord. The design carries many improvements from the PrintNC V3.0 and seems to have addressed the main issues with that design.

It is too early to know the full extents and capabilities of the PrintNC mini due to small amount of builds out there.

I’ll definitely be watching closely at how it performs over the coming few months.

My build has finally begun and in this post I will take you on the building journey. Make sure you have read part 1 here.

The heavy duty z-axis

Due to wanting to use a BT30 ATC spindle I had to upgrade to a heavy duty Z-axis. The design I opted for was originally designed by Crabs Shed. I optimized some small aspects of it and added it to my design.

Milling the z-axis plates

Crabs Shed was so kind to help me with milling the z-axis plates for my diy cnc build.

He even made a video of him milling it:

Initial measurements

One of the first things I did after receiving the plates was mounting my Sorotec Blue line HGR20 rails to it and measuring the (relative) flatness of the base plate versus it.

I quickly noticed that the leftover milling marks had a significant impact on the relative flatness of the mounting surface.

As the rails I uses are P – grade and my carriages have p1 preload it adds additional requirements on the rail mounting surfaces and alignment versus each other.

The image above gives you an indication regarding the accuracy of various parts of the rail and the various grades available. Regular Aliexpress “china” rails will be grade C or worse undefined.

I measured too much deviation of the z-axis mounting plate for my liking. I really wanted to keep this under 0.01mm or better hence I had to take it one step further.

Surface correction step #1

I got myself a set of whetstones from aliexpress in order to get rid of the milling marks and achier an overall surface which is more uniform.

In terms of improving the surface finish and flatness this did the most. It removed all the high spots and made the surface much more even.

Whetstone set

This set of whetstones was affordable at 25$ and I was able to get it locally quickly.

Aliexpress.com

Surface correction step #2

I got myself a small and affordable granite surface plate. It came with a full test certificate and all.

I got this since i wanted to scrape the high spots of the mounting surfaces flat.

In addition to this I had to buy the following items:

In total I spent about 2 hours of scraping. Can significantly speed it up if your initial plate is flatter or if you use power tools such as a dremel to do the difficult work for you.

I was not completely happy about the Diamant bluing paste. When indents get smaller than 0.01mm you will have a very hard time seeing it. There are better alternatives out there albeit those are difficult to get as an European.

When fully assembled, I measured the parallelity of the top plate vs the mounting surface of the Z-axis and it is well below 0.01mm. Good enough for me!

Mounting the Lusintun ATC spindle

Time to start drilling holes in the 100mm spindle mount.

The spindle mount was too tall so I could not fit a bigger size drill in my BF25L. So i had to put it on the side and ream it to the final size of 8.5mm.

How heavy is the Z-axis?

I would have guessed that it would have been heavier but 25.4kg it is!

New goodies

As I’m still waiting on quite a few milled parts for my cnc build I put my focus a bit on other aspects of the build.

HGR20 linear rails covers

To protect my nice precision and preloaded linear rails I know that I needed to get myself some linear rail protection. After looking on Aliexpress I found myself some.

I got 6 of the 600mm length ones. This is probably a bit too long for my build but I was afraid to get ones which were too short. It has many “internal supports”. This enables you to cut it to size when needed. You can find them here on Aliexpress.

Tool length measurement

In addition to that my tool length measurement sensor just came in. It seems like quite a solid and professional item. I also made a separate review of it. For those interested you can find it here.

Finishing up the machine frame

As I had too little concret last time, I went to Hornbach, got some more and finished up the second leg of the machine frame.

I cut some extra holes in the legs as I wanted to make sure it was completely filled.

After grinding it open I observed that it was completely filled. So there is no need for doing this. the total weight of a single concrete filled leg is 120kg now.

Mounting for the machine feet

I started cutting some 120x120mm squares out of 10mm thick steel plate to have a surface to mount my GD-60 levelling feet to.

After cutting and tapping it was time to grind off some paint from the legs and weld them to the frame. It made a little mess of my previously beautifully painted kern red paint.

I will have to clean up the spatter a bit and re-paint the damaged parts.

Conclusion – Mightymill build Part #2

Getting the z-axis mounting surfaces flat and dialed in was initially one of my bigger concerns. But after wet sanding it and scraping I was quite happy with the result.

Additionally I was able to make some progress regarding the machine frame.

Now it is finally ready to go to the next step and that is mounting and aligning the alu extrusions and rails to each other. And start making my design into reality.

Make sure to subscribe to the newsletter to make sure you don’t miss any updates.

Feel free to reach out to me on discord

My build has finally begun and in this post I will take you on the building journey.

The machine frame

My machine frame is built out of 120x120x4mm thick steel tubing. I chose this size as I could get it relatively cheaply. It is massive overkill, however. As my shed only has a small door it needed to be able to come apart easily so I opted for various bolted connections instead of welding it fully (which would have been much and much easier).

I added some extra supports for the machine bed which is not shown in the image above.

For painting I opted for Nelfamar Vinyl Plus which as a marine grade vinyl paint. I used this is various outside based projects before and it is easy to use, durable and quite affordable. The color I opted for is RAL3002 “Kern red”.

Concrete time

I wanted to take advantage of the vibration dampening properties of filling the machine frame with polymer concrete such as UHPC or Durfill. I describe the workings of that in this post here.

As Durfill is relatively expensive when you use it in low quantities I opted for using off-the-shelf pre-mixed concrete from the local Hornbach which at 2.75 EUR per 25kg bag is a steal. I added Moertelshop’s Flup4 to it to enhance its properties.

To cover the bolts I added grease to them and tighten them to the steel base. After semi-drying the concrete they were removed. I still need to do one of the steel legs. When I have some time in the coming weeks I’ll proceed with it.

Aluminium extrusion

For the machine itself, I used 80x80mm aluminum extrusion for the Y axis and 120x120mm extrusion for the X axis. The gantry risers are also 120x120mm aluminum t-slot profile.

Again to enhance it’s stiffness and vibration dampening properties I have filled these with UHPC polymer concrete. The specific product I used is Durfill by Durcrete.

It performs better than regular concrete or run-of-the-mill UHPC. As the volume for these aluminum extrusions is limited I chose the more expensive but better-performing Durfill product.

Polymer concrete time – Durfill

Durfill expands a little to ensure a good hold on the containing material. To ensure a clean mounting surface a spacer was created to prevent it from expanding past the edge.

Afterwards the spacer can just be washed out by washing it out with water. This method is extremely simple but works well.

With the Durfill, I just followed the mixing instructions and poured it into the aluminum extrusion. Whatever was spilled on the side, just wash it off with the garden hose.

After drying, the washed out spacers of the riser tubes look like this:

I did some video recording during this process. I’ll be uploading that somewhere in the coming weeks on my youtube channel.

The gantry tubes etc have also dried at this point.The weight of the table frame is about 400kg after filling it with concrete and mounting it togther.

Mounting and adjusting the Linear guides

For my design I utilized Sorotec Blue‘s HGR20 rails with P – precision class and P1 preload.

One of the first things I had to do was to create an adapter in order to allow two HGW20CC carriages and my dial indicator to be mounted to it. With a hole in the middle to tighten the t-slot nut and bolt. If you are interested in the drawing of this just hit me up on Discord.

After finishing this on my HBM BF25 mill (glorified drill press) I mounted it on my Y axis rail.

Buy these t-slot nuts!

I used the regular “twist and turn” t-slot nuts before and they are… very bad for this application. After switching to the model on the right all my adjustment issues went away. I was able to align the whole rail in just a few minutes instead of an hour of tinkering. Do not cheap out on your (t-slot) nuts!

Get it here

After using the premium of style of t-nuts the adjustment went swiftly.

Video of the linear alignment

I aligned the distance between the blocks to about 10um. This took about 15 minutes of work. Good enough for me

Conclusion – Mightymill build Part #1

I was able to make a good amount of progress so far. The next step is starting with the Z-axis and getting that going. Make sure to subscribe to the newsletter to make sure you don’t miss any updates.

Feel free to reach out to me on discord

Continue reading