Another Build Thread

Infrastructure Weekend

Before we can get started with this build, I need a sturdy place to mount the router and I need power. That was my focus this weekend. I have an old Sears workbench (anybody remember Sears?) that is made primarily of sheet metal, but has a very heavy top deck. I'm not sure what the deck is made of, but it is heavy and 1-5/8" thick. The deck measures 24" x 72". I have been using it to collect "stuff". Now it will be re-purposed as my router stand.

The top deck is not secured to the frame so I removed it to see how sturdy the frame was. It felt a little wobbly so I added a 2' x 4' piece of plywood to the back which stiffened it up considerably. The plywood will also serve as a place where i can attach power panels, etc.

The router will require 110 VAC power as well as 220 VAC power. The 110 VAC is already available on a 20 amp circuit. We used to have a swimming pool out back when the grandkids were little but it is long since been dismantled. I had a 20 amp 110 VAC circuit dedicated for the pump and filter so I re-routed the wiring back into the shop and installed a 220 VAC outlet next to the existing 110 VAC outlet. Then I removed the 20 amp single pole breaker and installed a double pole breaker to give me 220 volts at the outlet. There will only be 2 cables going to the wall. All the rest of the electrics will be on the stand or machine.

The router itself will be attached to a sheet of 3/4" plywood measuring 48" x 31-1/2". This will all get fastened to the top deck described above. The attached pictures give you an idea of what we accomplished this weekend.

CNC Router Specs

Here are the specs that I am using for the router build:

• Travel: X-Axis 22.44” (570 mm)
• Y-Axis 30.5” (775 mm)
• Z-Axis 4” (100 mm)
• Linear Guide: V-Groove Bearings
• Linear Drive: GT-3 Timing Belt
• Linear Drive Z-Axis: ACME Lead Screw
• Drive Motors: 345oz-in (2.45 Nm) NEMA 23 Stepper Motors with DM542T drivers
• Controller: RMH v3.1
• Construction: V-Slot aluminum extrusions – 2040, 2060, 2080
• Spindle: Vevor 1.5 kW, VFD, 220 volt, 24,000 RPM
• Rapid Speed: 200 ipm (inches per minute)
• Cutting Speed: 1/4" end mill, full width cut, 0.100" depth of cut, 50 ipm, material - hardwood (This is a fairly easy cut and is probably less than half the true cutting capacity)

Previews of coming attractions

My next posts are going to cover spindle selection, stepper driver selection, and most important - controller selection. Then we can start wiring this puppy. Stay tuned.

Albert

Stepper Motor Driver Selection

As I mentioned at the start of this thread, I wanted to provide some insight as to why I chose the various components to build the CNC router. I am also trying to get caught up to the point where I can report on current activities. So let's get through the background info starting with the stepper motor drivers.

I confined my search to stand alone units only. I saw some packages that incorporated four drivers in the same package. I believe for ease of troubleshooting and repair that stand alone units are best. But, my O' my, what a bewildering array to choose from.

My intent was to evaluate three different drivers from three different price ranges in a side by side comparison. In addition, this would give me a chance to test the NEMA 23 motors that came with the OX mechanical kit.

The three drivers I chose to evaluate were:

• TB6600 rated @ 4 amps from Amazon for $10.99
• DM542T rated at 4.2 amps from Amazon for $28.99
• Gecko rated @ 6.0 amps for $100.46

The pedigree for the Gecko drivers speaks for itself and I was not going to spend a hundred bucks just to run a test. If the lower priced drivers failed to do the job, then I would consider the Geckos.

These drivers work better with linear unregulated power supplies (like a simple transformer/bridge rectifier/filter) than with a switching power supply, because a switching supply will limit the availability of power when it's needed most, at the instant the motor starts to move.

My test setup is a 36V 10A homebrew power supply, an Arduino to provide a step pulse input which ramps up slowly from zero to establish the maximum running speed, the driver, and the 4 NEMA 23 motors supplied with the OX kit. No shielding or filtering was used, because I wanted to see how good the shielding and filtering is inside the driver under adverse conditions. Peak current for my motors is 3.0 amps so I selected 2.84 amps via the DIP switches on the driver. The documentation indicates that this is the peak current setting and that the RMS current value is considerably less, so I may be able to coax more torque out of these motors once everything is set up and running. I also tried various microstep settings from 2 through 128.

Using the Arduino configuration I was able to ramp the speed of the motors up and down and reverse direction with a button push. The Arduino setup is shown in the accompanying diagram. I will post the code if anyone is interested.

Motors connected to the TB6600 driver skipped steps and made alarmingly loud grinding noises, even when standing still. Not one of the motors performed adequately at any RPM. When I dropped the DM542T into the same setup, ALL of my motors worked fine, with no noise or skipped steps, produced boatloads of torque, and ran cool and quiet.

Stepping motors should not be noisy. Listen to a quality 3D printer and decide if it sounds more like a flute, or more like a bunch of gravel pouring out of a dump truck. If it's the latter, it's almost certainly because the driver is poorly designed. All that extra noise is power being converted into heat and noise instead of useful work.

I ordered three more DM542Ts for my CNC setup and tossed the TB6600 right in the recycling bin. You can spend a lot more on something like a geckodrive but you won't see a boost in performance commensurate with the 300%-500% price increase, unless you have very specific torque or voltage requirements.

The next step will be to test the drivers and stepper motors with the master controller. So I mounted everything on a breadboard and am waiting to finish wiring the controller so I can test the operation of everything before it goes on the router.

CNC Controller - Part 1

I spent a lot of time and research before making a decision regarding the type of machine control I wanted to implement. This is also the riskiest purchase I will be making which I will explain as we go along. Naturally, one of the major factors would be cost. But there are other factors to take into consideration such as ease of implementation, software requirements and environmental considerations. In the end, the decision comes down to whether you want a PC controlled CNC machine or do you want a stand-alone machine with integrated controller.

Let’s discuss PC control first. To begin with, you need a dedicated PC for the CNC machine, preferably a desktop unit. It does not have to be a high power PC, but it is still an additional cost and it has to operate in a dirty, dusty environment. Then you have to buy the software to run the CNC. The standard appears to be Mach3. Why would you go with anything else when there are so many resources to refer to? But the cost is $175 per license. Then things start to get tricky. You need a so-called breakout board to interface between the PC and the stepper drivers and other peripherals. There are lots of cheap Chinese devices out on the market and there are lots of horror stories associated with them.

If you want to avoid the risk of dealing with junk then the Smoothstepper seems to be the unanimous choice. But that is another 200 bucks. Then I am still going to install all of this gear into an enclosure and wire it all up which will cost at least another $200. To top it all off I would want to have an MPG (Manual Pulse Generator) in order to facilitate moving all three axes manually.

So what about stand-alone controllers? Well, the first thing is you don’t need a dedicated PC. You can do all of your design on a remote PC, transfer this to your CAM software (on the same PC), then load the G-code file onto a USB drive and plug it into the CNC machine controller. Seems pretty simple, but do I need to buy any additional software? The short answer is no. What you need to buy is the hardware with integrated firmware. At the present time the best stand-alone solution seems to be (for me at least) the 4 Axis 500KHz Linkage Offline Motion Controller System RMHV3.1/DDCSV3.1. It is available from Amazon for $275 and includes a 100 Pulse Handwheel MPG with Emergency Stop. In summary here are some specifics:

• 4.3 inches TFT screen, resolution ratio: 480*272;
• Input power: 24~32VDC;
• Output power: 0-10V, mainly for spindle speed
• MPG Resolution: 100PPR;
• 17 operational keys;
• Supports the USB flash drive to read the G code,and the size of G code file has no requirement;
• The highest uniaxial output pulse is 500KHz and the pulse width can be adjusted.
• It supports most common stepper motor drivers

From a cost and operating standpoint, my decision was the Offline Motion Controller System RMHV3.1/DDCSV3.1. Keeping my fingers crossed that this was not a big mistake! We'll know in the next few days as I get this prepped and wired.

A word of warning! This is definitely NOT plug and play. For one thing, the manual is written in Chinglish which is full of grammatical and syntax errors. Therefor if you are not familiar with concepts such as open collector outputs and optically isolated grounds I would advise staying away from this controller.

CNC Controller - Part 2

I wanted to minimize the number of wires and cables and I wanted a hand held controller similar to the CNC Router we have at work. I also did not want a separate MPG and its associated wiring to deal with. My solution was to disassemble the MPG and mount the individual components along with the controller in a 7” x 7” x 2” plastic enclosure. This way I could wire the MPG components directly to the controller inside the enclosure. The rest of the connections to the main electrical cabinet are made via a DB25 connector.

The RMHV3.1 has input connections for end of travel limit switches on four axes. I chose to go with soft limits and that saves eight wires plus ground that does not have to go to the main electrical panel and then eight separate pairs going to limit switches on the machine. I will install hard limits for home in the X, Y, and Z axes.
Before cutting the openings for the controller and associated components, I covered the enclosure with painters tape to make it easier to mark the cut lines and to protect the enclosure finish. The tape would be removed after all the cutting and fitting was done. I then measured and laid out the openings for the parts I harvested from the hand held MPG and mounted everything in the case.

Next step was to wire up the controller. If I were to utilize all the I/O available on the controller, it would require a 37 pin DB connector to link it with the main electrical cabinet. By selecting only the functions I intend to use (i.e. eliminating end of travel limit switches) I was able to utilize a 25 pin DB connector.
The connections to the RMSV3.1 are divided up into four ports:

• MGP Port – manually generates pulses to drive stepper motors
• Stepper Port – Connects to stepper motor drivers
• Spindle Port – Connects to VFD for spindle control
• Input Port – For all peripheral devices i.e. limit switches, probe, etc.

I started the wiring by connecting the individual components to the MGP port. This did not involve connections to the DB25. The individual components that were wired into the MGP port are:

• E-Stop button
• Rotary Encoder
• Speed selector switch
• Axis selector switch

Attached is a chart I developed as an aid to understanding and wiring the MPG components. SW1 and SW2 refer to the selector switches for speed and axis. The colors correspond to wire colors to aid in identification.

Next it was time to wire up the DB25. Once again I developed a chart to keep the connections and pin numbers straight. Colors on the chart matched the wire colors. The DB25 utilized crimp pins that were then inserted into the connector housing. This saved a lot of soldering. I used ferrules on the other end for insertion into the RMSV3.1.

Missing Pieces

I decided to take a break from controls and wiring in order to check out all of the mechanical components in the kit.

After unpacking all of the mechanical parts, I started to layout all of the structural members to see how everything would fit together. It became Immediately apparent that I was missing a 2040 member 750 mm long. This is the center structural member that supports the spoiler board. Also missing were four end plates that join the 2080 longitudinal members to the 2040 horizontal members.

Even more surprising, there were additional structural members included that I have no idea what to do with. These are:

2 pcs. – 2080 x 28” long
2 pcs. – 2040 x 18” long

If anyone has any idea what these additional pieces are meant for I would love to know.

CNC Controller - Part 3

OK, Great! I’ve got all the MPG wiring finished and also the wiring from the DB25 to the controller finished. However, I did not connect any wires to the Stepper Port. I’m going to connect the wires from my breadboard test setup to the controller.

I already tested the motors and the drivers using the Arduino as a pulse generator, so I know we’re good on that end. Now we get to see how the MPG works and if the controller can run the stepper motors.

Step 1 is to test the MPG unit. I connected the drivers mounted on the breadboard to the RMSV3.1 controller. Then I switched the controller MODE to MPG and selected the X-axis using the rotary switch on my control box. Then by rotating the encoder wheel I was able to get the X-axis stepper motor to turn. I tried all three speed settings using the rotary switch on the hand held controller and that seems to be working correctly. This was then repeated for the Y and Z axes.

There are three speed positions available, 1X, 10X and 100X. The 1X position results in a very clunky motion and in my opinion not very useful. The 10X gives a nice smooth motion and 100X results in a very rapid but smooth motion.

I switched the controller MODE to CONT and used the buttons on the RMSV3.1 to move the stepper motors. I was very pleased to see that all the motors performed very smoothly and quietly at 100% speed setting. I should also mention that I had the microstep setting at 8 and the current limit at 2.84 Amps.

Now that I’m satisfied that the RMSV3.1, MPG, stepper drivers and stepper motors are all playing nice with each other, it’s time to look at the overall electrical arrangement. The first step is to disconnect the drivers on the test bench from the RMSV3.1 and connect the wires from the DB25 in the hand held controller to the corresponding terminals in the stepper port. Then we need to migrate all of the components from the test bench into the main electrical cabinet which will be the subject of my next post.

Electrical General Arrangement

One of the issues I’ve been struggling with is the overall arrangement of the electrical controls and wiring. After considering all of the relevant factors I came up with the following design criteria:

• The only thing mounted on the wall will be two Edison receptacles, one for 110 VAC and one for 220 VAC; thus there will be only two cables going out to the wall, everything else will be mounted on the router.
• The RMHV3.1 will be mounted in a hand held box and this will be our primary control; in addition to the RMFV3.1 we will incorporate an MPG and E-Stop.
• The VFD for the spindle will be mounted in its own separate cabinet with EMI filter and 12 VDC power supply for cooling fan. (220 VAC supply)
• Everything else gets mounted in a separate cabinet (110 VAC supply); this includes 36 VDC linear power supply, 24 VDC auxiliary power supply, 24 VDC power supply for RMHV3.1, 12 volt power supply for system cooling, stepper drivers and everything else.

Main Power Panel
Inputs and Outputs
In order to finalize the layout it was necessary to define all of the inputs and outputs into the main power panel. The inputs would be:

• 110 VAC main power
• DB25 from the master controller

The outputs were determined to be:

• 4x cables to the stepper motors
• 3x cables to the Home limit switches
• 1x signal cable to the VFD cabinet
• 1x power cable (24 VDC) for the contactor coil
• DB25 connection to RMHV3.1 master controller

Main Power Panel Components

• IDE power entry module including fuse and EMI filter
• ON-OFF push button switch DPST
• 36 VDC power supply for stepper drivers consisting of:
• Toroid transformer
• Full Wave bridge rectifier
• 10,000 uF smoothing capacitor
• 24 VDC power supply for RMHV3.1 master controller. This is a small DIN mounted supply rated at 1 amp.
• 24 VDC power supply for Contactor coil
• 12 VDC power supply for cooling fans
• 4x Stepper Motor Drivers
• 2x cooling fans and filters

VFD Cabinet
I wanted to isolate the Variable Frequency Drive (VFD) from the main electrical cabinet for two reasons. The first reason is that it uses 220 volt AC power and I did not want to mix the two high voltages in the same cabinet for safety reasons. The second reason is that all VFDs generate a lot of electrical noise which can be transmitted back through the power lines and also radiated to nearby components.

Being extremely lazy I did not want to turn on two separate switches – one for 110 VAC power and another for 220 VAC power. That is the reason for the contactor. Thus when I turn on the AC power to the main electrical cabinet, it will energize the contactor and power up the VFD.

I had an old metal electrical box that I rescued on a dumpster diving expedition. My original thoughts were to mount the VFD in this enclosure. However as this build is progressing I am having second thoughts. One half of the router base is a set of drawers for storage. The other half is completely open space. At this point I am leaning toward mounting the VFD on the back wall of this open space along with the contactor and EMI filter. There should be enough air circulation so that I would not have to install a cooling fan.



I found a nice piece of aluminum plate in my garage stash and laid out the VFD components on it and they fit quite nicely. That ices it! The old steel enclosure goes back to the dumpster and the VFD and associated components will go inside the router base.



So to summarize:
VFD Assembly Inputs:

• 220 VAC @ 20 amps
• 24 VDC for contactor coil
• 0-10 VDC for spindle speed from RMSV3.1
• Open collector output from RMSV3.1 for spindle ON/OFF
• Control Ground

VFD Assembly Outputs:

• 0-220 V variable frequency to spindle

To be determined: Where to locate the 12 volt supply for spindle cooling pump and fan.

Mystery Solved!

Back in Post #12 I reported that I was missing a piece of 2040 that supported the spoiler board as well as end pieces that attach the 2080 longitudinal rails to the 2040 horizontal members.

In that same post I reported that I had 2 pieces of 2080 and two pieces of 2040 that I had no idea what to do with. They were not mentioned in the construction manual. While surfing the CNC net I found a video describing an upgrade to the standard OX build. It consisted of adding an additional 2040 horizontal member on each end to stiffen the frame and a couple of 2080 longitudinal members to support the spoiler board.

It seems as though this upgrade was incorporated into the kit, and that's a good thing!

The end plates that I thought were missing have also been included. However they are no longer simple triangle plates. See photos.

An unexpected surprise!

While taking a closer look at all of the structural members I noticed that the ends were pre-tapped for M5 threads. What a time saver this will be not to mention lubrication mess and chips to clean up.



Wheels

While taking a break from the electrics I Managed to get all of the wheels assembled. Each assembly consists of two bearings, a spacer and a wheel. The entire assembly is pressed together. I used a 5 mm bolt to ensure concentricity and put the assembly between a pair of washers. I threaded a nut onto the bolt and held the nut in a vice while I tightened the bolt. Worked like a charm!

Today I began working on the gantry end plates by bolting on the stepper motors. Next I attempted to install the rollers on the opposite side of the gantry plate and immediately hit a snag. The drawings call for M5 x 30 mm low profile screws for attaching the roller assemblies to the gantry plate. I searched high and low and could not find these screws. I went on line and ordered 30 of these buggers and now I have to wait on the USPS. Meanwhile I thought I might bore you with how and why I selected the spindle for my CNC router.



The Router Spindle

Like most everything else involving CNC routers, there are a variety of spindle types out there. Some made for wood, some made for metal, etc. As usual there are many factors involved when choosing a spindle for either your pre-build machine, or your homemade CNC machine.

If you are building a CNC machine, then you also need to make important decisions regarding the router spindle. Again, I realize a lot of hobbyists are bound by budget and that surely applies to me. After all, that’s part of the challenge. However, there are many choices to choose from within the same price bracket.

One of the most important decisions for the hobbyists, I believe, is the noise level. If you are anywhere near neighbors or your own house, then this is a huge consideration. You don’t want to be running an extremely loud machine for hours at a time.

Power
When looking at trim routers that can be adapted for CNC use, I see that they are all in the 1 HP range. This is equivalent to 746 watts of power. So this gave me a good starting point. However when you look at standard routers such as the Bosch 1617es and the DeWalt DW618K, they are both rated at 2.25 HP or 1.68 kw. This is sufficient power to more than handle any woodworking jobs I will attempt. Many spindle motors are listed by their rating in kilowatts instead of horsepower, so I will be searching for a spindle motor in the 0.8 to 1.5 kw range.

A 1.5-KW motor is a good choice for mid-sized to smaller machines, as well as even some large benchtop machines. You can perform the most common operations on wood and plastic workpieces as well as cut out parts from aluminum sheet.

220 volts vs. 110 volts
While it's said that an electric motor doesn't "know" the difference between 110 and 220, your electric service will. A motor running on 220 pulls half the amperage than on 110. Thus, if you want to use a 2 HP motor, it will pull about 18-21 amps on a 110 circuit. If, however, you want a 2 HP (1.5 KW) motor on your CNC machine running on 110, you'll have to furnish a 30 amp circuit which means a new breaker and heavier wire since pulling 20+ amps on a 12ga. wire can be done, but you're asking for trouble, a possible fire, and complaints from family members when the lights blink every time you turn on the machine, not to mention numerous breaker trips. Since the wiring is already in place, I can convert to 220 by simply installing a 220 volt breaker in the panel box. It’s a no brainer!

Air Cooled vs. Water Cooled
While searching through the advantages and disadvantages of hand held routers in the 1 HP range, one of the complaints was that these machines tend to run very hot, often to the point where the operator can no longer hold the router. Looking over the many choices available, it seems to me that 1 HP is the breakeven point for air vs. water cooling. There are other factors impacting how hot the router will get. Some jobs on the CNC may run an hour or more. Another factor is how hard the machine is working, i.e. type of material, depth of cut, etc. So, all things being equal, IMHO a water-cooled spindle is preferable to air cooled once we hit the 1 HP (.746 KW) point. Closed loop cooling systems for computer MPUs are cheap and readily available. In addition, water-cooled spindles are quieter than air-cooled.

Why Choose VFD Spindles
VFD spindles are a great alternative to using hand held routers to power your CNC machine. Hand routers are designed for small jobs in your hand not for running for hours on end. Hand routers often handle this unintended workload surprisingly well however they are in almost every way the wrong tool for the job.

VFD Spindle Pros:

• Significantly quieter than handheld routers
• Better speed range and availability of collet sizes.
• Much higher duty cycle – they are designed to run all day.
• No brushes to wear out.
• Higher torque in lower speed range
• Considerably better tool runout (accuracy) which leads to higher quality output and longer tool life.
• GCode Control (automatic control from software) – Most often (depending on the CNC controller) you can start/stop and set the speed of the spindle via G-code and also override these speeds from the machine control software in real-time.

VFD Spindle Cons:

• Price – doing it right with approved products will absolutely cost more than picking up a handheld router and bolting it to a CNC machine.
• Potentially RF Noisy (electronically not audibly) – Cheap or incorrectly commissioned/programmed VFD’s can send you down a rabbit hole chasing EMF noise-causing all sorts of failures during machining. It’s hard to chase noise without an oscilloscope and training.

At the end of the day the advantages of VFD significantly outweigh the disadvantages. Today we have pre-engineered systems that are truly plug and play.
And the Winner is….
Vevor 1.5kw 220v Water Cooled Spindle Motor Kit: VFD, Clamp, Pump, Pipe, and 14pcs ER-11

The kit includes the spindle motor itself, the VFD, fourteen collet inserts, a 5-m roll of tubing, a submersible water pump, and a spindle motor mounting bracket. The collet size this motor uses is ER11, which is relatively small and only allows you to use bits with a shank less than 7 mm (.276”). You could upgrade the collet assembly, but you would have to consider the limitations of the motor before attempting to put any really large bits in it. With that having been said, it is highly unlikely that we will be using tools with a shank greater than ¼” (6 mm).

• Size: 1.5 KW
• Cooling: Water cooling with pump and hoses included
• Control style: 220-V AC VFD with computer control
• Collet size: ER11
• Rated speed: 0-24,000 RPM
• Price: ~$270
• Available from: Vevor

The only concern I have is the quality of the VFD drive at this price point. However I will go with this package to start with. If the drive begins to present a problem I will replace it with a Hitachi WJ200-015SSF. One other note – I plan to ditch the pump supplied with this package in favor of a closed loop CPU cooling system.

Been a few days since my last post. Been extremely busy with work, college visits with my grandson and did I mention that I mentor a teen robotics club. Still I managed to get a little work done on the router.

Finished wiring the VFD electrical panel. The inputs and output wires connect to terminal strips at the top of the panel.


As mentioned in a previous post, the kit did not come with any 30 mm M5 low profile screws. I had to order some and they finally showed up. I finished attaching the wheels to both gantry plates. The bottom wheels are installed with eccentric spacers to permit a tight fit to the rails.


Next I worked on the X-axis gantry front plate by attaching the spacer blocks and 8 mm Acme nut block. I have some concerns about the nut block as it does not seem to be fitting up square to the rest of the plate. I will keep an eye on this and if it becomes a problem I will have to enlarge one of the holes and pin the block to the plate in order to get it into alignment.

I then attached the wheels to the spacer blocks. The wheels on one side are fitted to eccentric spacers for adjustment.




Finally, I have been pondering how to attach the beast of a spindle to the 2060 V-slot carriage. As you can see from the photo, the mounting bracket base is much wider than the 2060 Z-axis V slot. My solution is to manufacture a carrier plate that can be bolted securely to the 2060 with extensions at one end to accommodate the M8 spindle bracket mounting bolts. The carrier plate would be attached with eight M5 screws.

I ordered a 3/16" thick sheet of T6061 aluminum 12" x 12". I drew up the carrier plate in AutoCad from measurements taken from the spindle mounting bracket and the 2060 carrier rail. When my aluminum plate shows up, I'll cut out the part on the robotics club CNC router.

Last night while fitting all the pieces together I realized that the overhang from the spindle mounting bracket is going to interfere with the V-wheel mounting bolts thus restricting the amount of Z-axis travel to about 3". I would like at least 4" of travel and preferably 5" depending on the height of the spoiler board. As the build progresses I will make a determination of exactly how much Z-travel is available. At that point I may opt to buy a longer 2060 carrier and lead screw (250 mm). We shall see as we progress.

Some Thoughts on Spindle Mounting

If we go back and look at my first post, I attached a stock photo of a complete OX-CNC Router. I you look closely at the photo you can see clearly that the V-rail for the Z-axis is supported by four rollers. Because of the heavy overhung load due to the heavy spindle and mount, it would be highly desirable to have the extra rigidity provided by the fourth roller.



This would seem on the surface to be a straightforward modification. Simply replace the existing 3-hole spacer block with a 4-hole spacer block, add an extra pair of wheels and Bob’s your uncle. Hold on! Not so fast my friends. This is turning out to be a bit more of a slog than I first supposed.

Based on the stock photo showing four rollers, one would think that a 4-hole spacer block would be readily available from one of the on-line suppliers of OX kits. Such is not the case! I have scoured the internet looking for a 4-hole spacer block and have come up with zilch – nada. Oh, there are lots of 3-hole spacer blocks to be had, but no 4-hole, at least not that I have been able to come up with.

The next avenue I explored was making my own 4-hole spacer blocks. The original spacer blocks that came with the kit were fabricated from 12mm x 20mm anodized aluminum. This size is not readily available, at least in small quantities so I opted for ½” x ¾” aluminum stock which is easily obtained at fairly low cost. I drew up the part in AutoCad but ran into another snag. The eccentric adjustment spacer is a precision fit into a 7.13mm hole which is not a standard drill size. Grainger offers a 7.13mm reamer at the bargain price of $176.00. What I don’t want is a loose fit on the eccentric; otherwise I would be better off with no fourth wheel. So that essentially kills the idea of manufacturing my own spacer blocks.

So, how to get that fourth wheel installed without breaking the bank and re-designing the Z-axis? My solution is to buy a 3-hole spacer block, saw off each end and mount it to the X-axis plate above the existing 3-hole spacer block. This way I would only have to drill a pair of holes in the X-axis plate to accommodate the 5mm fixing screws. See the attached drawing for details.

So in order to make these mods I ordered a pair of wheel assemblies, a spacer block, some ¼” spacers, a pair of eccentric spacers and hardware for installation. While at it I also ordered an anti-backlash nut block to replace the one that came with the kit. So now we wait.

Electrical Cabinet

Meanwhile, I got started on the electrical cabinet by installing the exhaust outlets at the top of the cabinet and the fans in the bottom of the cabinet. The fiberglass reinforced plastic is tough as hell on saw blades. Completely destroyed two blades cutting the holes for the inlet and outlet.

Not much going on in the shop these past few days. Been working on the electrical panel layout and component placement. This will determine where to put the penetrations for the cables to the stepper motors and limit switches.
Referring to the attached drawing (sorry if it’s cocked a little), the toroid transformer, 10,000uF capacitor and 50 amp full wave bridge comprise the 36 volt power supply. The transformer is double wound with each winding rated at 7.5 amps. I am wiring them in parallel to increase the overall capacity to 15 amps.

In the upper right hand corner are the auxiliary power supplies. There is a 24 volt power supply dedicated to the RMHV3.1, another 24 volt supply for the contactor coil located on the VFD panel and a 12 volt supply for the fans. These are DIN rail power supplies and take up very little room in the cabinet. (See photo).

I used conventional terminal strips for the 36 volt DC and 110 volt AC distribution.

For the control wiring I decided to use DIN rail terminal blocks. These are quite convenient to use since all you have to do is strip the end of the wire, insert and snap shut for a solid connection. The ID for each terminal block corresponds to one of the wires coming from the DB25 connector except for the 24 volt RMHV3.1 power which goes directly to the DIN mounted power supply. In the process of laying out the connections for the terminal block, I realized I forgot to include the separate 12 volt supply and ground connections for the probe and limit switches. I will have to go back into the control box and add these.

Happy Thanksgiving everyone and thanks for watching my thread.

Continuing on from my previous post, I have spent quite a few hours working on this project with not a lot to show for it. Nevertheless I have been soldiering on with most of the time spent on mounting the electrical components to the sub-plate which will get installed in the electrical cabinet. A good bit of time was spent putting labels on all the terminals. In addition, I installed the DB25 connector into the side of the electrical cabinet as well as the 4 connectors for the stepper motors.





I got tired of waiting for the spacer block that I was going to partition (see post #23) so I took one of the existing spacer blocks and cut it up in order to prove the concept. Seems to work pretty well - now I just need the other spacer block to finish the Z-plate assembly.



How are we doing with the budget?

Way back in post #1 I estimated the cost of this project to be $2000. Every project manager has to track expenses vs. budget so this is where we stand to date:

OX CNC kit $514.86
CNC Controller $291.50
Electrical Cabinet $143.99
Spindle & Accessories $262.99
Stepper Drivers $ 84.76
Misc. Elec. & Mech. parts $396.97

Total Spending $1695.07

Been a busy week, but I finally got all of the mechanical parts I need to get on with the build. I installed the new spacer block and anti-backlash nut on to the front plate of the X-axis assembly. I test fitted the Z-axis 2060 V-slot and everything looks good. I adjusted the eccentrics so all the wheels turn and the 2060 slides up and down without binding.



Next I attached the NEMA 23 stepper motor to the X-axis gantry back plate. This was followed by joining the front and back plates together with the rollers and appropriate spacers sandwiched in between.



Next we moved on to the Z-axis assembly. Here is where we ran into a bit of difficulty. The Bill of Materials specifies the spacer between the motor and the end plate to be 40 mm. However, when I assembled the unit this resulted in the motor coupling hitting up against the end plate. I took everything apart and measured the spacers and found that they were only 38 mm long. To solve the problem I used three 1 mm precision shims to increase the distance to 41 mm. This gave me 1 mm clearance between the coupling and the end plate.




Next I attached the motor assembly to the Z-axis 2060 V-slot. Then I slid the X-gantry assembly onto the Z-axis V-slot and threaded in the Acme lead screw into the flexible coupling. Then I tightened the coupling and brought the lock collar up against the bearing and tightened it as well. To finish off the Z-axis assembly, I slid the lock collar and bearing on to the lead screw and then attached the bottom plate to the 2060 V-slot. I adjusted all the eccentrics so that the Z-axis assembly moved freely when turning the coupling by hand and that all the wheels were contacting the V-slot.

Now it was time to move on to the X-Axis gantry assembly. The X-axis gantry consists of three aluminum extrusions: two 2060 x 500 mm and one 2040 x 500 mm as well as two side plates with NEMA 23 stepper motors attached. I started by attaching one of the side plates to the aluminum extrusions. Next I slid the X/Z gantry assembly onto the aluminum extrusions as well as sliding in some T-nuts for attaching the reinforcing corner brackets. Next, I attached the other side plate as well as the reinforcing corner brackets and we now have a complete X-gantry assembly. Before going any further I will check to make sure everything is square and plumb.





Sooner rather than later I must get back to the electrical wiring, but in the words of Scarlett O'hara " Tomorrow is another day"

While checking to be sure that everything is square and plumb I identified two problems. One of the issues involves mounting the spindle, which I will address a bit later. However, my immediate concern is that the Z-axis is not plumb. I spent a couple of hours measuring and thinking about the problem and I believe I have identified the root cause.

Using a machinist’s square I checked the plumb of the left side spacer vs. the right side spacer. First the left side. I pushed the straight edge up against the spacer block and observed whether or not the base was sitting squarely on the work plate. Indeed it was as I could not discern any gap whatsoever between the base and the work plate. Note that the bubble is just a touch off center to the left. This is shown in the following photo:



Next I moved to the right side and repeated the procedure. Here is where things began to get a bit sticky. I slid the machinist square along the work table until it just contacted the spacer block. I immediately observed that the base of the machinist square was still firmly on the work table but the bubble had shifted considerably to the right.



This was certainly an indication that the right spacer block was not square and perpendicular to the work table. But by how much was the question. Next, I moved the scale until it contacted the entire spacer block. Then I observed the relation of the base to the spacer block, there was a significant gap as shown in the next photo.



So where do we go from here? One option is to scrap the entire project and take up basket weaving or needlepoint. But that isn’t gonna happen!

Here’s what we’re going to do: Dis-assemble the whole bloody works and find out why the left side spacer blocks are cocked out of plumb. Fix the problem and I’ll get back to you.

As far as the spindle mount goes, my original thought was to machine an adapter plate to attach the spindle mount. But that would not work because of interference with the spacer mounting bolts. I decided to order a totally different spindle mount and we will see how that works out. Talk about project cost creep!

Meanwhile, I’ve got a bit of electrical wiring to deal with.

Acknowledgement

Thanks for the input Tim. I’m not quite at the stage where I will do actual tramming, but I will use your input when I get down to the final tramming and leveling. One quick question: Steelers or Eagles?

Progress

This is going to be somewhat difficult to describe, even using photos; but I will give it my best shot.

I completely dis-assembled the entire gantry, then proceeded to re-assemble it step by step. This time, instead of tightening the bolts as I went along, I left everything loose and only tightened everything at the end when it was square and level. Also, I alternated the tightening sequence so that one side or one sub-assembly would be torque balanced as I proceeded. Another significant change in the re-assembly process involved the Z-axis support rollers. I will try my best to explain.

If you go back and review posts #23 and #28 you will see that I added a 4th support roller for the Z-axis spindle support. If you look carefully, you can see that the spacer blocks for both sides are machined identical; that is they are machined with 7.12 mm openings for a 6 mm eccentric spacer. Aha! But the instructions provided say to use a 6 mm aluminum spacer on one side and a 6 mm spacer on the other.



So, I removed the 6 mm aluminum spacers and replaced them with the 6 mm eccentric spacers. Now I had some adjustment available when I wanted to tram my router. Each eccentric provides up to 1.25 mm of linear adjustment, so if I adjusted the eccentric 1 mm on the right, I could compensate 1 mm to the left and thus adjust the perpendicularity of my spindle. Jeez, I hope that makes sense to whoever is reading this!

OK folks, here is the bottom line: after re-assembling the entire gantry, and incorporating the techniques mentioned above, everything is within 0.5 degrees of perpendicularity which I expect to correct when I do the final tramming.

And now, if you don't mind, can we get back to the electrical wiring?





The inclinometer images represent the top of the Z-axis stepper motor, the X-axis gantry rail, and the base that the entire assembly rests on.