Joe's 1000x600x200 steel and aluminium router started

Progress update!

I got back to the project after a break of a few months to work on other things. I ended up pouring epoxy to bring the X rails into plane, then set the master rail straight using a taut wire, microscope and some software as described here:
http://www.mycncuk.com/threads/12966...r-level/page27

I now have both rails mounted and the gantry moves very smoothly.
Attachment 28236

Next step is to fabricate the motor brackets and mount the ballscrews and steppers. I've ordered belts and pulleys and am waiting on these so in the meantime I started to fabricate a stand for the machine. Since the machine is itself a rigid structure this stand only serves to anchor it to the floor at a decent working height, hence the lack of bracing. All 4 legs will be bolted to the concrete floor then the machine will sit ontop on leveling bolts, finally the machine frame will be clamped hard to the stand so that it doesn't throw itself across the room (voice of experience...).

Cut and welded yesterday, painting in silver hammerite today. The more I use it the more I love this little MMA stick welder. welded the 3mm box section mitres at only 45amps! Almost as good as a MIG without any complications.
Attachment 28237 Attachment 28238 Attachment 28239

I'm looking for some recommendations on bed materials, machine will mainly be used for aluminium and composites, so preferably something that can take coolant. My last machine used epoxy soaked MDF which was far from ideal. The plate would need to be 800x1300.

I don't have a large enough piece of aluminium plate. So options I'm looking at are:

1. 20mm Tufnol, can't find this in sheets longer than 1200 so would have to alter the frame, this isn't really a problem. I'm not sure whether this is rigid enough to span the cross beams that are on 400mm centres??

2. Purchase a piece of 20mm aluminium plate large enough, quiet expensive.

3. Locate some aluminium extrusion T-slot bed profile 'planks', I can't find these either.

Thoughts?

Quote:

---

Originally Posted by devmonkey

I'm looking for some recommendations on bed materials, machine will mainly be used for aluminium and composites, so preferably something that can take coolant. My last machine used epoxy soaked MDF which was far from ideal. The plate would need to be 800x1300.

I don't have a large enough piece of aluminium plate. So options I'm looking at are:

1. 20mm Tufnol, can't find this in sheets longer than 1200 so would have to alter the frame, this isn't really a problem. I'm not sure whether this is rigid enough to span the cross beams that are on 400mm centres??

2. Purchase a piece of 20mm aluminium plate large enough, quiet expensive.

3. Locate some aluminium extrusion T-slot bed profile 'planks', I can't find these either.

Thoughts?

---

A bit more research on different materials, Young's modulus (GPa):

MDF	4
SRBP (paxolin/tufnol)	6.5
Aluminium	69
Steel	200

The deflection of a rectangular section spanning two supports due to a downward force in the middle is inversely proportional to the Young's modulus * thickness^3.

So taking 16mm aluminium plate as a reference, thickness required for equivalent rigidity of those materials are:
MDF 41mm, SRBP 35mm.

Or another way, 25mm SRBP is equivalent to 12mm aluminium.

From this it is probably ok to conclude 25mm SRBP supported on 4 sides over rectangular bays in the machine frame that are 840x320mm is probably sufficient.

Quote:

---

Originally Posted by devmonkey

A bit more research on different materials, Young's modulus (GPa):

MDF	4
SRBP (paxolin/tufnol)	6.5
Aluminium	69
Steel	200

The deflection of a rectangular section spanning two supports due to a downward force in the middle is inversely proportional to the Young's modulus * thickness^3.

So taking 16mm aluminium plate as a reference, thickness required for equivalent rigidity of those materials are:
MDF 41mm, SRBP 35mm.

Or another way, 25mm SRBP is equivalent to 12mm aluminium.

From this it is probably ok to conclude 25mm SRBP supported on 4 sides over rectangular bays in the machine frame that are 840x320mm is probably sufficient.

---

I would probably use one large piece as the bed, then a second piece as the sacrificial board.

30mm HDPE makes a nice stable bed and you can cut a grid pattern to make it into a Vacuum bed. I bought a 10x5 sheet for £500 so if you get in touch with plastics supplier you might get a small piece or cutoff piece cheap enough.

Quote:

---

Originally Posted by JAZZCNC

30mm HDPE makes a nice stable bed and you can cut a grid pattern to make it into a Vacuum bed. I bought a 10x5 sheet for £500 so if you get in touch with plastics supplier you might get a small piece or cutoff piece cheap enough.

---

Ok thanks, that is interesting, YM of HDPE is only 0.8, how much support did you have under it, or what was the clear span? I guess we are only talking about supporting the mass of the part being machined and the downward force of the cutter, this is probably minimal, like to take a punt on what this force is?

Looks like an 8x4 sheet of 25mm SRBP is £380, this would give me enough for the base board and two spoil boards.

Stand levelled on floor and machine craned ontop to check final position, needs to be moved out from the wall a bit so I can fit/service the ballscrews motors. Starting to look like a CNC machine!
Attachment 28249

Looks seriously serious! Make sure you have enough room to move around it while bending over and enough room to fit a screwdriver in all the places you didn't think you'd need to fit a screwdriver when you first thought "I'll put it here!".

Finished making the last set of plates, these are for the X and Y axis, ballscrew mounting. Hopefully this is the last of the aluminium work on the manual mill. I still need to make some stepper mounting plates but leaning towards doing those in 3mm steel.The white parts are 3D printed standoffs for the X axis ballnut holders, I didn't have any large enough ally stock so we will have to see how they fare. They get clamped between two aluminium lumps so should be ok, if not will replace when I have some bar stock.

Attachment 28250 Attachment 28251

Any they all fit:
Attachment 28252 Attachment 28253 Attachment 28254

Was still waiting for the pulleys this morning so decided to align the Z axis assembly and fit the ballscrew. I needed a 0.25mm/10thou shim between the ballnut holder and Z-plate, but it was extremely difficult to get the shim into place as the holder is buried between the rails inside the assembly. A bit of lateral thinking and I used some plastic lamination film cut from a normal A4 laminating pouch.

This stuff has the advantage of a) being quite hard after lamination,b) having a heat activated adhesive. So using an old iron I laminated 4 layers of the stuff onto the back of the ballnut holder, this formed the perfect shim. I trimmed it with a scalpel and cut the bolt holes. Worked really well. Z axis is now complete apart from lubrication points. These are a pain with the nipples supplied with the hiwin carriages and the ballscrew so will need some thinking.

Attachment 28265 Attachment 28266 Attachment 28267 Attachment 28268

Just as I finished the Z the delivery guy turned up with my belts and pulleys, so I spent a rather boring couple of hours, boring, reaming, cross drilling and tapping 8 steel pulleys. All done now. I only installed one M5 grub screw per pulley, each shaft will have a flat on it, hopefully this will be sufficient. I could do with some longer grub screws, or maybe just use an M5 bolt given the low RPMs.

Attachment 28269 Attachment 28270 Attachment 28271 Attachment 28272 Attachment 28273

Made the motor mounts today, 3mm steel cut from some scrap box section. I used a cheap screwfix holesaw, wasn't at all convinced it would work. Lots of cutting fluid and smoked like crazy but delivered a very clean hole, cutter seems like new still. Just cleaned them up on the linisher and ripped the end of my finger nail off, very messy.

Attachment 28274 Attachment 28275 Attachment 28276

They are now painted.
Attachment 28278

For the machine bed I got a few quotes for 30mm HDPE, various Phenolic paper boards, etc, all around £300. I called Smiths for a quote on 1300x880x20mm aluminium plate, quote was £274, then they found a pre-cut piece 1500x990x20 which I can have for the same price. Seems a no-brainer so am picking it up tomorrow.

I will get a 1200x600 piece of phenolic for the spoil boards, these are a standard size and just right for my machine and readily available from electrical wholesalers.

Picked up the bed today, 82kg of 20mm plate! Not sure how I'm going to cut it, too heavy for manoeuvring over the table saw. Will probably just use the circular saw very carefully...
Attachment 28283

Duplicate post.

Bit of a set back today. I noticed that the Z screw was operating smoothly in one direction but lightly jammed and clicked when the direction was reversed. I've taken it off of the machine and now believing there must be a ball trapped between two circuits, however I've now re-packed it twice and the exact same problem persists. I think there is probably a small error where returns are machined leaving a tiny gap that allows one outer circuit and the middle circuit balls to touch. The plastic returns didn't look to clever to me.

Anyway I've contacted Fred from BST to see what he suggests.

Fred is sending a replacement, what an excellent after sales service. In the meantime I've ordered a ball nut from Cardiff for £11 that should arrive tomorrow so I can finish the machine.

My spoil board choice solved itself today. I had to take the car in for a firmware update (I kid you not) and whilst I as waiting I noticed a small plastic engineering shop over the road. Knocked on their door and the owner was absolutely flat out making perspex screens for covid protection in shops and offices, he said he was 10x busier than he'd ever been before although there was now a world shortage of perspex, well done him!

Anyway I asked about foamed PVC and he went into his warehouse and pulled out a 19mm offcut 1300x700 which he sold me for £30, perfect!
Attachment 28303

Lovely machine. May I ask which aluminium you bought from Smiths? They have loads of different tooling plate names. Alplan, certal, contral etc.

Also, You have mounted your ball screw supports directly to the steel box section. How did you make sure they were aligned and in the same plane? Surely the box section wasn't that flat?

Quote:

---

Originally Posted by Nealieboyee

Lovely machine. May I ask which aluminium you bought from Smiths? They have loads of different tooling plate names. Alplan, certal, contral etc.

Also, You have mounted your ball screw supports directly to the steel box section. How did you make sure they were aligned and in the same plane? Surely the box section wasn't that flat?

---

This time I got a piece of KASTAL 300 from Smiths, but tbh all I do is phone my local branch and ask what they have ready to go, so long as it is cast milled plate.

The rails are set on epoxy what was cast in plane. The ballscrew supports are bolted to the steel and shimmed with the ballnut hard mounted to the gantry so it aligns perfectly. I. E. fit ball nut before end supports.

Edit:
This is what I use for shimming, £15 for a pack of different thicknesses from RS, works really well.
Attachment 28306

I'm waiting on some proximity switches to arrive. In the meantime I've started to work on the control side, my enclosure with psu and drivers is finished so this is about the actual machine control

In the past I've always used MACH and an old PC, this time I'm going try using a standalone DDCS controller, no experience of these other than checking it powers up.
Attachment 28307

The are 4 axis controllers (there is a thread on here somewhere) but have no capability to slave two axis together. My solution is to add a daughter board that will both slave two drivers and handle the gantry squaring. I don't think such a thing is available so I've started to build one. The basic system is an arduino nano with appropriate opto isolation that sits between the controller X outputs and X home input and the two X drivers and two X home switches.

The (untested) algo to deal with this is as follows. Under normal operation the arduino will mirror the X signals to drivers X1+X2. When a home switch is triggered AND the machine is moving in the homing direction the system will enter a homing state. It will suppress further steps to the motor whose home switch triggered whilst continuing to pass thru steps to the other motor. Once both home switches have triggered it will signal home to the controller. The controller will then back away from the home switches, the system will then perform the opposite logic and suppress steps to the motor whose switch releases first until both switches are released then clear home to the controller and return to the normal state.

The system never generates its own steps rather it chooses when to pass through steps from the controller, the hope is that it can sit passively and just do its job without the operator or controller having to do anything special. There will be a configurable max overdrive/underdrive number of steps that if exceeded will fault the estop. It will also be possible to set offsets for each motor home position so that the axis can be squared up without moving proximity switches or targets.

The DDCS has a sensible power domain isolation between the logic and stepper outputs on one side and home/limit inputs and outputs on the other side. My system will honor this isolation, on the switch side using opto isolators on the COM+/- domain. The stepper signal side will have a ground shared with the DDCS logic side, the stepper signals themselves are just 5v differential logic so can be connected directly to the arduino pins.

I've written the code, knocked up the schematic and just made a quick PCB so will assemble and test for when my home switches arrive.

Attachment 28308 Attachment 28309 Attachment 28310

There is also a tiny oled screen and a couple of buttons for configuration.

Once I've validated it works as intended the code can be shared.

Interesting idea, but I'd look at the pulse output frequency of the DDCS and your intended stepping rate, then the frequency response of the PC817s and the effective throughput you'll get on a 16MHz '328. You are limiting the performance of the machine. At the least I'd look to replace the opto's. Pass-through performance can be easily added with adding a 74-series logic device to allowing DDCS signalling to pass-through, or to augment from the Ardy.

Of course, your mileage might vary.

EDIT: not suggesting this, but have a look at the ESP32s - particularly with the onboard OLEDs, having a dual-core processor at 240MHz meant for a very easy protocol converter with fancy graphic display (was converting RS485 to RS232 at 1.2MHz, with one core dedicated to the graphics and one just copying from UART buffers)... similar but different kind of solution.

Clock of DDCS is 500khz, my code mirrors directly off the port so low overhead compared to arduino libs. I will verify propagation delay with the scope. PC817s are only on switches not step/dir, this is also the opto inside the DDCS, i've just had a look. Agree easy to build part of it with multiplexers but I'm not sure it is required. I've got a box full of ESPs here, again not required for this.

With 16x ustepping and 10mm pitch screws you are at 320 steps/mm so 500kHz gives 1500mm/s and a screw speed of 10k rpm, we only need 1/10th of this tops so I don't think frequency will be an issue.

Interested to see how this goes. Glad to hear you're bit-banging, though as the 328 is inline with the axis pulses to the stepper drivers it has to keep up at all time (not just homing) and I'd be curious if you can hold the signalling whilst I2Cing the display - interesting challenge.

I did realise after writing that the optos were for homing switches/reflection into the DDCS, so, yeah, will be fine - I'll withdraw that.

Quote:

---

Originally Posted by Doddy

Interested to see how this goes. Glad to hear you're bit-banging, though as the 328 is inline with the axis pulses to the stepper drivers it has to keep up at all time (not just homing) and I'd be curious if you can hold the signalling whilst I2Cing the display - interesting challenge.

I did realise after writing that the optos were for homing switches/reflection into the DDCS, so, yeah, will be fine - I'll withdraw that.

---

Yeh the I2C is the dubious part, albeit it is done in hardware, I haven't coded that yet, if it gets in the way it can be omitted. I just usually add them to all my boards these days as I have a drawer full and they only cost a few quid. In this case I had disable the uart to use the rx/tx pins for outputs so without I2C there is no way to get any debug out whilst testing. I did this so I could get a full 8 bit synchronous port write to the drivers rather that making multiple consecutive writes. Every true IO pin is in use ;-)

A very interesting project, especially as I was well on the way to completing something similar when I decided to put on my grown up trousers and upgrade my LinuxCNC version instead.
My version planned to have a board in the middle of the parallel cable from the PC to the BoB and used logic gates to select whether stepper control pulses were fed from the PC or the Arduino Nano (you just can't get away from them can you?!) and limit switch pulses fed back. When you selected gantry squaring the PC was left thinking the machine was idle and the Nano produced drive pulses which were gated to the two motor drivers by de-bounced and latched signals from the limit switches. Once that sequence was complete control was handed back to the PC for normal homing.

Under normal operation the stepper drive signals were only going through a couple of extra logic gates so no data speed problems were added and the timing of separately stopping the two motors was to be done by the discrete logic with no software delay to add timing errors. The nano was there to select who had control, generate step pulses at the two required speeds and set the direction of movement for the whole sequence. One of it's great weaknesses was that it relied on mechanically adjusting one of the end stops to set the squaring, though it would only have needed a few extra lines of code and maybe a gate or three to fine-adjust from the software.

I designed and built the whole thing including laboriously cutting and buzzing out both ends of a parallel cable but never actually put it into use having realised in the mean time that an upgrade of LinuxCNC was not as difficult as I'd first thought.

I look forward to seeing your video of the beast in action.

Kit

Hi Kit,

I still have a PC with a MACH license on it to revert to if this doesn't work. Just want to try and remove all the associated cruft (keyboard, monitor, desktop) I usually have around the machine to make it work. I've now drilled and assembled the board, tested connectivity then lacquered the copper side (just in case it is a keeper!).

Attachment 28311 Attachment 28312 Attachment 28313

I need some more solder paste mine is about 5 years past its best and someone has bent the ends of my nice tweezers meaning putting those resistors on was a PITA.

Cheers, Joe

It's drilling the bloody holes that's always the pain isn't it? I use SMDs as much as possible and bought a vacuum placement thingy off eBay a few years back. I keep the solder paste in the fridge, especially as inside my shed can get up towards 50C around Christmas and new Year.

I hope it works as you want.

Kit

Drilling?, if only we had a machine that could do all that for us :-)

And solder paste?, I've thrown more out than I've used, with it's damned stupid shelf-life (fridge or not...and a Northern UK climate). Right now 0.4mm rosin cored solder will do me and a decently "Large" iron tip. Oh, and flux... plenty of flux.

But, yeah, DevMonkey - hope it works and please report back. I need inspiration to continue my DRO/Pi controller for my mill.

Quote:

---

Originally Posted by Doddy

Drilling?, if only we had a machine that could do all that for us :-)

---

You have. It's called a CNC router. The trick is making the machine and your ability to register a pre-etched circuit board on it accurate enough. Something I haven't even atempted yet.

I do like using the solder paste. That and a gas soldering iron with a hot-air head makes hand-soldering SOIC chips nice sand easy and the 1.27mm (thats 0.05 inch for our 19th century readers) pad spacing is reliably makeable using the toner-transfer method for DIY circuit boards....without holes!!!

Kit

Quote:

---

Originally Posted by Kitwn

You have. It's called a CNC router. The trick is making the machine and your ability to register a pre-etched circuit board on it accurate enough. Something I haven't even atempted yet.

I do like using the solder paste. That and a gas soldering iron with a hot-air head makes hand-soldering SOIC chips nice sand easy and the 1.27mm (thats 0.05 inch for our 19th century readers) pad spacing is reliably makeable using the toner-transfer method for DIY circuit boards....without holes!!!

Kit

---

For crude through hole boards with thick traces like this one I use the toner transfer method with a £50 Brother laser printer from Argos, good results and takes <15 mins to make a board including etch and drill, I drill by hand on a tiny toy CNC machine. This process works well for double sided boards as well.

Attachment 28335

For higher resolution boards I use a CNC photo method on the same toy machine. Spray the copper with black paint then use a low power UV laser in the machine to isolation route it. Wash off the ablated paint then etch. If the board requires any drilling and/or is double sided I use a fixture so to re-locate it after etch. This process takes a bit longer but I can achieve 0.2mm features/tracks.

Anything more complicated with >2 layers gets sent to China, even if just a proto. All the SMD stuff gets soldered with hot air, agree very easy if your solder paste isn't like gritty chewing gum.

Quote:

---

Originally Posted by Doddy

Interested to see how this goes. Glad to hear you're bit-banging, though as the 328 is inline with the axis pulses to the stepper drivers it has to keep up at all time (not just homing) and I'd be curious if you can hold the signalling whilst I2Cing the display - interesting challenge.

I did realise after writing that the optos were for homing switches/reflection into the DDCS, so, yeah, will be fine - I'll withdraw that.

---

Hi Doddy,

So it turns out if you ditch the standard I2C/TWI wire.h library and display driver, crack open the 328p datasheet, write your own crude framebuffer and I2C non-bocking state machine that operates directly on the TWI registers you can tick the state machine at a 2-3us per tick penalty. Will put it on the scope and see if this is sufficient.

Main loop becomes:
1. Mirror stepper/home signals with squaring logic.
2. Update framebuffer if needed (using a 5x7 font this is 5 memory reads/writes per character)
3. Tick I2C state machine.

I could remove the framebuffer to make it a bit faster if required, this will effectively amortise the font copy part at one byte/I2C tick.

Cheers, Joe

Quote:

---

Originally Posted by devmonkey

Hi Doddy,
write your own crude framebuffer and I2C non-bocking state machine

---

"Crude"?, I prefer to think of this as "fit-for-purpose".

You know as well as I do the 328 is just one of a thousand different micro controllers that pollute the oceans and the Arduinio packaging is just a convenient way of throwing that chip into an accessible form for stupidly low cost - makes them a sensible solution for even the most simple of problems. There's no reason to use the Arduino libraries unless you're concerned about cross-platform compatibility or speed/ease of development (sacrificing code base, efficiency and real-time performance). I think you'd agree programming at the register level (I gave up at the instruction level many decades ago - I prefer to have a compiler at least) adds a level of complexity and care, but if that floats-your-boat (and it does, me) then hack away to your hearts content. At least you can be confident that you know exactly that the uC is behaving exactly as you want it to.

So I've just scoped the DDCS. An axis setup as per my machine, 1610 screw, 16x ustepping -> 320 steps/mm, max speed 10000mm/min.

As expected step frequency is ~53kHz, half period of 9.375us. Nyquist/Shannon tell us that this needs to be sampled at a minimum of 2f, i.e. 106kHz for perfect reconstruction. Puts an upper bound on my mirror code latency of ~9us, so to be safe say 5us. This is 80 cpu cycles, will it fit?? Should do, got to wire it in and test.

DDCS defaults set dir 7000ns (7us) ahead of step therefore code running at 5us will always see this. EM806 only requires 2.5us but configuring DDCS to deliver this will risk breaking the contract woth EM806.

Interrupt on change of state? Your time-sensitive processing is in the discrete control and not the I2C if you're using the TWI registers.

Quote:

---

Originally Posted by Doddy

Interrupt on change of state? Your time-sensitive processing is in the discrete control and not the I2C if you're using the TWI registers.

---

Yes it is possible you can register for toggle interrupts on any of the IO pins, trouble is my code is written in arduino/cpp at the moment, this means there is a stack to park so there is some overhead from servicing the interrupt. If the current polling approach takes too long will have a look.

So here is a baseline, the basic axis mirroring with auto squaring, but without I2C. Lag is 600ns after a lot of messing about I just stuck the logic in a lookup table and deleted all my code ;-) Yellow trace is step output from DDCS running at 10000mm/min, blue trace is the arduino step output, lag is shown in 'dt' bottom right.

Attachment 28337

You're not going to get better than that!

Quote:

---

Originally Posted by devmonkey

after a lot of messing about I just stuck the logic in a lookup table and deleted all my code ;-)

---

Hahaha love it. Sometimes 'ugly' solutions are just the best.

My dual motor homing / squaring board is now working with the DDCS. Video clip below. DDCS was told to 'home' the X axis:

1. Each motor stops when its respective home switch is activated,
2. Once both home switches are activated the board signals home to the DDCS,
3. DDCS now backs X away from home searching for switch release,
4. If one motor's home releases before the other this motor is stopped until both home switches are released,
5. Once both home switches are released the board clears home back the the DDCS and DDCS zero's X co-ordinate,
6. DDC now moves X a pre-configured distance away from home.

This logic only activates when a home switch is hit and the machine is travelling towards home. If travelling away from home either switch hit signals the DDCS (i.e. it behaves as a limit switch).

The only downside of this is that in normal operation driving the machine past home in the X- direction requires both switches to activate before limit- is signalled. This is not a problem unless one of the switches has fallen off.

Attachment 28364https://youtu.be/dyFAdsQIpxA

Call me stupid, but what exactly do you adjust to make the gantry square once the whole sequence has finished? What is the purpose of step 4? Do you have separate cotrol of the two X distances in 6?