Updated - Conversion of milling machine to CNC

[MeJasonT]

Ok some background so you can understand me.
Im a farmers son who hates spending money unless its absolutely necessary, you can never fix problems by throwing money at it.
and every thing i make usually comes from my stash of collected items or items gifted by friends (ie stepper drives from paint mixing machines).
Space is a premium but i have still not acquired the knowledge that my workshop isn’t a tardis.

So progress on the machine this week.

I continued working on the coolant system, I had to buy a coolant pump. The old pump from the machine had flow issues, it was like the flow rate from a urinating toddler - yes it dribbled more down itself than succeeded in aiming at the work piece.

A control panel was built which provided both flood and mist coolant options (Gcode M7 & M8), the plan was that with either option the flood coolant pump would run. option 1 would simply divert the flow directly to the work piece and option 2 supply coolant to the mist via a venturi so that i wouldn’t need a separate reservoir.

It failed

The (pneumatic) solenoid valve i was using to divert coolant flow just leaked like a sieve and its output was negligible. Before an academic jumps in here, you can use pneumatic and hydraulic parts in either system its the same only the media is different, air, water, oil. Pressure is a concern pneumatics wont work well over 120 psi and hydraulics well pressures in excess of 3000 psi is normal (Im trained to work with high pressure systems exceeding 10k psi).

My conclusion other than a leaking seal is that the 1/16 hole in the valve is more of a hindrance than anything else especially added to the fact i was using 1/8 hose.
The fitting size on the pump was a bit of a giveaway at 1/2".





im now in the process of "rip up and retry" so back to 2 separate systems with the mist having a reservoir, i could automate the top up of the mist circuits reservoir but it may not be worth it.

The machine panels have started to go back on and the probing circuits have been tested (both tool height setter and active probe)

The build has some limitations which i hoped would iron out, Pathpilot by default has a feed rate of 2 meters min, my setup will only allow 500mm/min which is fine for what i intend to use my mill for (im usually between 75 and 250 mm/min for most operations) But oh it would be so nice to get 2mtr/min. i suspect hat the bed ways on the mill being old metal guide ways is too much friction for the stepper motors to deal with. we are not going to put linear rails on it. Butchering such a fine old piece of British industrial heritage ( Elliott's in London) is just a step too far.

I have tinkered so much inside PP's configuration now I probably know as much about tormach's as they do.

A young gent as acquired my assistance in converting his mill, he had approached me as he also was wishing to covert his machine and had also acquired pathpilot. He has a Denford Desktop mill which i have started to call his Etch-a-sketch. His configuration also had to be dialled down, which is just as well as this thing would have thrown itself off the bench at 2 mtr/min.

At some point in the next week in going to grow a pair and try my first cuts, before then i need to do some serious clip recording and images taken to add to this post for you guys/gals.

Always been someone with more than one project on the go, this time is no different.

I have been quietly chipping away on three side projects but very much related, A Power Drawbar, an Auto Tool changer and a 4th axis.

so in the intro to this post i mentioned i liked building things from available scrap and not spending dosh. well.

The 4th axis is making its way slowly but surely out of the scraps box and into fruition

[MeJasonT]

Im also starting to accumulate bits for the ATC (Auto tool changer)



A schematic has been drawn up and a board etched and assembled for a controller



The board is ok but my firmware leaves somthing to be desired.
I have made the aquaintance of a good chap in Glasgow, Scotland who is building electronics for interfacing with Pathpilot David Loomes. If i were the spending type of guy i would have purchased a controller from David. We cant advertise things for sale on the forum but im sure pointing people in the direction of good people isnt against the rules - McMaster-carr, Sears and Lowes get mentioned often enough.

David Loomes CNC Products

[MeJasonT]

The 4th axis started off life as a paint mixing/shaker machine



To be fair my manual machining is as rough as a bag of spanners so i tend to manipulate to get what i want. In this instance i butchered a part from this old shaker as it was obviously designed and built to take alot of abuse and be hold a reasonable accuracy. I chose the "I" beam as its base as it was a solid frame which again was fairly square so that i didn’t have to try fabbing something to farmer boy precision.



The hub and chuck have been manually spun by hand on the table and the runout measured, It was 2 thou by pure luck. both top and side on to chuck.
The weld on the back plate looks like i used a mastic gun but i can assure you its not as bad as it looks, a flap disc will soon have it looking pretty. will it work as a basic 4 th axis, i think so. Its £500 for an asian import of 3k+ for a manufactured one. Did i mention not spending dosh.
all the interconnecting parts like the Gearbox shaft/cog and the drive cog in the end of the spindle were turned on the lathe to a reasonably tight tolerance.
As the spindle was already part of the shaker it and its mate were both matched. Using the "I" beam gave a flat surface on which to mount the hub, all things being equal that should keep the hub running true. The plan is to use the I beam on its side for 4th axis and at some point in the future the trunnion could expand to a 5th axis by mounting the 5th onto the back plate of the fourth and using the i beam stood on its end (shimmed up accordingly etc). Getting the miller back up and running is going to help immensely.

You will notice that the recycling theme extends to robbing parts for the ATC as well.

[Saltfever]

When you first mentioned GRBL I thought you were talking about GERBIL (a language I was exposed to in the mid-90s at work). Totally different ballgame and now you know my depth (or lack of) knowledge about Arduino stuff. While I have been around NC since the 70s and now CNC I have operated more or less in the user world of functioning systems and not the frustration (at least to me) of the developer world. I can see where development could be exciting, interesting, stimulating, and frustrating all at various times. It is all about the journey for all of us and the interest factor can change with the venue. I have always hated beta testing yet others delight in the challenge. However, I am now considering a new project and the knowledge gained from solving a problem offsets the frustration in time lost! But time is still so precious. Thus I am looking for fairly mature control software to minimize the time consumed in bug fixes. When you mentioned PathPilot it seemed a likely option compared to the others mentioned. I will be looking into it further.

You mention the speed of PP is limited to 2m/min. That might be acceptable for feed rate in many situations but it is pathetically slow for rapid traverse (non-cutting movement) for positioning. The minimum I would consider for positioning would be at least 10m-12m/min. For example the Haas mini mill is 30m/min!

What is the speed limiting factor in PP? Amps, accelerations, processor speed, other . . . ?

All the best..

[MeJasonT]

I think the limitations are down to hardware, until recently they were using stepper motors which averaged a max rpm of 1000 that is probably the source of the limitation. now they have switched to servos i have no idea what the upper limit would be. Its probably going to be dictated by the rigidity of your machine, as you probably know the industrial machines are constructed around a casted base weighing a few tonne. Accelerating and stopping a bed weighing 100kg on a solid machine would be child’s play but doing the same on a diy hobby machine it would end up with the machine lying on its side on the floor. I would guess that power has a lot to do with it as well. When you get into the larger servos you start looking for power units outputting 130v dc for example when you start looking at amperage they look like 10+ amps per axis the power units for the servos start to become hugely expensive for most diy projects. In order to obtain uninterrupted power supply the machines end up powered by 3 Phase mains supply. Everything gets big and expensive very quickly; Servo drivers are in the region of £1000 the motors are not far behind. You begin to realise how cheap a second hand machine really is when you start to figure out the cost of all the components. Simply take a look at the price of Tormachs 4th axis or their ATC and you begin to get an idea for how much the machine would cost to build.

The following link may be of interest to you


https://www.pearl-hifi.com/06_Lit_Ar...Design-Use.pdf

I tried everything to attach the pdf, zipped it up you name it but this text editor wouldnt allow me to upload it - so i have posted the link where i found it. Apparently its still available on their site.

[Saltfever]

Thank you for the Tormach link!

It’s a good technical discussion of the trades considered in designing their first machine. The meat of the doc for a DIY group is the first 104 pages that deals with mechanical engineering described in layman’s terms, physics, cost analysis, and decisions, resulting in that design. While the doc is 239 pages long the last half is devoted to their injection machine.

The value is that all the fundamentals of creating an electronic motion controlled, mechanical apparatus, are wrapped into the doc. It’s like having a beginner's Engineering 101 bible at your side to help you with your project decisions. It certainly touches on so many aspects that are germane to this forum that Jon should consider making it a stickie. Anybody building a machine should find it useful in the design of their own apparatus.

Young modulus hasn’t changed in the last few hundred years. Designing an adequate mechanical device is usually straight forward and in the case of a one-off DIY worst case, even over-design doesn’t hurt the pocket book!

The problem however is the rate of change in the solid-state electronic, software driven, motion controlled world. If that is not your forte one can fall behind rather quickly. For help in that area reviewing the Tormach thought process could be quite useful.

[MeJasonT]

I couldn’t agree more, having read the document Skimmed through it as engineers do) I did find it useful in the fact that I find motor and gearbox theory a magic art. There are so many different disciplines in engineering which allow one to gain access into different areas of the industry as a career. Oddly enough i started in electronics, i blame the Royal Navy at the start of my career for turning me into a clanky. I have covered by background at great depth on other posts, so now i have to stop jawing and stay on topic. You are very correct about the 101 Bible aspect of the technical note, you don’t usually get companies willing to share or make public their findings during the design process. Its that fine line between looking like a pioneer or showing the world any incompetence you may posses. It will be the dilemma many makers face before daring to share their concepts on this site i would dare argue. Im pleased to say they pulled it off with this white paper. The huge zoo animal in the room is always going to be cost, I was fortunate in the fact i was given a few stepper motors and drivers - or unfortunate as that is the limitation im stuck with. I would still use steppers nema 34's probably and enough torque to drive a truck. See i read it .lol. I know that the stepper drivers I poses are my greatest enemy but its always an option to upgrade them in future if the machine is build in a modular fashion. Blow me if they are now starting to teach such stuff and finding ways of building control panels with obsolescence built out of them (Machine Design magazine this month https://www.machinedesign.com/mechan...373J3983623F6C)

alas engineering has come full circle, we are back with pre Victorian mentality where academics rule and those creative thinkers are shunned as being uneducated and their ideas laughable. Its funny how they were the founders of the industrial revolution, like my hero Brunel who technically was a high school dropout.
Academics fear clever people, 80% soldiers are non academic being that officers are of degree standard and yet the boffins need the scroats to fight their battles.
Long live the 3D printer, makers and open source revolution.

[MeJasonT]

The 4th axis received a base plate yesterday. I wouldn’t normally share images of my welds to show off but today is different, I was vey pleased with them
just using a Bohler E6013 mixed metal universal rod 2.5mm and 80 amps. I cheat i use the amperage im told to use on the box the rods come in.
simple fillet weld where the plate was "V" prepped to make the channel.