MadMachine - let the fun begin

aubrey · February 16, 2018, 10:41pm

Hi all.

This is where I will be posting specifically about my CNC machine project.

I’m a retired artisan and often somewhat “weird” (hence my screen name) who had ALL his machinery and tooling (built up over years) consigned to a couple of skips by an idiot stepson.

BUT I NEED MACHINERY IN MY LIFE!!!

So I’m building myself a CNC mill.

It was difficult to decide which way to go as I literally don’t have tools - except for some spanners that fell into the spare wheel well of my car and many designs were way above the capabilities of what I have but about a week ago my luck changed and I was able to buy a 2nd hand I3 printer - so its going to be a MPCNC.

In South Africa we have a problem in that buying something from “overseas” sounds cheap but once you factor in delivery fees (a long way) and insurance (VITAL HERE) it is often simply not affordable.

So in many cases in my career I have had to use whatever is available locally and “make a plan” to get it working and I will probably be making MANY of these “plans”.

So I am going to post my trials and tribulations here.

I may be posting in the wrong place and if I am then if a moderator can let me know and if possible move it to wherever it should be.

In the meantime I need to get a couple of spools of PLA, a piece of pipe to test fit the parts to see if the dimensions are good and follow up on a lead I got for 2nd hand motors out of industrial printers at an affordable (for me) price.

And maybe tonight I will try printing my first part to see what is going on.

So, wish me luck as I will probably be needing it and to the guys who are more knowledgeable, please keep an eye here as I will definitely need more than just a little mentoring along the way.

Eventually the machine will be used to produce wood carvings and sometime in the future it (or its sister/clone) will probably get an A axis - and maybe the router will get more degrees of freedom but we will cross that bridge if and when we come to it.

aubrey · February 17, 2018, 4:27am

So using the calculator at http://stevecd.github.io/mpcnc_calc_react/ I have gotten the following measurements:

Required swept length - 90 cm (900 mm)
Required swept width - 90 cm (900 mm)
Required swept height - 45 cm (450 mm)
Total pipe (25 mm od) required:
3 off 927.9 mm (X axis)
3 off 927.9 mm (Y axis)
2 off 470.1 mm (Z axis)
Total pipe 6507.8 mm cut as above

Z lead screw (M8x1.25) - 1 off 468.1 mm
Toothed Belt
2 off 945.7 mm (X axis)
2 off 945.7 mm (Y axis)
Total Belt 3782.9 mm

Now I don’t know what the values in the “Settings” section is all about so I left it at whatever the default values were - I suppose I’ll find out at some time or another!

My quest for 2nd hand steppers was a bust AND I had a run-in with the idiots in blue - long story, later but it cost me 500 bucks because something was 8 inches away from where it is SUGGESTED that it be placed - Year end coming, chasing the ticket quotas is my guess…

Anyway, going to try to print the smallest pieces - the ties that help prevent the belt roller bolts from misaligning.
Why? Well this will be the 3rd time I’m starting a print on the I3 so screw ups are bound to come my way - rather on a low volume print than a high volume print…

Wish me luck.

vicious1 · February 17, 2018, 8:10am

You are in the right place buddy, welcome. Can’t wait to see how you progress.

That is an old one from a previous design. Do I still have it linked somewhere? This is the current one.

Ohhhhh ya gotta learn to suppress that urge…99% of the time it just makes it worse. Can’t teach an old dogs new tricks I guess. I did watch a friend lose her mind on them once, they apologized and left. I felt like I was in the twilight zone, it was glorious.

jeffeb3 · February 17, 2018, 8:37am

I am expecting this to be a fun build to watch. No pressure.

I found this site to be pretty useful for describing what is failing on a print:

http://reprap.org/wiki/Print_Troubleshooting_Pictorial_Guide

3D printing still requires a lot of skill, so don’t be afraid of a few dozen failures. The first layer is 10x more critical than you think. Hopefully you’ll get a hang of it before your stepson sells the printer.

aubrey · February 17, 2018, 3:35pm

Update:
Its 1 am local…

The first part print was a disaster!!! Filament was not feeding off the spool so it started printing air.
I hunted down a simple clip over spool holder that would fit my machine, sliced it and began setting the machine up.
I suspect that the heated bed may be warped as the adjuster at home position needs to be loosened more than it can be (the nut is no longer in contact with the surface).
Eventually got it more or less set up.
Then the filament started to grind - cleaned it all out.
Then I noticed that the extruder fan would not switch on - the other blower did…
Unwound the spiral cable holder - THE DARN THING WAS NOT EVEN CONNECTED!!!
To make matters worse all the places to plug wires in are used - NOW WHAT?
So I doubled up on the switchable fan connector. evidently the extruder fan runs when the machine is powered up but this is better than nothing!!!
Then I noticed that the positive side of the power supply cable pair was loose - It had just been stripped and twisted - out with the soldering iron and tinned the wires - heated bed power wires too.
Put everything back together - Evidently the board was changed at some point BUT its a crappy design, no space and mounted right up against the frame (the side with all the wires - USB port on the other side so can’t turn it over - last resort was to mount it with 2 screws but moving the right hand holes to where the left hand holes should be and wrapped everything up again.
And the little display on the machine frame does not work - needs to be updated (firmware) I had been told.
Started the spool holder print - forgot to replace the silicon extruder heat shield and the workpiece cooling shroud but let it go on.
So far so good - THEN just before it was about to do the closing on the top layer it lost communications…
Hope THIS is not going to become an issue!!!
So the first port is a dud - but its usable…
Put all the missing pieces back, cleaned the bed, rubbed Pritt (glue stick) all over it again and started on the 2nd one.
Then remembered that I had not checked the bed levels again…
Oh well, too late - but the bottom layer looked OK so I let her run.
It has just started to close the holes and when its done its sack time…
While I was typing this it finished - top layer too this time.
It does not look too bad but I think that just maybe I’m not pushing out enough material as you can see the boundary lines between the passes - little creases but not holes or drop-aways so I must remember that.
BED - NOW - Goodnight…

aubrey · February 17, 2018, 11:35pm

Oh don’t worry, I’ll get the hang of it all right!
As to the idiot stepson, his biological father gave his mother her first failed marriage and he gave her her second. I’m in Cape Town and she is “captured” in a situation and up in Johannesburg.
Hell has a special place for idiots like him - and I have an aluminium baseball bat (normally used to “keep the peace” in my boarding house) that I would dearly like to take for a test drive.
Mr Moderator: what is the policy here for placing links to off subject material?
In any case, I am attaching an image of a sculpture that had me crying like a baby for over an hour the first time I saw it - It lit a fire that is sending me down the path I am on as far as this project goes - I HAVE to make my own version of it to pay homage…
What the heck, Here is a link to more info

aubrey · February 17, 2018, 11:44pm

OK - when I posted the last entry I got a “Gateway not found” error (not uncommon on a Wordpress site) so I hit the reload button - then I got this:

ERROR: Duplicate reply detected; it looks as though you’ve already said that!

But I can’t see a “Yes it is duplicated” button to dump it so I’m overwriting the duplicate post (I hope!!!)

aubrey · February 18, 2018, 12:34am

WHAT IS WRONG WITH THE UNIVERSE? ;>}

Just when I thought it was safe to “go back in the water” I discover that the 57 mm hole in the spool allows the spool to drop on the axle which allows contact between the rim of the spool and the machine frame.

I also show the “failed side arm print” in the photo - the infill sure looks “nice” ;>}

So off to Thingiverse and I found THIS

A bit “adventurous” for a newbie?

You bet!!!

But that is how you learn!!!

aubrey · February 18, 2018, 2:37am

First print: unmitigated DISASTER !!!

The piece around the edge didn’t even bother to adhere.

So it was started again and while it was in the final warm-up stage I hit the bed with the Pritt stick in a VERY liberal fashion.

Well, the bit around the edge stuck OK.

HOWEVER the first layer was all over the place.

The individual lines were not lying flat and parallel to each other, there was gaps all over the place - it was a mess!!

And in the center it looked like the wall of the hole was drooping off to the side - NOT GOOD!

Almost aborted the print when I had a brainflash: IDIOT!!! There is a thread in the middle of this thing and that is the start of it!!! (Carefully removed the cursor from the “Abort” button)

So I increased the flow rate to 106% (sounded like a nice enough number) and on the 2nd layer things started to look better, everything was smoother so I let it run and had breakfast.

Thinking about the crappy first layer I came to the conclusion that something was definitely out so I sat eating and thinking (sometimes I multi-task - what can I say!)

Then it hit me: The problem is with my setup procedure!!!

I home the machine, lift the Z a few millimeters, Go X0 Y0, home Z again and use a piece of paper as a feeler gauge to adjust and repeat on the other extremities and do this a few times.

BUT THE BED IS THE THICKNESS OF THE PAPER BELOW MACHINE X0 !!!

So the first layer is not being extruded “into” the bed but rather above it!!!

That means bad adhesion and 1 layer’s worth of material having to try to take up the volume of its own layer PLUS the layer represented by the thickness of the paper !!!

I was given Repetier Host v2.0.5 with the printer and I remembered that there ar some pre-defined and programmable functions available (and the print seems to be doing well as I write - SEXY thread!!!) so I am going to use 4 of them for setup purposes.

Basically, I will home the machine and each function will move Z whatever the thickness of the paper is above Z and then to the X and Y coordinates of the corner I’m busy setting up.

Then repeat until I’m happy with the bed leveling.

That way the first layer will be extruded INTO the bed, ensuring better contact and “stickiness” to the bed and the extruder nozzle will be able to "wipe it to the correct height which will squash excess material a little sideways into the material next to it.

AAAGGGHHH !!!

While writing I went to have a look at the progress - IT STOPPED EXTRUDING!!!

Something is jammed up SOLID - I can’t even get the filament out!!!

So it looks like I will be getting intimate with the extruder head.

As I suspected, the first layer is a mess - if you hold it up to the light you can see daylight in places so after I troubleshoot the extruder I’m going to use the new setup procedure and try again.

I’ll be back later.

aubrey · February 19, 2018, 8:31pm

While I have been having an “enforced” period of down time due to technical reasons I have been doing some thinking (DANGEROUS!!!)

Anyway I got to thinking about the design of the extruder head and I am wondering about a couple of things.

Firstly, on most of them that are available commercially I notice that the part that actually engages the filament and forces it into the “melt chamber” is an actual gear.

Really? - The shape of a gear tooth is designed TO BE A GEAR TOOTH!!!

It is NOT designed to be a traction device and is in fact a pretty lousy shape to be a traction device IMHO.

It has rounded edges that are IDEAL for creating slippage at the least provocation.

Why is this? Or is it simply using whatever is at hand to sort of get the job done?

The next thing I was wondering about (mainly because it is the actual thing that is causing my down-time) is the piece of M6 threaded stainless rod with the hole in the middle that in my case has a PTFE liner which is actually the problem in my case - clogged and stuck solid!!!

What is the use of the liner - it is simply another point of failure that has been introduced into the system!

Someone told me that it is a legacy thing from the days when the “standard” filament size was over 3mm and when things moved to the 1.75mm diameter they simply put in a “spacer” to make up the difference.

But is the PTFE not then also acting as a heat barrier and reducing the efficiency of the heat transference from the metal part to the plastic that is being melted?

So would a design that is closer to the 1.75mm standard not be better?

It could be reamed to a smooth and tapered bore to help with extracting blockages.

Like production injection molders (which are cyclic as opposed to 3d printer heads which are continuous) it is all about hydraulics.

In printer heads the “ram” is the filament entering the enclosed system through a fine tolerance orifice and once it is inside, melts and becomes part of the “fluid” that is extruded out the nozzle.

So would a smaller (not too small) “melt reservoir” not work?

If it was not semi-insulated by a chunk of PTFE then the transference of heat from the SS part to the material would be faster and everything would happen more efficiently.

I have also been lurking around YouTube and was particularly interested in a number of videos about “making your own filament” and the problem that everyone seems to be having is keeping the diameter of the extrusion to within a very fine tolerance around 1.75mm and this is understandable given the fact that the filament needs to “seal” the orifice through which it enters the melt chamber so that the required hydraulic pressure is maintained which forces the melt out of the actual print nozzle.

In extreme cases the filament can jam up going in (filament too large) or not provide enough sealing (filament too small) in which case the melt will be forced back up along the path that the filament is entering the melt chamber and solidify, thus causing a jam up that way.

So as it stands now, the diameter of the filament as it comes off the spool is critical to the success of the whole process as it provides a seal (or leak) to the melt chamber AND allows the accurate calculation by the software of the volume of material that is being extruded out the print nozzle.

But is there not a way that this can be worked around?

Would it not be great if a printing head could accurately and “clog freely” work with a filament of minimum diameter 1.75mm and maybe up to about 2.00mm?

Maybe I should give some thought to this issue while I am eventually waiting for the components of my MPCNC to print…

And while I am waiting for the MPCNC mill to finish the jobs it will be running.

vicious1 · February 19, 2018, 9:23pm

Many things have been tried. Have a look at the extruder wiki. There are hundreds of attempts. The gear needs to pull equally in both directions without cutting the filament in half and engaging well. As long as one of the gear or the bearing has a grooved surface it is actually a great option. Some are even geared on both sides.

PTFE makes printing with PLA a breeze. I have throats with no liner but the settings are more finicky. The reason yours went bad is it was overheated and baked the PTFE dry. No more than 260C and even then mine only last about 1 year. Luckily a set of 2-3 new liners a about $2.

The extruder will handle all sorts of size variances, the print will look like crap though. But yes again this has been solved as well, you can get a device that measures the average diameter as it goes in and adjusts the feedrate accordingly. As for making your own filament…not that easy. Plastic can only be cooked so many times, and you always need new fresh material and nowadays it cost just as much a laser calibrated professionally made filament. Not to mention the drying that needs to happen.

aubrey · February 20, 2018, 7:59am

Hi Ryan - points all well taken!

I was kind of ranting a little bit - I tend to do that hence the name.

While I was wondering around in my boredom I did come across some guy that was experimenting with using something like a drywall screw to move the filament.

His argument was that there would be no slippage due to multiple points of contact.

Anyway, thanks for the reply

aubrey · February 20, 2018, 9:23am

I have come to the conclusion that I bought someone else’s pile of spares!!!

A spare frame, spare bits and pieces - got it all working sort of - and sold it to me.

Why do I say this?

Well, there is not a single washer on the Alison (thats her name now - she can be a b1tc# but if you treat her well she works…) and there are an assortment of miss-matched cap slotted and cap screws all over the place.

And I’m not going to even talk about the wiring…

Another big thing that I noticed is that the fan was in the wrong place!!!

In the photo I have mounted the motor to the right of where it was and drawn a line to show that it was cooling the extruder gear and the stepper motor but nothing was getting to the cold end (?) because the bottom of the extruder assembly was blocking any air movement to that destination.

[attachment file=53429]

So the Stainless Steel “melt chamber” was getting hot and the “cold end” was too which supports what Ryan said earlier about frying the PTFE liner.

But the show must go on - I need to get my spool holder printed so that I can carry on with the MPCNC parts but it looks like I have to find a cooling shroud for the “melt chamber” to print first - and maybe some harness clamps to control the rats nest of wires rubbing all over the place and bending where maybe they should not be bending…

But in the meantime I cobbled together a bit of a Heath Robinson contraption to get the air from the fan to where it is needed.

Here it is - its NOT pretty!!!

[attachment file=53430]

And a view from below…

[attachment file=53431]

But to carry on trying to get Alison to print anything without some sort of cooling down there would be a waste of time and filament - so it is what it is!

The first thing that I desperately need is to finish off the filament spool mount and that means printing the universal spool holder.

The other day printing the small part of that assembly lead to the blockage disaster but now things that should be cooled are being cooled.

So I set up the job (using a modified setup procedure that seems to have gotten rid of the first layer problem) and printed the nut part again.

That came out pretty nice in my inexperienced view - seems that the nozzle is sort of smearing or smoothing out the line if material as it is laying it down which begs the question “Is that a good thing?”

I think that maybe it is because it is forcing the new layer into hard contact with the previous layer which probably means good layer adhesion.

Or maybe it is NOT!!!

I started the “bolt” part of the assembly and made what is probably a rookie mistake - I didn’t flip it over so that the large part was at the bottom.

Which means that I started printing the narrowest (43mm) part and the widest (80mm) part would be at the top.

Well, something happened - the nozzle somehow climbed sideways off of the piece, the steppers skipped a whole bunch of steps and proceeded to print in mid air…

Fortunately enough of the part had already been printed so that it would at least be functional!!!

[attachment file=53432]

On the left is the first failed nut, center is the “good” one and on the right is the the “failed but usable” bolt piece.

I’m using Repetier-Host with the extruder set to 190 degrees, the bed set to 50 degrees and I’m using glue stick. Even though there was a collision between the piece and the nozzle, nothing broke free from the bed.

EDIT: THIS IS INCORRECT !!!
However the “Feed Rate” is set to 100 and I think that maybe I should dial that back after the first few layers have been printed and the infill starts happening.

I THINK THIS IS CORRECT !!!
At some point or other I must have enabled “Easy Mode” which hides the “Flow Rate” setting. I am going to have to rectify that and see what it is set to and play around from there.

Anybody got any thought?

jeffeb3 · February 20, 2018, 10:27am

What are you using as slicer software? I use the prusa modified slic3r:

Ryan has some settings for the normal slic3r, and they are a very conservative, but good starting point:

https://www.v1engineering.com/import-extruder/

Generally, you want the first layer to be slower, and the outside layers to be slower, and the infill to be as fast as you can get it without making mistakes.

The first layer is critical. When it cools down, if there is a fat part and a skinny part, they will cool at different speeds, and cause warping on bigger parts. The first layer should be smooshed a little, but not really smooshed. I don’t know a better way to describe it with words, but if it’s not able to extrude the plastic it’s trying to, or it’s smooshing it enough that it’s flowing out the sides, then that’s too much smoosh. When you get it close, a smooshed part will have bell-bottoms or elephant feet on the bottom because the first layer will be too smooshed.

Are you printing PLA? I really like printing on blue painter’s tape instead of gluestick. But a lot of people like the glue stick too. When you get better at getting the bottom layer right, you can move to PEI, and be really happy. But if you smoosh into PEI, it will really stick, and you’ll ruin the PEI getting the part off. PEI sticks to the scent of PLA. If your bed isn’t flat enough for the larger parts, you can add a brim. It’s a hack, but it does help with the warp.

Here’s some more info that’s pretty good, to keep you busy:

http://reprap.org/wiki/Triffid_Hunter's_Calibration_Guide

aubrey · February 20, 2018, 10:18pm

Hi Jeff - I’m busy enough but I will get to all the links.

I’m using Repetier Host to run everything, Cura to slice and PLA.

Look, I know that it is early days on the printer but at this point I need to get the machine and myself to the point where I can start printing the bits and pieces for my MPCNC.

Some of those parts have got MASSIVE print times and I will probably be able to hunt down a lot more info on the net while I am waiting…

I am actually pleasantly surprised that of the 8 prints I have run up to now.

I have had 4 “fails” but 2 of them were “usable”.

My biggest shock was seeing just how “well” Alison performed even though she has a lot of issues (some pretty serious) that I still need to sort out.

Once that is done and I know more about setting up for a print (I need to get a dial gauge to get the leveling process spot on) , I will probably give her a major service and lube job and start the printing marathon.

aubrey · February 21, 2018, 3:01am

Oh my Giddy Aunt Mary!!!

I printed the little boat test model and it is an unmitigated disaster!!!

And to make matters worse, I don’t have a CLUE where to begin sorting it out!!!

I have seen this test model looking “not so good” but maybe I have set a new standard of “LOUSY” …

HELP !!!

vlerherg · February 21, 2018, 7:56am

That looks like severe under-extrusion and/or a partial clog resulting in not enough and intermittent plastic actually making it out of the nozzle!

barry99705 · February 21, 2018, 9:45am

Yea, I agree with Vincent, under extruding. Usually the filament has suggested temp range on the spool. I usually print 10 degrees hotter than the lower temp.

aubrey · February 21, 2018, 10:00am

Hi Vincent.

I have basically spent the whole day printing benchies - most of them big failures and the last one, although a failure, is definitely getting in the right direction.

And what you say is starting to make sense.

A number of failures happened when it just stopped extruding plastic. To get it going again I had to remove the filament and cut of a bit because the front 10 or 15mm was noticebly thicker than the “virgin” filament.

Eventually I bumped the extruder up to 210 degrees and the “Flow Rate” up to 110% AND turned off the retract function and the material cooling fan just for luck.

I got this:
[attachment file=53484]
As it was printing I was very pleased to see that the bow was looking pretty darn good and small details like the anchor chain ports, the ?exhaust? hole in the stern and ?fish pole holder? above it were not looking too shabby at all.

But then it came to the cabin section!
[attachment file=53485]

But here is the thing: The nonsense only starts when the nozzle has to move many and far distances without extruding - thats when the stringing starts and once it gets to the upper portion where extrusion happens at a more or less continuous and constant rate things start looking a heck of a lot better again.

And at least with the retract turned off it was not clogging.

So I think that I have either got a “made up of parts” extruder here which is not working very well or like I mentioned in a previous post, that “gear” (which is showing more than a little wear and tear) is not up to the task any more.

The cheapest option is to replace the gear and see how much difference it makes and thats what I’m doing tomorrow.

vlerherg · February 21, 2018, 10:14am

Dialing in a printer that is an unknown quantity can really be a pain and is an exercise in patience and creativity. From what I can tell from the pictures, it does look like you had a vast improvement, until it came to the point where there would have been a lot or retracts and non-print moves. I know you said you turned retraction off, but it really does look like it was retracting and then not priming sufficiently before beginning the next extrusion. In Repetier there is a setting to add a slight additional advance when it finishes the retraction move to combat this. With retraction all the way off, I would expect to see extra filament from oozing in the form of strings and little zits where it moved over a previous layer while continuing to ooze.

That said, your extruder does sound like a piece of work! For one, it is critical that you have adequate cooling on the cold end above the heat break. When the nozzle is at temperature, you should be able to touch above the heat break and feel no heat. Should be cool to the touch. If not, you need better cooling up there. Also, in a previous post you mentioned that the PTFE liner was probably cooked. Rather than decreasing friction, a bad liner will increase it and make your filament more prone to jams, especially during retract moves.

As for overall extrusion, did you accurately measure your filament diameter in several places and then input into your filament profile? That can make fine-tuning difficult or require extra multipliers to get right otherwise. Also could be an issue with steps/mm in the firmware, since this is a cobbled together printer and may not be using a standard drive gear. Optimally you should have perfect extrusion with the multiplier at 1.0 (100%). Anything else is an indicator of something not being set right.

Once you overcome the inconsistent extrusion issue, they rest should be a matter of dialing in settings.