Author Topic: GRBL Multi-Axis Grinder  (Read 2545 times)

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #15 on: December 24, 2020, 08:00:09 am »
Another small update and a warning.

Eddy the video’s I pointed to should be disregarded, via way of explanation.
After grinding the first cam (the six mm stock) I prepared a 10 mm piece to re-do
the same test, At this stage I had  mounted a small reservoir with a needle valve to
regulate the flow of the water in the wicking system.
I generated the file in CB and touched off the work,filled up the reservoir and pressed start
there are 21 step overs in the tool path.
While the grinder was running I fiddled with the needle valve getting the water delivery rate
right, wasting a lot of water and about half way through the job I could see that the water was
not going to last and raced out to a tap at the front of the shop, but by the time I did this and got back to the grinder it had finished and I'm guessing that maybe 10 laps of the stock had been
This was enough to run at least a few of the tool paths dry.
On the surface everything seemed OK the grind was ok.
In preparation to do the test again, Noticed that on a spare stepper motor, which was
at least 100 mm behind and to left the wheel  had a grey sheen to it.
Giving it a rub left a stain on my hands like the powered graphite lubricant does.
The stuff has the consistency o f talcum powder and must have been floating in the air
to get over where it did so it's a bad idea to grind carbide without wicking or flood coolant.


Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #16 on: December 24, 2020, 08:07:17 am »

After doing the tests above, I needed to modify the grinder guard (to enclose it more)
after doing this I re-installed the guard and wheel.
When I spun it up, I noticed that it was no longer running true.
Upon checking the problem out I discovered that the shoulder machined into the 
looked a little rough machined, I tried to clean it up,but short of dis-assembling the grinder and cleaning
 it up on the lathe it was about as good as it was going to get, and as I'm going to make another grinding wheel
spindle I didn't want to do this.
So I had to make a true set hub in pic this means that I can adjust the wheel true if I have to exchange
it for a straight wheel for example.

The next problem was the braking mechanism, I had made it integral to the C Axis spindle
 thinking that it was a neat and tidy solution.
And on the surface it worked ok  until I forgot to undo it and gave the Axis stack a push to move it to the
90 degree position, the brake didn't slip, the bolt holding the brake drum to the C Axis came lose and as the
Axis baseplate is 300mm long it's like using a 300 mm spanner to undo it.

So I had to rethink the breaking mechanism and used a articulated quadrant arm which  fits under the Axis base
and is pretty compact and works well. The only disadvantage compared to the integral brake is that it is
limited to 125 degrees of rotation compared to unlimited .

The next problem poked it's head up when measuring the cam I had just ground, it was over size by 0.6 mm.
I made a mistake in logic at this point and broke out the test stand and indicator and
proceeded to take some measurements.
I put a new blank in the spindle touched it up against the wheel set the indicator and  used a
300mm shifting spanner to try to 'move about' the axis base plate.

I got +- .6 mm and this is where I fell down, I thought that the C  Axis spindle was bending
but if I had just indicated the base plate itself I would have saved half an hour of head scratching
as upon doing this I measured 0.02 mm of flex in the base plate and as the C Axis Spindle
sits on the plate and the spindle centre height is 185 mm, it was flapping about like a flagpole and it
surprisingly small amount of force albeit with a 300 mm lever to bend a 300 X 400 X 12 mm plate
with two 40 X 40 X 2 mm SHS welded end to end underneath as slides to push it around the bench.

Anyway to remedy the problem was easy, I just ran some 6mm X 40  X 400 bar along the lines of force directly
underneath the spindle on the bottom of the plate and some 20 X 2.5 mm flat bar + some 6mm X 40  X 400 bar on the top.

I was feeling pretty happy at this point until indicating the the blank again, this time
the flex was 0.06 ten X better than before but still quite a bit.
The flex could only be coming from the linear bearings, they are good quality NSK's
and felt ok before being installed, I checked then in a catalog of mine and there are not the preloaded type so the
movement I was seeing was probably ok, so I went from a mono rail configuration to two rails spaced as far apart
as possible and this got to 0.01\0.02 ish so I'm going to give it a test run an see what error I get on the finished part. 

The other thing I did was to (at the grinder zero position) where it will mostly to positioned
doing the hooks that I made it to do, was to install a post and locking mechanism turning
the Axis baseplate from an cantilever beam to a  doubly supported load.

« Last Edit: December 24, 2020, 08:22:02 am by dave benson »

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #17 on: June 25, 2021, 03:39:41 am »
Another small update.

I've been using the grinder for for a while now and am pretty pleased with
the grinds I'm getting, it's not a tool and cutter grinder by any means but
it is working to it's design spec for the task it was intended to do.
I have reground some engraving bits, and D bits.

What I have to do is make some belt and pulley guards, and a bearing cover.
I had intended to make the guards by hand out of sheet metal, but don't have a roller
or stand alone bender, do have a small unit that is used with the 20 ton press although
small stuff is a bit fiddly to do.

The bearing cover on the other hand needs to be fabricated from three parts assembled
and welded and some intricate machining done, and so after some deliberation I looked into
3D printers to print the cover and the belt and pulley guards.
I tried to do my due diligence as much as I could and found this young man on Utube
who has made quite a few videos about the Ender 3, which seems to be a good beginners
printer so I pulled the trigger and bought one, which has arrived today and over the weekend
I will assemble it and get a Benchy done.

I have looked at the various slicers and found that the Prusa slicer will run on my desktop pc
win7 32 bit and have played around with it and so will start with that, I may install Cura on my
laptop and give that a test drive as well.

What I've learned is that just like using lasers for engraving, it's a glacial process and many
users mod their printers from the start. A lot of problems can mitigated by having a more
structurally rigid machine, so I'm going to use CB to make some gussets to stiffen up
the machine after getting some experience with it.
When I get to the point of being able to get a decent print, I'll post the photo's of the guards
and bearing cover.


Offline EddyCurrent

  • CNC Jedi
  • *****
  • Posts: 4739
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #18 on: June 25, 2021, 21:44:15 pm »

I have a Creality CR-10. At first glance it appears lacking in certain areas but I find the print quality is nevertheless excellent.
I did however upgrade the bed heater and that was a huge improvement.
« Last Edit: June 25, 2021, 21:46:09 pm by EddyCurrent »
Made in England

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #19 on: June 26, 2021, 03:58:14 am »
Hi Eddy

I've binged watched many videos while waiting for the machine to arrive, and have found some
channels that are geared towards printing problems and printer upgrades and others where
It's more about  'look what I made with my printer' for example casting moulds,press tooling and
harmonic drives and even chain mail.
I'm going to start off with PLA and if those prints look successful I would  like to move on to PETG
as this (by all accounts) prints like PLA but has better mechanical characteristics.
once I feel comfortable with the machine and the slicer I would like to print (after some upgrades)
Nylon, but that's a long way off yet, baby steps first. as the old saying goes
get good at what your doing first and then get fast.
I've bookmarked this guy's round up of fifty printing filaments as a general reference.

anyway the time has come and I'm off after posting this to assemble the machine and do my first print.
I do have an IR thermometer and will keep an eye on the temperature gradient of various
parts of the bed as it heats up, I did order as an extra a removable magnetic build plate as I did
see a couple of videos where people struggled to get a print of the built plate and damaged it.

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #20 on: June 26, 2021, 13:25:56 pm »
Well I set up the machine and the only thing I had to adjust was  the Z stepper motor spacer to get
it aligned properly, but that was it.
I printed a test bed leveling file from thingiverse which worked straight out of the box with
no tuning of the parameters just load and slice, (Prusa slicer) this was encouraging, so with
what was left of the PLA that came with the machine downloaded a test square with X,Y,Z on the faces
and it too just worked, the cube's dimensions should be 20 x 20 x20 and it turned out to
be pretty close X 20.13 Y 20.04 Z 20.29 so a little tuning will be required but not much I
suspect. I ordered  kg of PETG to do some more tests and print the belt guard for the
grinder. Before getting the printer I must admit I was a little skeptical and dismissive of the
3d printers ability to make anything useful and practical, but I've changed my mind and wish I had
got one sooner.

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #21 on: June 27, 2021, 01:45:01 am »
With what was left of the PLA filament that came with the machine, I printed 2 gears
one a module (1) 15 tooth gear and a module (1) 15 tooth cycloidal gear.
Designing the gears if FC and slicing the file and loading onto the printer took less than
five minutes, I loaded the SD card into the printer and pressed run and didn't look at until it
was finished, wow I'm sure this is just beginners luck and when I try to make some more
complex parts like compliant mechanisms and flextures I may receive a reality check. :)

But for now 'colour me impressed' I can not believe just how easy the process is to get going even
for a beginner like me, and if your thinking of getting one for your children as a birthday or
Christmas present as a gentle introduction to CAD then don't hesitate.

Right now I have less than half of the filament, that came with the machine (wasn't a lot)
so will try to make a module (1) rack to go with the gear and see how they mesh together.
The PetG won't get here for a week so this is about all the testing that I can do on the machine
for the moment.


Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #22 on: June 28, 2021, 01:32:54 am »
Ok I printed a module one rack and ran the gear up and down it and it's
pretty good, I did orientate the work piece along the axis with the least error +0.04 over 20 mm.
I printed 50 mm of rack and as can be seen in the pic it's ok.

I've ran out of filament now and can do no more until the PETG comes.
I've been advised not to make any adjustments to the settings until the new filament
arrives as the filament that comes with the machine actually prints "not that well"
compared to filament that comes from a reputable filament manufacturer.

The only thing I will do now is square up the Z axis motor spacing bracket as it is
not flat and square, this is cocking the leadscrew in at the top of the printer about 5 mm.
Fixing this stops  ringing at the top of a tall print.

I drew up the grinder cover and the slicer tells me that it will take 6 hours to print, which I'm ok
with as I will be doing something else while it prints in the background.

At this point it's been a relatively painless process, the machine came very well packaged and with
all the tools required to assemble it and most importantly the instructions were well done and
easy to follow.

However if the high speed printing bug bites me, I would make one from the ground up of the core xy
type making the frame out of heavy 40x40 extrusions  (bolted to a flat surface) and reduce the moving
mass as much as possible.
This way you can stop the ringing at it's source rather than try to mitigate it with software.


Offline EddyCurrent

  • CNC Jedi
  • *****
  • Posts: 4739
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #23 on: June 28, 2021, 08:23:14 am »

I can see you have the 3D printing bug now  ;D
Are you using 100% fill for those objects ?
Imagine those people who bought a printer and have no facility to produce their own 3D models, there must be a lot of them collecting dust now.
Made in England

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #24 on: June 28, 2021, 11:37:47 am »
HI Eddy

I can see you have the 3D printing bug now

Yes I have, as I've seen the potential  for home use, for example there is guy on you tube, who I originally
started watching because he participated in the land speed record trials in the 50 cc motorcycle class he failed in
his attempt and decided to make a motor at home, at this point he had very little in the way of machinery or the
skills to make what he needed, however he has a fierce determination to achieve he's goal and has failed many
times but has shown that he is willing to do what ever it takes  including up skilling, getting the machinery or if
that's not possible then excepting help from the comment section, anyway he has 3d printed carbon fiber manifold
which still needs to be finished on a mill (mating faces) made planar.

I'm waiting with much anticipation to see if those components hold up (to the heat) mostly, and if they do I'll be
fully on board and will make a printer to specifically to do this.
The concern I have is that the chopped up carbon fiber is suspended in a polymer and just how well the polymer
will hold up.

I have looked about and there is a German company that uses two separate heads one for the polymer filament
and one which has the continuous carbon filament flowing through it which is only cut when needed making parts that
are comparable in strength to the metal alloys used machinery.

I've already priced the 20% chopped carbon\ 80% Nylon filament and if you compare it to the cost
of metals and also the ability to produce items not possible or very difficult to make with subtractive machining then
it actually compares very well.

I don't want to sound like a fan boy as there will always be subtractive machining, but I've seen
the potential of 3d printers.
There's a NZ rocket company printing rocket engines as well as many in the US and I bet
Rolls-Royce would be looking into it as well and probably have been for a long time.

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #25 on: June 28, 2021, 11:42:46 am »
Are you using 100% fill for those objects ?

No 20%, I've left everything as default except the extruder retraction, I intend to make a profile for each filament
type and maybe one each for higher\lower speed printing when I know more.
I made the wall thickness 3 mm, and to be frank I don't know if 20 % is enough as I'm not experienced
enough to have an intuitive feel for things, as I do for milling or lathe work.

Imagine those people who bought a printer and have no facility to produce their own 3D models, there must be
a lot of them collecting dust now.

Yes which is a shame really, I have made my models with the linkstage fork of 0.19 Freecad by RealThunder.
which is a huge improvement as not only can you make the UI look similar to Fusion but under the hood
the sketcher workbench performs in a similar way, they have gone a long way to fix the topological naming problem
where it had been very easy to break your model, I like it a lot.

I know you use Rhino which is a very capable cad system,so you will be fine, but for those who have found
Freecad wanting in the past I would urge them to have another look as after the Autocad changed Fusion there
has been a torrent of new blood in the programming area and things seem to be gaining pace.

On one forum I haunt, it was made mention that at a freeware conference the Blender
people were talking to the Freecad guy's, and this would push development ahead in leaps and bounds.

For those still coming to grips with parametric modelling there are websites like Grabcad, Thingiverse which have 100's of thousands of files and maybe half a dozen that I've personally visited which have 10's of thousands of models, The Grabcad models are 'by and large' excellent quality and in a variety of formats including STL.

There are definitely some rules that must be observed when using a STL to make a engineered component that has to conform to specific dimensions and tolerances.
I may speak about this more when I have a component to show.


Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #26 on: July 06, 2021, 02:44:16 am »
The PETG came a couple of days ago, and after formulating an profile
for the filament and printer and doing a calibration block and a dimensional
accuracy test, I set about doing some Spindle guards.

The first one was a simple rounded box shape and printed well except for
the inner lip. I had not included any support as the slicer was set to Auto
meaning that if it thought that the job required it, it would include some.

The redesigned second attempt printed the inside very well, however the flat
faces that were parallel to the bed turned out terrible as there was no support
for the bridging moves.

For the third attempt, I removed the faces parallel to the bed altogether and
this printed well on the inside and except for a couple of goobers on the face
that was attached to the bed.

It also was good for structural rigidity (quite rigid layer adhesion wise) and this
is the one I'm going to use.

This took 5 hours to print as once I realised where optimisations can be made
in regard to model orientation on the print surface and and more generally avoiding
surfaces in the model that would be hard to print.

Five hours might seem like a long time, but the machine ran unattended and the cost
of the filament was for all thee was 6 dollars.

With two very similar models, the print time and model structural integrity can both
be greatly optimised with a little forethought.

I could have kept the parallel faces, and use supports, this however added another
couple of hours to the print and I didn’t want to spent that much time on it.

I'm fully on board the 3d printer train now and will re-make the coolant system
as prototyping this way is cheap easy and fun.

So in summary, the PETG prints very well and the model is more than strong enough
for this application. The slicer software is the best freeware software I've ever used
It's blindingly fast compared to CB or FC on the same model, and I've not managed to
break it yet. Kudos to the Authors, well done.


Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #27 on: July 06, 2021, 02:50:04 am »
The rest of the photo's
« Last Edit: July 06, 2021, 02:55:08 am by dave benson »

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #28 on: July 06, 2021, 02:51:37 am »
OK back to the grinder, I've re-ground some engraving (D) bits (broke the 0.1 mm tips) of
quite a few until I got my act together.

The spindle works fine up to 200 rpm in position mode (600 rpm at the motor)
and for velocity mode, I'm going to add a multiplexer\data selector (one chip)  and an oscillator
so that I don't have to simulate this with Gcode as I'm doing presently, as the files are
large and I'm coming up against the 32 bit memory limit of win7-32bit.
Se e the .nc file
I will either use a wider belt on the X\Y Axis, (6mm now) to 12 mm or use 6mm metal\carbon fibre re-enforced belts as these ones are the cheap 3d printer type and not up to the task over  time.
I did price some mini ballscrews, but needed  some time to recover from the “sticker shock”.

« Last Edit: July 06, 2021, 02:56:38 am by dave benson »

Offline dave benson

  • CNC Jedi
  • *****
  • Posts: 1468
    • View Profile
Re: GRBL Multi-Axis Grinder
« Reply #29 on: July 10, 2021, 01:59:02 am »
Another small update.
Earlier I mentioned that I would like to print practical structural parts from
Carbon Fibre and have found this young man (Stefan) from cnc kitchen
who has build his own testing equipment to test various properties of
printed parts. In one video he used Fusion and FEM to model and improve
a Shelf bracket. I've printed all of the 340 mtrs of PETG  that I have learning
about the properties of the material and the best orientation to print a strong part fast.
Like all machining processes FDM has strengths and weaknesses meaning that
some parts are well suited to printing while others are better suited to traditional
machining methods where you have a larger range of materials and with an appropriate
Mill\Lathe can machine much faster.
Even though the pure filaments are isotropic, you lay the layers down in a
manner that means the print exhibits a grain structure similar to wood kind\sorta meaning
that they can be between 40-60% stronger in one axis compared to another.

This is the Guy I've been keeping an eye on as he was going to print a Carbon Fibre
inlet manifold for a two cycle engine, and he has made the part and ran the engine.

Printing details video:

Running video:

It went very well for him, nice looking part that functioned well too.
It turns out that he is using the same printer as mine with the only modification being
the addition of a MicroSwiss Extruder and a firmware update.

I've ordered the Parts to be able to Print 80% Nylon\ 20% Carbon Fibre.

Including an extra complete all metal hot end with Stainless Steel heat break which is ok for
Nylon not CF.

A2 Tool steel nozzel.
Bi Metal heat break   copper\Titianium  for better thermals allowing the use of the existing
heater with a little slower printing speed.
A dual Z Axis kit.
and will have to make an enclosure.

The first thing I will make is some fixtures for the grinder as according to the videos I've seen
the CF is in the region of 40 % stiffer for the same cross sectional area than the 6061 I had planned to use.

If this works out , I'll build a fully enclosed coreXY design printer Running Klipper with CF parts
printed on the Ender3 with some post machining on the mill, I have watched some builds where
the builder ordered most of the parts and then downloaded and printed the rest out of
PLA or PETG and watching the assembly process was appalled at the fit and tolerances they were
prepared to accept.