Well I seem to be in a bit of a rut at the moment. The WorkBee is starting up in alarm state and needs homing each time. Homing clears the alarm but sending it to its zero position just resets the alarm and its back to home again... ditto Ditto. After homing I can jog the WorkBee and I've tried to extend the work area zero as close as I can to the 550 x 520 mm available. I've also set the zero point well inside the maximum area but get exactly the same alarm every time.
Also sent a short video and some still images to Ryan for comment
Not a happy bunny at the mo

 

Ryan - Email with attachments sent

 

OK working - Thanks Ryan/Oosnest. I have tried a few small file generated in Inkscape to try the WorkBee with. My problem was with the soft limit setting in GRBL I had it set to =1 and changing to =0 seems to have cleared the issues with only traveling between home and zero.
Need to finish the router mounting and try cutting some wood.

 

@Ryan Lock Thanks Ryan - Been practicing on my guitar headstock logo gradually taking down to the final size and trying different feed/plunge rates with the small mills. Broke 1 of my 0.4mm mills today due to a wrongly set Z zero after a cutter change...
Pilot error hopefully I won't do that again

 

Right I've had a play without too much bother about what the WorkBee was doing as long as it was moving in the right direction. So I spent some time yesterday with a steel rule and began the calibration process. I used a distance of 400mm to calibrate to. The Y axis calibrated very easily, I had a few problems with the X axis mainly due to the motor misbehaving. I had to up the current to the motor adjusting the pot, this needed to go almost all the way clockwise before the motion became smooth across the X axis and the distance traveled became consistent. Before adjusting the current pot travel along the X axis was ± 3mm over the 400mm travel so any calculations were a waste of time.
As to the calculations (found on the Sparkfun tutorial pages and the github.com grbl config page) I found that my initial Y axis travel: circa 401.5, requiring an adjustment of $101 from 200 to 199.2528. . I finally ended up with $101 = 199.27 and $100=199.6 to give me repeatable 400mm readings over the 400mm distance.

The calculation is: New $ setting = 200 x (Distance set / Distance measured), So for me S101 (Y axis) = 200 x (400mm / 401.5) = 199.2528

I tried cutting a 190mm circle after making these setting changes and the result was very very close across many different diameters

I still need to adjust the Z axis and will use a dial gauge to do this over a 20mm distance.

 

CNC Router Bits | RouterCutter
Trend Chamfer And V Groove Cutters | Trend Direct UK

 

Turk - not used these guys but might help: 2D and 3D carving, engraving and lettering router bits, solid carbide and tipped | RouterCutter

I've been buying cheap off the bay but having problems at present with breakages. Mind you the cutters are 0.6 and 0.3 mm so I'm slowing the feed rate down and reducing cutting depths at present but I think it might be the router that's the problem (run out), so looking at getting a smaller water-cooled spindle 0.5 - 0.8 Kw as I don't need heavy cuts (save my ears too!)
I've also added a second router body support to stop the router tilting during cuts. I used a single ring support to make the two ring supports - then repeated with the two ring support and it made a slight difference to both the shape of the cuts (65mm circles) and the rigidity of the router on the Z axis...
Francis

 

Problem running a Fusion 360 file
I'm playing with the free download (hobby). It's a simple level the top of a small piece of wood toolpath. It posts the toolpath OK as grbl but when I come to run it with UGS it lifts the Z axis just throws up an error message:
>>> G3Y-1.863J3.012
(error:20) Unsupported or invalid g-code command found in block.(error:20) Unsupported or
invalid g-code command found in block.
(ALARM:1) Hard limit has been triggered. Machine position is likely lost due to sudden halt.
Re-homing is highly recommended.(ALARM:1) Hard limit has been triggered.
Machine position is likely lost due to sudden halt. Re-homing is highly recommended.
[MSG:Reset to continue]

Not sure if the >>> G3Y-1.863J3.012 is the error line as prior to that the gcode was just a list of ok ok ok... after other commands.

Guess I need to try editing that line out of the gcode file and do a re-run

 

There's the simple code and the flat surface I'm running as a test....

 

Hi Ryan, @Ryan Lock No extra stops just the limit switches. I'm zeroing out to the same point each time.

What I've just done is to design an identical item in Ct2D and Fusion 360 and print out the initial gcode and compare the two. There are a number of G's in Fusion 360 that Cut2D doesn't.

So I edited the Fusion gcode to make it similar to the Cut2D and try and run it



Fusion 360 G-code test (underlined text was edited out)

%

(1001)

(F360Test)

(T1 D=6 CR=0 - ZMIN=-2.49 - flat end mill)

G90 G94

G17

G21

G28 G91 Z0

G90

(Adaptive2)

M9


T1 M6

S5000 M3

G54


M8

G0 X-56.8 Y-54.74

Z9

Z4

Z-1.29

G1 Z-1.89 F1000

X-56.794 Y-54.737 Z-1.984

X-56.774 Y-54.728 Z-2.075

…………………………………………………………

Francis

 

It did Ryan, @Ryan Lock I had an error at the end of the run, looking at the ending of the code the same references were repeated so next time I will remove those as well. The error didn't show until after the test cut had completed, but whether that would effect a longer sequence I'm not sure.

Now working on a more complex (for me!) 3D design in Fusion 360 and will take the same codes out and see what happens...

 

This ran after editing the Fusion GRBL/gcode (see above). UGS doesn't seem to like code for coolant etc...

 

@Yellers What I ended up doing was to make a light skim at the surface of a piece of hardwood. then two deeper cuts and used vernier calipers to measure the actual depths of the cut against what was asked.

For me there wasn't any need to adjust the Z axis as cut depths were as close as I could measure with my calipers. The X & Y axii were my main concern as distances in these directions are far larger than the Z axis motion.

 

@Ryan Lock Not sure Ryan, really new to F360 but it's simple enough to edit a few lines of code from the start of a sequence. If I were in business then I would look for a more automatic solution. They might even be required if the cnc xPro is wired to activate coolant etc...

 

Hi IanC,
I bought a second spindle a 1.5Kw ER11 water cooled - with a 65mm OD. The 1.5Kw 80mm OD spindle was too heavy. I will build a smaller CNC for that.
A lot of my CNC use is with small fine cutters (0.3 - 3mm) with the occasional 6mm these cutters need slow feeds so machining times can be quite long - circa 90 mins sometimes. That's why I went for a water cooled spindle to avoid an over heating problem with an air cooled spindle.
I thought quite hard about dropping to the ER11 collet size but decided that I can still get larger cutters with ¼ inch shafts should I need them, also longer reach cutters are available with ¼ shafts.

You will need to think about what your wanting to machine with your WorkBee and whether ER11 collets will allow you to use all the cutter profiles/sizes you need. If not then ER16 collets.
If you can find one an ER16 - 1.5Kw spindle with an OD of 65mm would be better. Most were 80mm OD and that puts the physical weight of the spindle too high. Whether you go for an air or water cooled spindles is a matter of choice, the extra tubes and small pond pump/water container can make things looks messy with a water cooled one but you can machine away all day without worrying about over heating.

When you do get going use very slow feed rates until you get a better idea of your machines capabilities. Tooling manufacturers will quote feed rates for their tools, these are usually for production machines with associated coolants/extraction. The WorkBee is a great machine but it's not a high level production machine - use 25% of the quoted feeds and you'll be close.

Good luck