A Second CNC machine

Last March, I began installation of a second CNC machine capable of work up to 60 x 144 inches in my 3rd floor studio. My first machine is now dedicated to wood carving while the second machine is for drawing and painting so I can now work several projects at once.

This turned out to be a MUCH more difficult installation than the first. The new machine is almost twice the size of my first -- all the heavy pieces had to be carried up three flights of stairs and then bolted together like a giant erector set. The rails of each axis had to be carefully and precisely leveled and fastened exactly parallel -- not an easy task with bubble levels and ultimately I bought a laser level which made the process possible -- until then, I was really starting to pull my hair out (and I don't have that much hair to pull anymore)! The bed and vacuum plenum engineering was much more problematic than I'd imagined because they had to be built up from several interlocking panels. In the end, it turned out that there were a number of electonic and electrical problems which I had a LOT of trouble diagnosing and repairing, so most of my work these past several months has been more engineering than art, but the machine is FINALLY complete and I am ready now to make some larger and more complex art...

New CNC machine with bed complete and two of three vacuum plenum panels installed. Holes cut in bed (one visible in foreground) for hookup to vacuum which provides suction through trupan ultra-light MDF surface).


One of the three 4 x 5 foot plenum panels with waffle cuts for vacuum distribution


detail showing vacuum waffle cuts on bottom surface


vacuum plenum complete

July 11, 2008 update...

vacuum hose hooked up to fitting on table bottom


3-zone control of vacuum to table via simple knife gate valves


New CNC machine complete with table planed flat.
All that's left to do now is install compressed air hookup for painting and SWEEP UP!!!

-- Mike

Labels: equipment

  ¶ 9:48 AM 1 comments links to this post

  Self Portrait, painting in acrylic on linen, 60x40 inches

Today I completed something a little bit 'different'... Looking back to the velvet paintings displayed at the 5 and dime when I was a kid -- maybe they're still up there today -- hula girls, Elvis in all his glory, typical 60's kitch in stiff opaque color practically glowing out of that black-black velvet... Well, I didn't so far as to actually paint this on black velvet (which would have been pretty nice in a retro-leisure-suit sort of way -- but I'm just not that 'cool' I suppose).

So this is painted on stretched linen in transparent titanium white acrylic on a carbon black ground.

self portrait, Oct 11, 2007, 60 x 40 inches, acrylic on stretched and primed linen
click image for nice enlargement

Just getting started
click image to view Windows Media Player movie of painting, 700kb download, 35 seconds)

I first painted the entire canvas black, then applied many layers of white paint in order to build whiter and whiter lines out of previously painted white lines -- so each successive overpainting made that area whiter and more opaque.

The only dark values in the painting come from the black underpaining -- nothing but successively more opaque white was used to produce the image.

Up close, it's pretty intense and interesting because of all the stringy white squiggles and circlets from which the image is constructed...

-- Mike

Labels: equipment, painting

  ¶ 12:28 PM 0 comments links to this post

  Painting with Paasche Flow Pencil

Painting blue/black acrylic paint on self-portrait #2

I'm working on a short series of painted self-portraits at the moment. I've mounted a Paasche Flow Pencil (very much like an airbrush, only without the air -- the 'nozzle' is conical and a needle valve seats in the nozzle in order to control the flow of paint. The Paasche is not a very high-precision device, so fluids which aren't very viscous (like water) flow out quite fast, even when the valve is 'closed'.

I suppose I could'a figured this out in advance, but it seemed to me that coffee or tea stained paper would make a good first trial image and clean-up would be very easy... WRONG! WHAT A MESS!

So viscosity about like cream works quite well.

First attempt -- connection to solenoid broke and pen never lifted so paper got messed up!

I removed the little paint-pot which came with the flow pencil and replaced it with a water bottle I'd modified for the purpose. The cap has the paint outlet mounted in it and I mounted a vacuum-relief/pressure port (to force more viscous paint to flow using a bit of air pressure) on the side (a tube runs to the 'bottom' of the bottle so air can get in). This allows me to fill the bottle with paint (or whatever) without making a mess, mount the cap, and connect the hose to the pen before inverting the bottle and allowing paint to flow. Works well.

Because the pen doesn't have many parallel surfaces -- it's all cones and curves and joints, it wasn't easy to conceive a way to hold and actuate it. In the end I decided to make it easier for me to construct the gizmo and so I made it to kinda work backwards (power-off allows paint to flow -- so when it stops, paint keeps running). A single solenoid both lifts the pen off the paper and releases the trigger, stopping the ink flow while the pen is moved to a new painting location. When the solenoid is 'off', a spring pushes the pen down onto the paper or canvas and actuates the trigger at the same time.

The paper surface on my 2nd attempt is now badly abraded with 'fuzz-balls' all over the surface in the darkest areas -- next attempt I'll use a heavily sized paper and adjust the height of the flow-pencil so it JUST BARELY TOUCHES THE PAPER!

I'm VERY excited about the possibilities for color work, both directly (painting flat areas of transparent color), and especially indirectly (using weighted gestures again) since the flow-pencil produces a VERY sharp-edged line -- much cleaner edge than my airbrush!

-- Mike

Labels: equipment, painting

  ¶ 8:42 AM 1 comments links to this post

  Painting machine 84 x 45 inch watercolor painting underway

My new 17-value airbrush characters seem to be working – now if I can only get my airbrush to not clog up after an hour or 30 minutes or 90 minutes – otherwise it’s gonna require constant supervision which ain’t worth it (and BORING)!

Apple QuickTime movie of painting underway

Here’s an image of my spreadsheet calculations of area – I’ve made a set of ‘gestures’ – diamonds, and squares in various sizes and an ‘X’ and a ‘+’:

The ‘X-area’ column is the list of gestures D for Diamond, S for Square, X for diagonals, and P for Plus and the 3-digit number is the area in thousandths that an eighth inch line produces in a 1 inch box.

These are designed to be overpainted in each of red, yellow, blue, and black according to the color and value content of the corresponding area of an image. I’ve got a sample image partially painted in yellow, blue, and red now and it’s pretty much electrically brilliant! There are 17 to the 4th theoretical color possibilities (83,000 plus variations), so PLENTY of variation!

THEN I had to write the program to analyze the image and write the ShopBot code – my code reads an initialization file I wrote in ShopBot code which defines the characters and the subroutines to call them, then the program goes on to analyze each pixel in the image and generate the ShopBot code to position the machine to the pixel and do the painting. The SO,3,1 turns on the airbrush and SO,3,0 turns it off. M2 means Move in 2 dimensions (x,y) – J2 means FAST Jog in 2 dimensions (x,y)

Sub output_sbpcode_for_cmyk() ' 1/26/2007 gesture generator... Mike Lyon
' THE NEXT TWO LINES ARE FOR CONVENIENCE IN INITIALIZING CURRENT DRAWING LOCATION
' IT'D BE A GOOD IDEA TO ENTER THIS LOCATION IN THE FORM ALONG WITH &scale
PrintLine(3, "SO,3,0")
PrintLine(3, "J2,8,45")
PrintLine(3, "' MRL-generated GESTURE CODE from image")
PrintLine(3, "' position bot at top left of drawing first!")
PrintLine(3, "&xstart = %(1) ' remember beginning x value")
PrintLine(3, "&ystart = %(2) ' remember beginning y value")
PrintLine(3, "&scale = .4 ' this is 1=100% of 1 inch, .4 = .4")
PrintLine(3, "&rrr = 0 ' initialize rows (Y) variable")
PrintLine(3, "&ccc = 0 ' initialize X variable")
skipit = 0
lines_written = 10 ' TO ACCOMODATE TWO LINES ABOVE -- NORMALLY SET TO 0
PrintLine(3, "GOTO ROW_1")
While Not EOF(1) ' include the gesture subroutines file
sbp = LineInput(1)
PrintLine(3, sbp
lines_written += 1
End While
For rrr = 0 To rawimage.Height - 1
PrintLine(3, "ROW_" & rrr + 1 & ":")
PrintLine(3, "&rrr = " & rrr)
lines_written += 2
For ccc = 0 To rawimage.Width - 1
' getbrightness returns value from 0 to 1 where 0 is black and 1 is white
zz = Math.Round((1 - rawimage.GetPixel(ccc, rrr).GetBrightness) * 16, 0)
' 0 is digits to rt of decimal
If zz > 0 Then
If skipit > 0 Then
skipit = 0
End If
PrintLine(3, "&ccc = " & ccc)
PrintLine(3, "GOSUB " & zz.ToString("00"))
lines_written += 2
Else ' zz = 0
skipit += 1
End If
If Button5.Text = "END" Then Exit Sub
Next ccc
Label20.Text = "Writing " & rrr & " --- " & lines_written
ProgressBar1.Value = 100 * rrr / rawimage.Height
Call updateform()
If Button5.Text = "END" Then Exit Sub
Next rrr
PrintLine(3, "J2,&xstart,&ystart ' back to initial location and ready for next color")
PrintLine(3, "END")
lines_written += 2
End Sub
lines_written += 2
End Sub

This is actually only the working part of a MUCH longer program – the long program gets input from a screen, checks to make sure it all makes sense, and then calls the appropriate operation, in this case the subroutine above… Here’s what the screen looks like when the program is running (just finished generating the ShopBot code for the color ‘cyan’ for a Jim Winter drawing:

COOL, eh? I added the new code onto my existing program so I wouldn’t have to rewrite all the file and form handling – can’t believe it actually works!!!

I’ve defined some subroutines named after each of the 17 possible values (00 through 16 is white through black) which call the gestures I’ve designed to produce the values -- – the first routine listed below is for ‘black’ or value 16 – it calls .313” square, .438 square, .234” plus, and .338” diagonals in turn to produce the ‘shading’ in that area.:

Then I defined the movement for each ‘gesture’ – these are shown below as ShopBot subroutines – you can ‘call’ the subroutine (GOSUB D375 calls the .375 square inch diamond, for example) – the starting corner position is assumed to be in the variable “&xstart” and “&ystart” and the scale factor in &scale. Then the row and column indexes are in &rrr and &ccc variables – in a 100 row x 50 column image, for each rrr (1 through 100) I position to each column ccc (1 through 50) and draw the gestures for that position.

Finally, here’s just the very beginning of the completed code to produce 0.4” squares (&scale = .4) for a very STUPIDLY large drawing 200 squares high by 110 squares wide… Next time I’ll start a bit smaller :-)

SO,3,0 ‘ turn off the airbrush
J2,8,45 ‘ jump to the beginning of the drawing – remove this for normal operation – this is for debugging only
' position bot at top left of drawing first!
&xstart = %(1) ' remember beginning x value
&ystart = %(2) ' remember beginning y value
&scale = .4 ' this is 1=100% of 1 inch, .4 = .4
&rrr = 0 ' initialize rows (Y) variable
&ccc = 0 ' initialize X variable
GOTO ROW_1 ‘ jumps around the gesture subroutines which follow below and get right to the painting!
' this is the start of a gesture subroutines file
' scaling of &scale = 1 (100%) makes marks in a 1" square
' at &scale = 1 (100%) marks are to be 1/8" wide
' gesture call names are LETTER followed by number nnn
' D = diamond shape of .nnn area
' S = square shape of .nnn area
' X = crossed diagonals of .nnn area
' P = 'plus' (vertical & horizontal) of .nnn area
' main calls are 00 through 16 and represent values (tone or shade) from white to black
16:
' this is entry point to produce darkest gesture (value 16)
GOSUB S313
GOSUB S438
GOSUB P234
GOSUB X338
RETURN
15:
' entry point to produce next to darkest gesture (15)
GOSUB S188
GOSUB S438
GOSUB P234
GOSUB X338
RETURN
14:
GOSUB S438
GOSUB P234
GOSUB X338
RETURN
13:
GOSUB D375
GOSUB S063
GOSUB S438
RETURN
12:
GOSUB D375
GOSUB S438
RETURN
11:
GOSUB D250
GOSUB S438
RETURN
10:
GOSUB D375
GOSUB X338
RETURN
09:
GOSUB D250
GOSUB S063
GOSUB X338
RETURN
08:
GOSUB D250
GOSUB X338
RETURN
07:
GOSUB S188
GOSUB X338
RETURN
06:
GOSUB D375
RETURN
05:
GOSUB S313
RETURN
04:
GOSUB D250
RETURN
03:
GOSUB S188
RETURN
02:
GOSUB D125
RETURN
01: ' entry point to produce next to lightest gesture
GOSUB S063
RETURN
00: ' entry point to produce blank (do nothing)
RETURN
D375: ' entry point for 37.5% diamond
J2, &xstart + &scale * ( &ccc + 0 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 1 )
M2, &xstart + &scale * ( &ccc + 1 ), &ystart - &scale * ( &rrr + 0.5 )
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 0 )
M2, &xstart + &scale * ( &ccc + 0 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,0
RETURN
D250: ' entry point for 25% diamond
J2, &xstart + &scale * ( &ccc + 0.125 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 0.875 )
M2, &xstart + &scale * ( &ccc + 0.875 ), &ystart - &scale * ( &rrr + 0.5 )
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 0.125 )
M2, &xstart + &scale * ( &ccc + 0.125 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,0
RETURN
D125: ' entry point for 12.5% diamond
J2, &xstart + &scale * ( &ccc + 0.375 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 0.625 )
M2, &xstart + &scale * ( &ccc + 0.625 ), &ystart - &scale * ( &rrr + 0.5 )
M2, &xstart + &scale * ( &ccc + 0.5 ), &ystart - &scale * ( &rrr + 0.375 )
M2, &xstart + &scale * ( &ccc + 0.375 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,0
RETURN
S438: ' entry point for 43.8% square
J2, &xstart + &scale * ( &ccc + 0.0625 ), &ystart - &scale * ( &rrr + 0.0625 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.0625 ), &ystart - &scale * ( &rrr + 0.9375 )
M2, &xstart + &scale * ( &ccc + 0.9375 ), &ystart - &scale * ( &rrr + 0.9375 )
M2, &xstart + &scale * ( &ccc + 0.9375 ), &ystart - &scale * ( &rrr + 0.0625 )
M2, &xstart + &scale * ( &ccc + 0.0625 ), &ystart - &scale * ( &rrr + 0.0625 )
SO,3,0
RETURN
S313: 'entry point for 31.3% square
J2, &xstart + &scale * ( &ccc + 0.1875 ), &ystart - &scale * ( &rrr + 0.1875 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.1875 ), &ystart - &scale * ( &rrr + 0.8125 )
M2, &xstart + &scale * ( &ccc + 0.8125 ), &ystart - &scale * ( &rrr + 0.8125 )
M2, &xstart + &scale * ( &ccc + 0.8125 ), &ystart - &scale * ( &rrr + 0.1875 )
M2, &xstart + &scale * ( &ccc + 0.1875 ), &ystart - &scale * ( &rrr + 0.1875 )
SO,3,0
RETURN
S188: ' entry point for 18.8% square
J2, &xstart + &scale * ( &ccc + 0.3125 ), &ystart - &scale * ( &rrr + 0.3125 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.3125 ), &ystart - &scale * ( &rrr + 0.6875 )
M2, &xstart + &scale * ( &ccc + 0.6875 ), &ystart - &scale * ( &rrr + 0.6875 )
M2, &xstart + &scale * ( &ccc + 0.6875 ), &ystart - &scale * ( &rrr + 0.3125 )
M2, &xstart + &scale * ( &ccc + 0.3125 ), &ystart - &scale * ( &rrr + 0.3125 )
SO,3,0
RETURN
S063: ' entry point for 6.3% square
J2, &xstart + &scale * ( &ccc + 0.4375 ), &ystart - &scale * ( &rrr + 0.4375 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 0.4375 ), &ystart - &scale * ( &rrr + 0.5625 )
M2, &xstart + &scale * ( &ccc + 0.5625 ), &ystart - &scale * ( &rrr + 0.5625 )
M2, &xstart + &scale * ( &ccc + 0.5625 ), &ystart - &scale * ( &rrr + 0.4375 )
M2, &xstart + &scale * ( &ccc + 0.4375 ), &ystart - &scale * ( &rrr + 0.4375 )
SO,3,0
RETURN
X338: ' entry point for 33.8% 'X'
J2, &xstart + &scale * ( &ccc + 0 ), &ystart - &scale * ( &rrr + 0 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 1 ), &ystart - &scale * ( &rrr + 1 )
SO,3,0
J2, &xstart + &scale * ( &ccc + 0 ), &ystart - &scale * ( &rrr + 1 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 1 ), &ystart - &scale * ( &rrr + 0 )
SO,3,0
RETURN
P234: ' entry point for 23.4% 'plus'
J2, &xstart + &scale * ( &ccc + .5 ), &ystart - &scale * ( &rrr + 0 )
SO,3,1
M2, &xstart + &scale * ( &ccc + .5 ), &ystart - &scale * ( &rrr + 1 )
SO,3,0
J2, &xstart + &scale * ( &ccc + 0 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,1
M2, &xstart + &scale * ( &ccc + 1 ), &ystart - &scale * ( &rrr + 0.5 )
SO,3,0
RETURN
‘ THIS IS THE START OF THE ACTUAL DRAWING routine to produce Jim painting -- black layer:
ROW_1: 'nothing in this row
&rrr = 0
ROW_2: ' nothing in this row
&rrr = 1
ROW_3: ' nothing in this row
&rrr = 2
ROW_4: ' nothing in this row
&rrr = 3
ROW_5: ' column 24 is first area to paint
&rrr = 4
&ccc = 24
GOSUB 01 ' draw gestures for gray value = 1 of 0-16
ROW_6:
&rrr = 5
&ccc = 25
GOSUB 07
ROW_7:
&rrr = 6
&ccc = 26
GOSUB 02
&ccc = 144
GOSUB 02
&ccc = 145
GOSUB 04
&ccc = 146
GOSUB 04
&ccc = 147
GOSUB 01
ROW_8:
&rrr = 7
&ccc = 26
GOSUB 10
&ccc = 138
GOSUB 05
&ccc = 139
GOSUB 08
&ccc = 140
GOSUB 11
&ccc = 141
GOSUB 11
&ccc = 142
GOSUB 13
&ccc = 143
GOSUB 15
&ccc = 144
GOSUB 15
&ccc = 145
GOSUB 15
&ccc = 146
GOSUB 15
&ccc = 147
GOSUB 15
&ccc = 148
GOSUB 15
&ccc = 149
GOSUB 14
&ccc = 150
GOSUB 13
&ccc = 151
GOSUB 13
&ccc = 152
GOSUB 13
&ccc = 153
GOSUB 12
&ccc = 154
GOSUB 11
&ccc = 155
GOSUB 06
ROW_9:
&rrr = 8
&ccc = 26
GOSUB 09
&ccc = 27
GOSUB 07
&ccc = 133
GOSUB 02
&ccc = 134
GOSUB 07
&ccc = 135
GOSUB 10
&ccc = 136
GOSUB 13
&ccc = 137
GOSUB 14
&ccc = 138
GOSUB 15
&ccc = 139
GOSUB 14
&ccc = 140
GOSUB 12
&ccc = 141
GOSUB 12
&ccc = 142
GOSUB 10
&ccc = 143
GOSUB 08
&ccc = 144
GOSUB 09
&ccc = 145
GOSUB 09
&ccc = 146
GOSUB 09
&ccc = 147
GOSUB 10
&ccc = 148
GOSUB 11
&ccc = 149
GOSUB 11
&ccc = 150
GOSUB 09
&ccc = 151
GOSUB 12
&ccc = 152
GOSUB 12
&ccc = 153
GOSUB 13
&ccc = 154
GOSUB 13
&ccc = 155
GOSUB 14
&ccc = 156
GOSUB 16
&ccc = 157
GOSUB 12
ROW_10:
&rrr = 9
&ccc = 26
GOSUB 02
&ccc = 27
GOSUB 12
&ccc = 131
GOSUB 05
&ccc = 132
GOSUB 13
&ccc = 133
GOSUB 14
&ccc = 134
GOSUB…

It goes on and on like this – two or three of these commands for each of the 22,000 squares in the drawing – then the next color paints over them, then the next color, and on and on for WEEKS!!!

But it works!!!

-- Mike

Labels: equipment, painting

  ¶ 1:39 PM 3 comments links to this post

  Painting machine tests -- photos and a movie

Yesterday and today I've been experimenting with the new computer controlled painting device and the results are encouraging -- a few interesting problems (maybe these will turn out to be 'characteristics') which I hope to resolve, but -- well, see for yourself (click the play button on the movie -- about 1 minute long):

-- Mike Lyon

Labels: equipment, painting

  ¶ 2:18 PM 2 comments links to this post

  computer controlled paintbrush gizmo complete

Computer controlled paintbrush gizmo up and running as of 30 minutes ago!!!

Parts for airbrush assembly – the airbrush itself is just above the angle iron with the solenoid already installed. Solenoid plunger above it, little bent aluminum airbrush retaining clips above airbrush – the angle iron gets bolted into the aluminum pipe and holds the airbrush vertically in the center of the pipe – the plastic pipe shims the assembly out to proper dimension for my holder… Ink/paint bottle at left and “Z” clip to keep the plunger from falling out is next to plunger – it’s a damned simple solution given how long it’s taken me to figure out!

The new paintbrush assembly almost ready for installation!

The completed assembly (minus ink bottle) is mounted in machine – the interface board came from ShopBot – it’s very simple board with five relays driven from program controlled outputs in the ShopBot controller (the ribbon cable you see accordion-folded runs from top of the board over the cable umbilical of the machine and into the main ShopBot control box… I use only one of the relays (and 24v power from the controller) to energize and deenergize my little solenoid which pushes the button on the airbrush…

I mounted the relay board onto the Z-axis of the ShopBot in a little plexi box I built this afternoon – the box should keep the board clean when I’m routing instead of painting… The cover is open at the bottom and just slides down over the plexi back in some nicely fitted grooves in the sides of the cover. You can see the brush assembly mounted into the machine at the bottom of this photo.

Here’s a view of the business-end of the assembly neatly mounted in the machine – the air hose has a quick-release at each end so it’ll be a snap to change over from painting to drawing to routing – but I’m going to need to build a 2nd generation drawing control to fit the relay board the same way and using a 24v continuous duty solenoid which I now have on hand instead of the intermittent duty 110v solenoid I used for the previous two drawings (and melted one of the coils in the process, unfortunately – that won’t happen again once I’ve converted over to the 24v device – and I won’t have to worry about 110v finding its way into the delicate electronics (or my delicate body, either)…

You can see the 24v solenoid wires (white wires) about 1/3 up from bottom left, and solenoid plunger on button of airbrush and black air hose coming in from right… Since it works in reverse of pen-lifter, I have to program a new converter before I can use it (the pen solenoid got powered when it was to STOP drawing – the brush gets powered when it is to PAINT, so my old program is reversed and I need to make a simple change in order to generate the proper code for the new airbrush… Should be testing with actual paint tomorrow!!! YIPPEE!!! But… The thing does actually WORK right now, the computer turning the airbrush off and on with command, so I’m ready to go!!!

-- Mike Lyon

Labels: equipment, painting

  ¶ 6:18 PM 2 comments links to this post

  VACUUM PLENUM for press

Labels: equipment, Woodblock

  ¶ 5:39 PM 1 comments links to this post

  Computer Carving Machine operational! The Large Press installation is almost done now! This week I'm cutting the final parts -- the vacuum plenum for the press itself... This is a somewhat complicated task it turns out, but I believe I've got it figured out.

I've successfully moved my ShopBot CNC machine from the 3rd to the 1st floor... After cutting a large vacuum plenum to hold 4x8 foot by quarter-inch plywood sheets flat and solidly in place on the machine, I experimented yesterday by carving the dozen blocks for my (unimpressive image, I'm sorry to say) prints for Baren Forum Mythical Beasts Exchange (#27).

My kids gave me a nifty and TINY little all-digital video camera for XMAS, and I used it to shoot a 2 1/2 minute video of the computer-guided tool in action -- suitable for Windows Media Player, you can VIEW IT HERE, but you may want to turn down your volume as the router is very loud and obnoxious... What you'll see are some brief cuts of V-bit carving to outline the printing shapes, then clearing outside the V-carved areas with a 1/4" down-spiral -- I've also shown a couple of views of the control screen -- the first shows each line of program code as it's executed -- I think we're at about line 14,000 in the shot -- later there's a view showing the cutting head position in X (length) Y (width) and Z (height) coordinates.

In spite of the unimaginative image, the blocks turned out quite nicely! I'll be printing in pinks and greens for this one...

-- Mike Lyon

Labels: equipment, Woodblock

  ¶ 1:13 PM 1 comments links to this post

  Garage Drawer Humidor in ACTION! So... I'd been giving a LOT of thought to one difficult question -- HOW, exactly, to make the sliding drawer humidor MOVE? I looked into available liner actuators -- there are a TON of them available, but they're all so darned expensive, and all require some kind of 'programming' to make them work... Prices seem to start at around $1,500 without controls -- YIKES, that's a lot of money! So I 'backed up' to consider pneumatic cylinders and valves -- I've already got a small but very noisy compressor, and a cylinder and valves might run closer to $500 -- still too much, really. So I thought I'd just use ropes and pulleys and counter-weights and struggle with it manually -- but this humidor/drawer moves awfully far -- about 70 inches and this didn't seem like a very cool or efficient solution.

Finally, I had a major "AHA" night before last when Linda, my wife, hit the button to open our automatic garage door! YES, that's IT! So... Yesterday I did a little research on the Internet and found a nifty 'screw driven' garage door opener manufactured by Genie and widely available for about $170! I ran over to Home Depot and bought one -- they had about 50 of 'em on the shelf -- includes all the electronics and controls I needed, plus the linear actuator -- they sell so many of them that they are CHEAP for what's in the box, I think!

Last night and this morning I assembled the thing with some modifications to the limit switches so they'd work upside down which was the best way for me to mount the thing to my humidor drawer. Went together EASY and works like a charm! The seven inch per second action is pretty ideal for the way I want to use it, and I can actuate it by foot or by knee very easily -- even came with three remote control devices (which I don't imagine I'll need), so I can very easily open the drawer, register the paper to the block while the drawer holds the paper up above the inky block, hold it in place with both hands and hit the switch to close the drawer -- looks to me like it will work like a charm! I'm VERY happy today!

Here's the humidor in action (remote in one hand, camera in the other, animated from still photos as the drawer moved from closed to open) -- YOU GOTTA LOVE IT! That's a full 4x8 foot drawer moving easily back and forth -- it's working so very well!

Here's the view from underneath and behind the humidor with the drawer closed. The slide has actuated the garage drawer opener's 'open' limit switch at upper left and the linkage from the slide to the drawer is visible at the top left to center. The opener is supported at the far end by an aluminum triangle visible on the right, and the wire coil to the 'foot switch' is visible in the background.

Barely worth mentioning, as I might actually disconnect this and go with one of the remotes -- but this is the 'foot switch' -- modified by hot-melt gluing a small block of wood to the button so it extends far enough above the body of the thing to be 'clicked' by a tap of the foot...

-- Mike

Labels: equipment, Woodblock

  ¶ 6:23 PM 1 comments links to this post

  Sliding Humidor nears completion...

March, 2004 printing table with sliding drawer humidor -- very sturdy table made of 1/4" plywood pieces which 'lock' together with simple hooked tabs and slots. Takes about five minutes to assemble or break down and packs into a 2" thick box for transporting.

This 'small' humidor is a sealed plexi box which dovetails onto the table. It has two open-top drawers with blotter in the bottom to keep paper damp and ready for printing. The bottom drawer slides open just over the block by pressing a foot 'pedal'. A sheet may then be pulled up over the lip of the drawer and registered onto the block while supported by the drawer. As the foot pedal is released, the drawer slides closed under the paper allowing it to settle onto the block very sweetly!

The new humidor is a very scaled up and mobile version of a traditional Japanese damp-stack. The bottom of the drawer is a 4 x 8 foot sheet of 6mm Coroplast (corrugated plastic sheet) which supports the damp stack -- 1st the damp blotter, then the dampended sheets (up to 4 x8 feet), then 'sealed' with another sheet of Coroplast. The top sheet will be supported by a simple lifting bar, similar to the way silkscreens are generally supported above the paper until the screen is dropped -- a pair of short retractable legs will hold the Coroplast above the paper stack during registration (while the drawer is open over the block). When the drawer is allowed to close, the 'legs' will be retracted by the structure of the drawer support and the Coroplast will settle back down, sealing the dampstack until it is time to print the next sheet. Printed sheets will be peeled up from the back and slid down over the front rail of the press into a similar drawer (not yet completed) which opens to the front mounted under the lower rails of the press. When a press pass is completed, the whole stack of printed papers will be picked up as a sandwich and flipped neatly back into the open paper delivery drawer (I HOPE)!

Because I want to be able to move freely around the press while brushing up the block, I've left a 24" aisle open between the press and the closed paper delivery drawer, so the drawer has to move quite a long way to arrive over the block! Each 'glide' of the large drawer is a bit over 87 inches long, stiffened by a 30x60mm aluminum extrusion. To prevent tipping, the drawer frame will be bolted into the floor (which is my hardware store visit, as they supplied me with the wrong size mounting hardware late last week -- ugh)...

side view of paper delivery drawer (slid back a bit to allow drilling into floor underway)

detail showing drawer corner and glide assembled to end plate -- simple 'stops' (not shown) are bolted into the ends of the fancy extrusions to prevent over-travel in either direction.

The drawers are quite simple -- the long front and back are 1" square x 1/8" aluminum tube with 1-1/2" x 1/8" flat bar screwed underneath to provide a lip for the CoroPlast bottom. The sides are 1-1/2" x 1/8" aluminum angle, and there's another bar mounted in the center between the front and back tubes to prevent bowing during open and close actions.

The drawer support is a simple trestle, held parallel by four lengths of 1/2" galvanized pipe tensioned with all-thread running the length and piercing the the 3/16" aluminum end plates and the 2" square tube legs -- tightened with four nuts and washers on each end. Bottom tube is bolted to legs and floor to prevent tipping.

press and paper delivery drawer which is shown partially open over the press bed

Next step: assemble the vacuum plenum for the press, assemble the bottom printed paper receiving drawer, and finally decide how to get the drawer to open and close while I hold the paper registered to the block with both hands! Getting REAL close to printing again, at last!

By the way, the press 'action' is a dream come true -- glides so easily and surely across the bed regardless of pressure setting -- I hope the thing PRINTS every bit as smoothly and beautifully as it seems to operate!

-- Mike

Labels: equipment, Woodblock

  ¶ 8:41 AM 4 comments links to this post

  Press almost complete! Several people have asked how the press is going... It's really complete, now -- just waiting for some materials for the paper-handling gizmo to arrive in order to start working! -- Here's a photo:

-- Mike

Labels: equipment, Woodblock

  ¶ 12:25 PM 4 comments links to this post

  Drive section assembly complete... HEY! Last night and this morning I fitted the drive assembly and bearing guides for the top roller -- and amazingly, everything fit like a charm -- even the keys fit the keyways in the shafts and sprockets! It was especially gratifying when I installed the chain around the two sprockets (what a MESS of grease those chains are!) and the connector link just slipped in precisely -- what a perfect fit -- the result of all those difficult chain-ring calculations!

close-up of press drive

back-side assembly

drive side bottom view

top roller bearing guides test fitted OK

my very messy installation this morning

But one serious problem has now become apparent -- I originally conceived the press with the 1/8 inch bed held (like a drawer-bottom) by the inside slots of the main rails all around. Somehow I managed to under-dimension the press bed by about 3/8 inch, so it is able to slide around quite sloppily, and there's sufficient deflection side-to-side in the main rails to allow the bed to slip out of its track in the center. I've got 1/2 inch between the ends of the rollers and the side rails, though (to make it easier to keep the press bed swept free of wood chips and other debris), so I'll repair the problem when I attach some lengths of 1/4" x 2-1/4" plate to the inside of each top rail and then screw the bed down tight to those, using some flat-head machine screws countersunk into the bed very close to the rails. That should serve to stabilize both bed and side rails without too much work and without interfering with the smooth passage of the rollers across the bed.

-- Mike

Labels: equipment, Woodblock

  ¶ 11:48 AM 0 comments links to this post

  Press bed in place and frame leveled Here's a photo from last night after installing the bed and leveling the press in place -- closer and closer!

-- Mike

Labels: equipment, Woodblock

  ¶ 6:57 AM 1 comments links to this post

  New press stands on four legs! This afternoon, the new press finally stands up on its own four legs and just after this photo was shot, I mounted the main end-plates and the bottom roller now sits safely in its bearings hanging in position on from the press rails... Everything has fit perfectly so far... I'm waiting for the inevitable 'disaster', of course, but it seems to be coming together incredibly quickly this week!

Mike Lyon takes a quick breath after assembling the press onto its legs.

See previous description of this project

Labels: equipment, Woodblock

  ¶ 4:36 PM 2 comments links to this post

  Large format Woodblock Printing Press This is a progress report (with many photos) showing a long-term project underway -- my design and manufacture of a large (about 5x10 foot) woodblock printing press for shin moku-hanga (new Japanese woodblock prints).

Late in 2004, I decided to produce large woodblock prints -- larger than I was able working alone and using a baren... Of course, I'd considered using assistants, but prefer to work alone and on my own schedule. So I began to think seriously about what sort of 'press' I might like to use. I tested the method on some medium scale (32") blocks for "Anthony" using my 30 x 60 inch bed etching press and the very lightest pressure and it worked much better than I'd expected, so I began by researching existing large intaglio presses (I'd first looked into the possibilty of building a huge hydraulic plate-press, but it appeared too heavy and much too costly). Large intaglio presses are also quite expensive, and the best were specifically designed for very thin plates and didn't have sufficient room between the bed and the top roller to accommodate both my blocks and the vacuum plenum I require to hold them flat. None of the manufacturers seemed willing to modify their designs to my specifications -- that came as something of a surprise to me, but I understand their reluctance better now. There were also some significant paper handling problems to overcome while registering large papers single-handed (to be described in a future posting).

Because of space considerations in my studio, a moving bed press was undesirable (requiring a footprint twice the length of the bed, so I first tried to design a very stiff bedded press with a single roller. Deflection of the bed seemed to me to be an insurmountable problem with this method, so I decided a more conventional two-roller scheme would be easier to engineer. I ultimately decided to design a somewhat unique two-roller 'wringer' type press in which the bed would be stationary and the rollers could move back and forth across the bed. Because the rollers work in opposition, the bed could be very thin -- just stiff enough to support the block prior to printing. I got some 'help' from Doug Forsythe's $25 plans to 'build your own etching press' -- these made for a more confident start!

One of the most important decisions was exactly 'how' the rollers could be made to accurately move back and forth along the press bed. There are many options for 'linear motion' solutions -- after much consideration and comparison, I selected some off-the-shelf components from Bishop Wisecarver (BWC.com) who not only provided the excellent (and expensive) aluminum extrusions I chose for the frame components, but also the compatible carriage and rail system which will convey the rollers along the frame rails.

carriage and rail illustration from Bishop Wisecarver

After many false starts and much consideration about exactly how large I might want to print, I decided to dimension the press to enable printing full size blocks cut from widely available 4x8 foot 'standard' sheets of plywood.

drawing of the final press plan

Getting the measurements right, figuring out how long things should be, how to make everything fit together, especially in the drive section where a number of parts had to be coordinated was challenging! For example, consider the two sprockets and chain in the drive -- that has always seemed so simple... Until I tried it! There are a HUGE number of choices of chain sizes and sprocket sizes, and after figuring out what weight of chain would work, then there's the desired mechanical advantage, and finally, you can't just measure chain to the nearest 16th inch, the links are fixed, so you have to calculate how many links, then the shaft centers can be figured and fit exactly... LOTS of new stuff to learn and calculate -- I must have tried 100 different combinations of chains and sprockets before deciding on one set which would work... In the process, I gave up on the configuration which would probably have been best -- placing a light weight chain inside each of the long rails which support everything, then meshing a sprocket into that chain on each side to drive the rollers along... That would'a been very cool, but I wasn't patient enough to redesign EVERYTHING in order to make it all fit... Compromises, compromises...

detail of the plan showing the rollers, bearings, sprockets, etc... kinda hard to read, eh?

So eventually I decided on a design I thought would work using as many 'off the shelf' components as I could find, and making the rest. Because I thought I could machine aluminum myself using the same CNC machine (from ShopBot Tools) I've been using to carve large blocks since March, 2004, I designed many of the parts to be cut out of 1/2" aluminum -- here's the drawing from which the parts were machined:

aluminum parts machining plan

Below is a photo of the machining in progress -- it was more difficult than I'd imagined, and I broke a few bits out of ignorance, but if I have to do it again, I'll be able to accomplish it much more easily. Live and learn!

machining the aluminum parts

view of machine and completed aluminum parts

the completed aluminum parts along with some of the bearings, handles, etc to be mounted

OK, I got a little fancy and engraved a logo on the faceplate, then filled it with pigment laden epoxy and sanded it down flush -- kinda pretty!

The rollers were another challenge... The bottom roller had to be machined to very close tolerance in order to exactly engage the press bed. I was also concerned about potential deflection in the 1.25" center shaft and performed some fairly complex calculations in order to satisfy myself that deflection would be minimal under load. Eventually I settled on a design which I had bid at several local machine shops. My design called for four large 'washers' to be welded to the shaft and the inside of schedule 80 black pipe, then for the pipe to be turned to precise dimension. But it turns out that schedule-80 pipe is unsuitable for turning, as it contains stresses which won't allow an accurate turning to be produced, so I had to go with much more expensive seamless DOM (drawn over mandril) tubing.

the drawing for fabrication of the rollers

the BEAUTIFUL nickel-plated rollers themselves!

view of assembly area, the 10 foot long press bed is at the top in the background, the aluminum extrusion rails are at the left, the two nickel plated steel rollers are visible as well

Detail of the roller ends -- a local machine shop did a fantastic job fabricating these, even turning down the welds to make smooth chamfers! They are almost stand-alone works of art themselves!

So I am finally actually assembling the thing! So far everything fits perfectly, and I'm feeling optimistic that it's all going to really work! The rails and carriages have been assembled to the frame sides, seem to work as I'd imagined, and look good!

Assembly detail showing BWC carriage mounted to carriage rail and carriage rail mounted to press rail, and press rail attached to leg and brace.

Total cost has been about $6,500 (yikes! But still SO MUCH cheaper than even a well-used intaglio press of similar size)

To be continued (Mike Lyon -- 10/31/2005)...

Labels: equipment, Woodblock

  ¶ 10:47 AM 4 comments links to this post