MicroMimesis is the name for my company (to be). Right now I only have a store set up on Shapeways .. but .. I DO have the Domain Name – MicroMimesis.com registered. My plan is to supply a few items for the Model Railroad and Military Modeler. At some point I need to set up a website/store .. but .. need products first!

Locomotive Numbering System

I have a new loco for my layout . “On30 Modern 4-4-0 American outside frame loco (DCC) .. c. 1920’s. I was thinking about weathering it and realized that it needed lettering and numbering before even starting the weathering. That got me to thinking about the number on the cab.

There was a discussion on the Model Railroader forum on this – Locomotive numbering

So .. I am going to use that as a guide. I’ll do that in a Page to keep it easier to find than a page like this.

What am I up to?

what am I up to?  More like what I’m NOT up to.   I have a layout in progress, multiple P48 conversions in progress, family and work.  Ed’s giving me a space here to post my progress, as little as it may be.


Take care

MeshLab DAE export error when importing into Sketchup

Sketchup model.

  • Length Units Decimal Inches
  • Created a cube 25.4 x 25.4 x 25.4 : cube25p4onside.skp
  • Exported as a .date : cube25p4onside.dae
  • Imported the mesh into MeshLab
  • MeshLab: Filters » Cleaning and Repairing » Remove Duplicated Vertex
    — Results were 16 duplicated vertices removed. Explanation : Each face of the cube has four vertexes. Where three faces join there is a corner shared by all three faces – a common vertex. For the cube that means the same vertex has been duplicated three times so two are redundant. A cube has eight corners or sixteen redundant vertices.
  • Exported the mesh as cube25p4onside.stl
    — A STL file does not contain data stored as a specific system of measurement – inches, mm etc. It is simply stored as numbers. For this cube then it is stored as 25.4 .. that is all. The trick is that Sketchup exports in inches without regard to the Length Units the model is created in. If you created a cube with Sketchup set to Decimal mm you would have a cube 25.4 mm on a side. When you export the DAE file it would be in inches .. 1 x 1 x 1. By creating the model in Decimal Inches but using mm for lengths .. we have a cube that actually is 25.4 x 25.4 x 25.4 INCHES. Exporting the DAE and importing into MeshLab then saving as a STL we have simply a file with sides of 25.4 UNITS. When we upload to Shapeways we simply tell them that the model is in mm. There are two reasons for doing this.
    — (1) Sketchup has a problem with radii smaller than 0.018″ – it will refuse to draw a circle and has problems with interseting mesh at that size. In O scale that 0.018″ is equal to 0.864″ Full-Size. Obviously this would be a problem. If we multiply that 0.018 x 25.4 we get 0.4572 which obviously is way outside the 0.018″ radii issue. In fact if we close one eye and look at it sideways .. with us modeling in Decimal Inches but using mm lengths .. that Sketchup minimum radii issue of 0.018 (call it Pseudo mm) converts back to 0.0007″ for O scale or 0.034″ Full-Size. QED – Problem Solved!
    — (2) The Shapeways materials information pages give you the information to design a model for 3d printing. In this are minimums for each material which include things like ‘Minimum Supported Wall Thickness” and “Minimum Supported Wire” and “Minimum Embossed” amoung other data. This is all expressed in milimeters. That means we can quickly check the width of a wall to ensure it meets the minimum for the material we are working in. If I am designing for FUD (Frosted Ultra Detail) which is my “Go To” matrial for my projects, the minimum wall thickness is 0.3 mm – which again means that I simply have to use the ‘Tape Measure Tool’ to make a quick check that I am within the minimum.
  • * Exported the mesh as cube25p4onside.dae
    — This is one of the points for this exercise.
  • Imported cube25p4onside.dae into Sketchup
    — Now we use the ‘Tape Measure Tool’ and we see that with our Sketchup universe set to Decimal Inches .. the cube measures 1000.000 on a side. What is up with that? To continue let’s use netfabb
  • Open cube25p4onside.stl in netfabb.
    — The netfabb dashboard sees the cube as 25.4 mm on a side (it defaults to mm on import). In netfabb we can export a STL file as STL (ASCII) – so let’s do that – cube25p4onside_ascii.stl – which we can open in Notepad.
    — cube25p4onside_ascii.stl – just a bit of the code – the first 15 lines:

    solid cube25p4onside
    facet normal 0.000000 0.000000 -1.000000
    outer loop
    vertex 25.400000 25.400000 0.000000
    vertex 0.000000 0.000000 0.000000
    vertex 0.000000 25.400000 0.000000
    facet normal 0.000000 0.000000 -1.000000
    outer loop
    vertex 0.000000 0.000000 0.000000
    vertex 25.400000 25.400000 0.000000
    vertex 25.400000 0.000000 0.000000

    — We can see from this that within the STL file we have the vertices numbered as 25.400000 – again .. when uploading a file to Shapeways telling the website that the file is in milimeters works great

  • So the question is .. how did MeshLab get from a STL file with a cube of 25.4 units on a side to exporting a DAE file with 1000 inches on a side?
    — Let’s open up the DAE file as exported from MeshLab and look at the two ‘source’ loops ..

    <float_array id=”shape0-lib-positions-array” count=”24″>25.4 25.4 0 0 0 0 0 25.4 0 25.4 0 0 25.4 0 25.4 25.4 25.4 25.4 0 0 25.4 0 25.4 25.4</float_array>
    <accessor count=”8″ source=”#shape0-lib-positions-array” stride=”3″>
    <param name=”X” type=”float”/>
    <param name=”Y” type=”float”/>
    <param name=”Z” type=”float”/>
    <source id=”shape0-lib-normals” name=”normal”>
    <float_array id=”shape0-lib-normals-array” count=”36″>0 0 -1 0 0 -1 0 1 0 0 1 0 1 0 0 1 0 0 0 -1 0 0 -1 0 -1 0 0 -1 0 0 0 0 1 0 0 1</float_array>
    <accessor count=”12″ source=”#shape0-lib-normals-array” stride=”3″>
    <param name=”X” type=”float”/>
    <param name=”Y” type=”float”/>
    <param name=”Z” type=”float”/>

    — First Loop: Appears that it is happily reporting 25.4 units. It referrs to – array count=”8″ – “-positions-array” – accessor count=”24″ (followed by 24 numbers, 12 of which are 25.4 and 12 which are 0. Since a cube has 8 vertexes and each needs a x,y,z coordinate I am *assuming* that count=”8″ is the number of vertexes and count=”24″ is the x,y,z coordinates of each of those vertexes.
    — Second Loop: This time the array count=”36″ followed by 36 numbers – 0, -1 or 1 – and the accessor count=”12″ – with a reference to ‘normals’. A normal is if you are looking at the outside or inside of a face (3d talk). It was at this point my brain began to shut down. A cube has six sides so I can *guess* (close cousin to assume) that the count=”12″ is referring to the outside and inside of each of the six faces. What about the count=”26″ .. well .. something to do with the vertexes of each of the 12 outside/inside faces .. and my brain tried to do a re-boot at ths point.

  • So .. we already know that Sketchup exports a DAE file as inches no matter what Length Unit you modeled in. Importing a DAE file back into Sketchup it is doing something different .. obviously, changing that 25.4 to 1000.
    — if 25.4 * X = 1000 …. then …. X = 1000/25.4 .. so X = 39.37~ .. and why does that seem familiar. Ah Ha! 1 meter = 39.3701 inches.
    — 0.0254 meters = 1 inch
    — AND multiply both sides by 1000 and we get 25.4 meters = 1000 inches. Soooo. Sketchup is taking that DAE file adn thinks the 25.4 is meters and converting to 1000 inches. Huh. That made no sense .. I can export a DAE file from Sketchup all day and import it back without the units getting mixed up .. and .. and .. Ooooo. Exported from Sketchup .. the DAE file I imported was created in MeshLab … (pause for effect)
  • The files are different with the Sketchup DAE file adding various Sketchup specific items such as materials. What caught my eye was this:
    — from the MeshLab exported DAE file:
    — from the Sketchup exported DAE file:
    <unit meter=”0.0254″ name=”inch” />
    — The change in which axis is UP is no problem .. BUT .. look at the additional line that Sketchup added .. . To which I go .. Okaaaayyy. This DAE file was exported with a cube 25.4 inches on a side with the Units set to Decimal Inches (reason in the first paragraph).
    — So I went back to Sketchup and created a cube using Units as Decimal Milimeters and exported the cube. The line is exactly the same. The difference is that the units exported are 1 and 0. The units exported when the Units are Decimal Inches are 25.4 and 0.
    — So. With Units as Decimal Millimeters the length is 1. Sketchup converted the 25.4 mm internally before export ..and then exported as inches .. so the length of a side is 1. When I had Sketchup set to Decimal Inches but sides to 25.4 it exported the Length of a Side as 25.4 .. in both cases then when importing the DAE back into Sketchup it uses that bit of code. To repeat .. <unit meter=”0.0254″ name=”inch” /> .. or let’s look at it in mm ..

To repeat without all the Gobbly Gook ..

The problem is that MeshLab exports what appears to be a vanilla COLLADA file. Sketchup uses their own version which adds information such as for materials. As was stated by TIG the Sketchup DAE file has that line .. < unit meter=”0.0254″ name=”inch” /> which is not in the MeshLab DAE file. Without this Sketchup goes nuts with the sizing as it has no instructions.

I exported a 1 inch cube as a DAE .. brought it into MeshLab .. conveted to STL and saved as STL. Reloaded the STL and then exported back to DAE. I then imported that MeshLab DAE into Sketchup. The cube NOW measured 39.370 on a side (some rounding going on here by the way)

That is when I saw the missing ‘conversion code’. If you rework that formula a bit for millimeters you get <unit millimeter=”2.54″ name=”inch” /> .. which in a roundabout way let me to – 1000/25.4 = 39.371 (and change). which is the number of inches in a meter.

Sooo. Take the crazy dimension you get on a MeshLab DAE after importing into Sketchup and divide by that 39.370 (applied the rounding) to convert back correctly.

Woofer Worries

ETraxx_woofers_finish_3Once upon a time .. long .. long ago .. I made some custom speaker pods for my 350Z.

Sometime in the past one of the woofers died. Yesterday, I went looking for a replacement.

The original woofers were Rockford Fosgate P28S8 – 8″ Punch Stage-2 (8 ohm) subwoofers. These were chosen as I could tuck them away nicely in a space on either side of the rear strut and not lose any ‘trunk’ space.

The choice of these speakers was in parallel with the choice of amplifiers. I used a Profile CA600M amplifier to power the subs. The CA600M RMS Power @ 4 Ohms is 1x300W (bridged). The Punch Stage 2 subwoofer is rated at 200 watts RMS/400 watts peak as stated above so fits nicely within the CA600M RMS of 1x300W bridged (150W per speaker). I chose 8 ohm speakers so hooked in parallel so the amp shows a 4 ohm load.

Etraxx_woofers_MDF_1This was years ago and I found a Rockford Fosgate P2D4-8. Everything looked good until I looked at the dimensions. This speaker has a mounting dia of 8.72″ with a bolt hole dia in that mounting plate of 8.03″ .. and finally a trim ring dia of 8.96″. The problem is .. my pods have the speaker/trim ring recessed .. and that older woofer’s recess is 8.375″ dia.

Well Darn.

Google Book Links

Google Books
As long as you live in the US you have access to a vast array of information in Google Books. I usually modify the search so that it displays only Free Google Books .. and often set a range on the dates. This is .. really .. an amazing resource for those of us modeling the late 19th and early 20th century.
The California Lumber Merchant – March 15, 1923
pg 35 – ATTEMPT TO SALVAGE CARGO – Attempts will be made by the San Diego Motorship Vaquerro to salvage a part of the lumber cargo of the four masted schooner Watson A. West which was wrecked on San Miguel Island last week.
Pacific Motor Boat, Volume 8, October 1915
pg 21 – shows a side and top view of a Five Masted Auxiliary Lumber Carrier. The description has her at 225 ft. long, 32 ft. 6 in. beam, 18 ft. draft and have a capacity of 1,500,000 ft. of lumber ..

Tofu Brains

Well .. I’m told that “A mind is a terrible thing to waste”. That may be true, must be true, surely is .. true. It is also terrible when what brains you DO have turn to tofu.

I am renovating my house and have been for a couple years now .. in between doing “other stuff”. Yesterday, I was in the middle of putting sheetrock up in my shop. (my shop. My house is two stories .. living upstairs and the ‘shop’ downstairs. It was a garage at one time until some past renter ran her car through the wall. It was repaired as a shop, storage area .. whatever.)

I stopped without putting up the last piece of sheetrock as I needed to finish wiring to the switches at the bottom of the stairs that controls the outside light and the light up on the porch. It made sense to finish the wiring before continuing the sheetrocking as I still had access to the back of the electrical boxes and easier to run wire when you are not staring at a blank wall. Just saying.

Got everything hooked up .. the exterior light working. Check. The three-way switch that controls the light upstairs on the porch .. not so much. Checked .. “ummm .. yeah .. my voltage tester showed I had the switch wired correctly. Went upstairs and inside .. checked the three-way switch there. Correct. Getting juice where I am supposed to when switches are thrown .. but .. no light.

Pulled the light off and checked the wires. No power.


Ahhh. My Tofu Brains.

Tracing wiring ..crawling around in the attic .. and I pieced together the backstory. I had forgotten that originally power was routed from the overhead light between the Kitchen and Living Room over to a switch on the wall to control it .. AND .. to the three-way to power the porch light.

.. and .. the way it was wired it is in series not parallel. That’s why it isn’t coming on. I could fix that by completing the circuit up in the attic but not going that route. I need to replace the wires anyway. I’ve already replaced all the wiring except for the kitchen area (where the switches are). The house was built about 60 years ago and the wiring didn’t have a ground wire. So .. today I will be cutting into the wall to get access to the old wires so I can run NEW 14-2 w/g NM cable.

Since this means I will be crawling in the attic, burrowing though insulation .. and HEAT .. this promises to be a nasty day.

Need .. coffee!

Open Boats

tcomobI have a small book titled “The Construction of MOdel Open Books” by Ewart C. Freston. Published in 1975 by Conway Maritime Press Limited, Greenwich, London.

This was some 38 years ago and in the text the author mentions that he uses “Seccotine” as an adhesive. According to Wikipeda this is a “(…) a brand of refined liquid fish glue used for gluing paper and card and as a binder in gesso, which remains flexible after drying. It is also used in the mounting of preserved insect specimens, as it can be dissolved in water if the specimen must be removed for further investigation.

You can still get it evidently – Seccotine

In any case .. I got sidetracked there. I was looking over the bit in the book where the author is talking about hollowing out the shell. In the past this part has puzzled me – as it (to me) made little sense. It says ..

Surprising as it may seem, with a boat such as the one I am describing, a thickness of 0.002″ can quite easily be reached, and with practice even less. The shell of the 28’0″ New Bedford Whaleboat, of which there is a photograph in the Introduction, has a thickness of 1/2″, which, at a scale of 1/4″ = 1’0″, is 0.0011″.

I was just looking over that again .. when I focused in on the passage “a thickness of 1/2″, which, at a scale of 1/4″ = 1’0″, is 0.0011″.

Well. It seems that whoever did the proofreading for this book must have been “math challenged” .. 1/2″ at a scale of 1/4″ = 1’0″ is 0.011″ .. NOT .. 0.00011″ .. so the earlier statement of “(…) a thickness of 0.002″ can quite easily be reached (…)

Now .. I am MUCH less puzzled .. at least by a factor of 10.

MicroLux Tilt Arbor Table Saw

85870_RSo .. I broke down and bought the MicroLux Digital Table Saw from Micro-Mark. I have been cutting strip-wood to length for years .. first using various miter boxes with (for me) limited success. Later I modified a Harbor Freight chop saw to do that with better results. I have also purchased relatively serious quantities of scale lumber online for some relatively serious damage to my bank account. It occurred to me that in addition to the other things it could do .. cutting my own scale lumber would pay for the saw in time. When I looked at the saw at Micro-Mark I saw (sic) that it was “New and Improved” ….
New and Improved at the Same Low Price!

The MicroLux Tilt Arbor Table Saw is more powerful, more accurate, and smoother running than any other saw its size. It has a variable speed motor, so it will cut metal and plastic as well as wood. And now, the rip fence is equipped with a DIGITAL readout . . . set-ups for accurate rip cuts couldn’t be easier! This is, without a doubt, the most feature-packed machine ever designed for the scale modeler. Includes a calibrated miter gauge with its own adjustable fence, blade guard, anti-kickback attachment, and 80-tooth fine-cut steel blade for balsa and basswood up to 1 inch thick, hardwood up to 1/4 inch thick (hardwood up to 1 inch with optional carbide blade). Blade diameter 3-1/4 inches with 10 mm hole.

84904_RI had placed my order and then at some point I was looking at the accessories. The main one that I had been interested in was the plastic inserts – the blank blade plates. Since the saw is a tilt-arbor there is a gap in the blade plate to allow the saw to tilt. The problem is that this gap allows thin strips of wood (scale lumber) to fall through that gap. The answer was to replace the stock blade plate with these plastic blank plates and then by raising the blade through the plastic cut a nice, narrow slot which allowed cutting that costly scale lumber.

In the photo we have the old saw #80463 showing the plates. Note that the plates secure with a couple of screws.

85870~3_R.. and then we have the “New and Improved” version. The blade plate is shaped differently .. and most importantly attaches differently. No screws .. but similar to the way a battery cover snaps into something. There are a pair of hooks on one end that you engage, then drop the plate down so that a catch depresses and snaps into place. The question that comes to mind immediately is .. how can a blank (read plastic) plate replace this to cut stripwood? The plate sits on a recessed ledge and it might be possible to drill and tap a threaded hole at either end to mount a plastic blank – but it would be nice in Micro-Mark had a blank plate to use without any such modification.