Best Of Pennsy Position-Light Signal Head - Shapeways Project?

[towl1996]

Jim,
Thanks for the link. Waiting to see how the 'six pack' turns out, definately the way to go. And the back to back sounds interesting.

[nscaleSPF2]

towl1996,

Have not forgotten about this.  Report to follow shortly.

Jim

[nscaleSPF2]

Sorry it took me so long to report.  It took a couple of days to figure out what happened to the "six pack" and what if anything can be done to fix what happened to it.

Dear reader, You will remember that I promised to create a model that contained of 6 of the latest signal heads in this thread, and place an order for them.  This was because the 6 could be bought for almost the same price as a single head.  Anyway, I created an array of 3 by 2 heads, and oriented them the same way that the previously printed single heads were oriented.

When I received the six pack, I discovered that Shapeways reoriented the model.  Instead of the front of the head being oriented up (as most of the previous parts were printed), the tops of the heads were oriented up.  How do I know this?  Because the fronts of the signal heads had the "box-car side" striations that others have noted in other posts.  The striations are only located on the front face directly under the hoods, and the bottoms of the hoods are also pretty rough.  Bummer.  At least the basic geometry of the parts looked the same as those previously printed.

Have I mentioned before how frustrating it has been to work with Shapeways?  I think so.  Please allow me to bash them again.  There, that felt better.

The question now is what to do about this.  Given that the printer volume is 127 x 178 x 152 mm (from the Shapeways website.  I am assuming W x L x H) it would take about 3,000 heads to fill this volume.  Not a practical solution, since even then there is no guarantee that Shapeways would not turn the 3,000 piece model upside down.

I'm fresh out of ideas here, guys.

Anyone know where we can rent one of these printers and a technician to run it for a day?

Anyone know of another company that owns one of these printers and actually cares about their customers?

Other ideas?

Jim Hale

[nscaleSPF2]

Oh, oh, I got it!  Amazing what the human brain can come up with.

Instead of a 3x2 array of heads, how about a 2x2x2 array.  A cube, if you will.  Two rows of two heads, with another two rows of two heads positioned above them.  If I make this model correctly, there will be no incentive for Shapeways to rotate it.  They only rotate it to make more space for themselves in the printer.  I'll try this, and post the results.  That is, if there is anyone still following this thread.

Jim Hale

[VonRyan]

Oh, oh, I got it!  Amazing what the human brain can come up with.

Instead of a 3x2 array of heads, how about a 2x2x2 array.  A cube, if you will.  Two rows of two heads, with another two rows of two heads positioned above them.  If I make this model correctly, there will be no incentive for Shapeways to rotate it.  They only rotate it to make more space for themselves in the printer.  I'll try this, and post the results.  That is, if there is anyone still following this thread.

Jim Hale

I'm still following this with much interest.
I just don't have much to contribute since I don't have any technical knowledge to contribute, but I agree that the cube may indeed be the answer to the problem.
The plan for the layout I'm currently intending to construct doesn't need any signals, but I'll certainly be looking to get some of these in the near future.
I figured I should start building up a fleet of signals now so that they're ready to go when I need them.


-Cody F.

[John]

I ordered some type D signals from shapeways .. they have 48 of them on one job .. you might contact the designer and see how he did it ..

[nscaleSPF2]

I ordered some type D signals from shapeways .. they have 48 of them on one job .. you might contact the designer and see how he did it ..

John, I'll be most interested to see how these turn out for you.  Please let us know.  Thanks, Jim.

[John]

John, I'll be most interested to see how these turn out for you.  Please let us know.  Thanks, Jim.

I think Phil said he had ordered some ..

[3rdboxcar]

Oh, oh, I got it!  Amazing what the human brain can come up with.

Instead of a 3x2 array of heads, how about a 2x2x2 array.  A cube, if you will.  Two rows of two heads, with another two rows of two heads positioned above them.  If I make this model correctly, there will be no incentive for Shapeways to rotate it.  They only rotate it to make more space for themselves in the printer.  I'll try this, and post the results.  That is, if there is anyone still following this thread.

Jim Hale

It is still possible for it to be rotated although unlikely to happen, there is a way to ensure the print is rotated the way you wish but it makes the cost nonsense [you have to make 3 axis from an origin the size of the bounding box then position your heads in the corner of the axis].

Alexander

[nscaleSPF2]

It is still possible for it to be rotated although unlikely to happen, there is a way to ensure the print is rotated the way you wish but it makes the cost nonsense [you have to make 3 axis from an origin the size of the bounding box then position your heads in the corner of the axis].

Alexander

Alexander, I thought of this possibility, also.  But would it not also be possible for the technician at Shapeways to invert this kind of model?  Unlikely, yes, but possible?

Jim

[3rdboxcar]

Alexander, I thought of this possibility, also.  But would it not also be possible for the technician at Shapeways to invert this kind of model?  Unlikely, yes, but possible?

Jim

Yes they could still invert the model but they could not put it on it's side which is what you do not want, using the axis theory means the surface you want printed horizontal would be horizontal whether it was right way up or inverted 180 degrees.

Since your heads are small a cheaper option would be to make a sprue with 50% of the heads at right angles to each other, that way you should be guaranteed to get 50% printed with the horizontal face you want. [if that makes sense]

So far everything I have had printed is either 0, 90, 180 or 270 degrees so it is a 50 / 50 chance of getting the face you want printed on the horizontal.

Alexander

[nscaleSPF2]

Alexander,

Wow, this is a big revelation to me.  Having such a high scrap rate is not desirable, but ok if that is what it takes to get a good part.

As far as the signal head in this thread, I found that it was not ok to have the front face "inverted".  Some of the early parts were printed inverted and the hoods had a rough, striated surface.  You can see the photos earlier in this thread.

Thanks very much for the information.

Jim

[unittrain]

Which LEDs do I need for these I got some yellow ones from Ngineering a ways back will have to check size. What is the best size to use.
Thanks

[peteski]

Which LEDs do I need for these I got some yellow ones from Ngineering a ways back will have to check size. What is the best size to use.
Thanks

Earlier in the thread it was stated that the signals are designed for SMD 0603 LEDs.  Industry standard for surface-mount LEDs with 0.060" X 0.030" footprint.  You can buy those for pennies each in quantity from many electronic suppliers (such as Mouser or Digikey).

[nscaleSPF2]

I can report success regarding the "8 pack" of signal heads printed by Shapeways.  Close inspection showed that these samples are as good as any of the previous ones.  Here is the link to these parts.

https://www.shapeways.com/model/1235465/signal91.html?li=productBox-search

For those of you who are considering ordering these, you might want to wait.  Bob Knight reported in the week-end report that he is close to opening the TrainCat shop again.  This may include his much slicker version of the PRR position signal.  Bob, if you are reading this, then I sincerely wish you the best and hope that your shop becomes active again.  All of us N scale addicts appreciate your fine efforts.

Jim Hale