• The Plug Track functions are experimental and still being developed. Some of the earlier pages of this topic are now out-of-date.

    For an updated overview of this project see this post.   For some practical modelling aspects of using Plug Track see The Book of Plug Track.

    Some pages of this and other topics include contributions from members who are creating and posting their own CAD designs for 3D printing and laser-cutting. Do not confuse them with Templot's own exported CAD files.

Experimental Plug Track: 3D-printed, CNC-milled, laser-cut

Quick reply >

Hayfield

Member
Location
Essex
Ralph

Whilst it never took off Exactoscale produced turnout kits in P4, they were highly detailed but the design whilst allowing ( or would it be correct saying needed gauges to be used owing to the assembly design not being 100% accurate) a degree of adjustment prior to gluing, has proved in my mind that the concept is sound. If the rails are correctly formed and positioned there will be no need to adjust anything

I have had the opportunity to try one of Wayne's EM gauge turnout kits and had the pleasure of trying out one of the prototype 3D printed switch sections, and have come to the conclusion this method may well be the answer of having the basis of an easy to build system to make turnouts and crossings. In fact far more flexible than any previous designs as each turnout/crossing is bespoke

The one area left to be built by the builder requiring skills is the rails, though I assume if this method catches on some enterprising modellers will be able to supply preformed vee and switch rails. But this is the one area that is left to the skills of the builder.
 
_______________
message ref: 5326
Location
Manchester
I am sure you are right but I have to confess to never building anything other than ply and rivet. The cost of these turnout kits has put me off trying anything else as the costs for a ply and rivet turnout are minimal in comparison. I find it very encouraging that there are kits available for newcomers to finescale and I wish these manufacturers every success, they deserve it after all the work they have put in perfecting their techniques.

Ralph
 
_______________
message ref: 5327

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
On both the BIBO (and the MINIBO with Templot's backlash correction function) I've achieved very good accuracy by careful measurement and adjustment of the shrinkage settings, better than +/-0.1mm. It's important for the socket sizing of course.

I shall be very disappointed if I can't achieve the same on the Neptune by tweaking the settings. As usual it's proved impossible to get this sort of information from any website or by asking anyone, the only way to find out is to get one and see.
.
Neptune 2S now assembled and working. :)

First off was to get one of these, which has just arrived:

https://www.amazon.co.uk/gp/product/B0855DBVXF

The card slot on these printers is for a MicroSD / TF card, as the edge connector on the internal control board. Those TF memory cards are hardly big enough to see, let alone handle, so the adaptor cable was essential for me. I can now use ordinary SD cards. For the present I have installed the Neptune in the computer den, so I don't actually need to use the card, I can control the printer over the USB cable from Cura on the computer:

neptune2s_printer.jpg


But it can't stay there, it is in the way, so it will have to move to the 3D-printing den eventually, which is too far from the computer and it will have to run off the card. It's better that way anyway to avoid computer glitches and having the computer tied up for hours.

The good news is that the timbering bricks are just as good as from the BIBO, chairs fit the sockets fine, and at around +/-0.02 mm that is 5 times better than the Elegoo claim of +/-0.1mm . Perhaps that relates to other polymers than PLA.

I used Cura 5 rather than the bundled version of Cura, and the same settings that I use on the BIBO. The only change was to increase the amount of retraction -- this is my first Bowden tube extruder, the BIBO and MINIBO are both direct-extrusion.

Assembly took several hours to get everything square and straight to my satisfaction -- starting with disassembly of the parts supplied poorly assembled, and starting again at the beginning. :(

Once assembled it is working great, but I can hardly recommend it to someone new to FDM printing. It would be better to get one more fully assembled and ready to go -- assuming the assembly has not been done by the same person who did this one. :)

But it has proved the point I was unsure about -- this type of printer can produce timbering bricks of the required dimensional accuracy, regardless of the web site specs. The build plate is bigger than the BIBO, so I shall probably use it more often. I'm using grey polymer on here instead of brown, to identify easily which bricks are which.

cheers,

Martin.
 
_______________
message ref: 5328

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
Hi Martin,

To improve the resolution I swapped the X and Y steppers for ones with half the step angle and used the smallest belt drive cogs available on my (now antique) Folger printer. It improved the X-Y resolution a lot.

To minimize backlash the drive belts are held in tension with tie wraps.

Cheers!
Andy
@AndyB

Thanks Andy.

I'm not really trying to improve this printer for my own use. The object is to find out what is possible with the typical FDM 3D printer many Templot users have, and set the program defaults accordingly.

These printers all seem to be clones of the same basic design, using the same components. Some of the parts in the Neptune are identical to those in my 5-year-old BIBO printer.

They all use stepper motors having 1.8 degree steps, giving 200 native steps per revolution. They all use a process called microstepping where the motor current is shared between adjacent coils in various proportions so that the motor hovers, somewhat imprecisely, between two native steps. This reduces the torque available but increases the effective number of steps. Typically there are 16 microsteps between each native step. When it gets to the next native step the motor delivers full torque and recovers any imprecision in the positioning. This means that under light loads, such as in a 3D printer, you can reasonably use all the microsteps to determine the resolution in travel/mm. Under heavier loads such as in a machine tool the motor will tend to cog to the native positions, and you lose the resolution.

It's interesting that these small steppers use exactly the same 1.8 degree design as the much bigger industrial steppers which we used 40 years ago to convert a Bridgeport miller to CNC control. At that time there was only one 50% microstep available from the stepper-driver modules we used, giving 400 steps per revolution. It seems those 40-year-old industrial stepper drives are still going strong and sought-after:

https://www.ebay.co.uk/itm/125395898412

I designed the control system entirely around 400 steps/rev from the motor and a 2:1 reduction to the leadscrews, giving 800 steps/rev on the leadscrew.

Most of these 3D printers use the same Robin Nano stepper-driver control boards:

https://www.amazon.co.uk/dp/B0B8841Q5H

supporting the Marlin GCODE interpreter. Which is why I was puzzled that the Neptune isn't claiming the same resolution as the others. I am now satisfied that it can deliver it, even if it doesn't claim it.

The BIBO uses instead the Gen-L controller board with interchangeable stepper-driver modules:

https://www.amazon.co.uk/gp/product/B07Q7X9MJ8

but the same Marlin interpreter.

40 years ago there was no Marlin, and I wrote a fairly primitive control system on an 8-bit computer to convert the component design to motor steps.

I seem to have strayed off-topic. :)

cheers,

Martin.
 
_______________
message ref: 5330

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
.
News update from Neptune. :)

I couldn't get on with the textured flexible build plate. None of my test bricks really stuck to it properly, and two of them were wrecked when they warped and came loose mid-print. None of them required flexing the plate to release them, they could all be released in situ by being simply flicked off the plate shove-ha'penny style. I tried thoroughly degreasing the plate, with no obvious improvement. My guess is that it is intended for chunkier solid models, and not designed for large thinnish objects such as the timbering bricks. All the review videos show it working well, so I don't know what I'm doing differently. Maybe PLA Plus behaves differently from standard PLA?

Turning the plate over gives a smooth shiny metal surface, and using the glue-stick it worked fine in terms of getting the brick to stick.

But the main problem with the flexible plate is exactly that, it is flexible. And not rigid enough to be flat in its own right. It simply conforms magnetically to the underlying heater plate. If that's not flat, nor will be the build plate. In my case despite levelling carefully at the four corners, it was concave and hollow in the centre. Sufficient to prevent the brick from sticking properly in that area. No doubt that is the reason for introducing the auto-levelling compensation on the Neptune 3. But that does nothing to deliver a part having a constant thickness. The bricks were fine for X-Y dimensions, but all over the place on Z.

I was entirely happy with the removable glass plate on the BIBO, so I have now done the same on the Neptune. It is flat and rigid in it's own right, and is clipped to the heater plate in 3 places, two at the front corners, and one in the middle at the back (slightly off-centre to clear the Y-stepper motor):

https://www.amazon.co.uk/gp/product/B07SQCQ8KB

https://www.amazon.co.uk/gp/product/B001J8G6Q4

It should in theory have some packing at each clip, to clear all possible unevenness of the heater plate, but it is working fine without any on a concave heater plate. Probably I should put a delay in the start-up GCODE to allow for a longer heat-up time, if it is not in firm contact with the heater plate.

And the result is -- it works fine, exactly the same as the BIBO. Bricks now a constant thickness all over.

The build plate on the Neptune is actually 235mm x 235mm (as indicated in the Cura profile for it), so that is another departure from the quoted specs of 220mm x 220mm. I think it could easily be used at 230mm wide (X), but maybe keep the 220mm limit front to back (Y), to leave room for the above clips, if used.

cheers,

Martin.
 
_______________
message ref: 5332

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
Does that imply that a Neptune 2 (rather than a Neptune 2S) would be sufficient?
I think the S is for Spring plate.
Steve
@Steve_Cornford

Hi Steve,

I wondered what the S was for. I imagined it was Second version, as in some cameras. But in that case the Neptune 3 would have been Neptune T. :)

From: https://www.tomshardware.com/reviews/elegoo-neptune-2-3d-printer

"The mat on the Neptune 2 is a solid board of what appears to be epoxy-impregnated fiberglass with a textured mat applied to the top-facing side. This solid mat has an average thickness of 1.5mm, as opposed to the 1mm average thickness of the flexible mat. The mat is held in place with four binder clips, and has just enough flexibility to allow a printed part to pop off after printing."

That doesn't sound very rigid, but obviously stiffer than the thin spring plate. The presence of the clips explains the quoted 220mm build size instead of the full 235mm plate size. They missed a trick in not updating that dimension for the Neptune 2S blurb.

I think I would still want to change to a rigid glass plate. But maybe that's just me.

But the main change in the 2S is surely the Dual Gear Metal Extruder? I don't know if the previous one was single-geared, or plastic, or both, but either way the quality of the filament-drive is important in a Bowden-tube extruder. To create string-free prints it's important to use precise retraction of the filament, and Bowden-tube extruders are known to be inferior to direct-drive extruders in this regard. It was another area I wanted to check on. To get comparable results to the direct-drive BIBO, I have increased the retraction from 3.0mm to 9.0mm on the Neptune. It's working fine now, but was very stringy with only 3.0mm, with strings blocking the sockets.

Presumably the Neptune 2 extruder had been found wanting for Elegoo to be making a fuss about the improved one on the 2S:

"ELEGOO Neptune 2S is upgraded to long-lasting all-metal dual gear Bowden extruder, providing easier and smoother feeding of filaments, reducing slippage and missed steps while increasing precision and accuracy."

Are you actively looking at FDM -- I thought you were in the laser-cut plywood camp? I don't know the costs of out-sourcing laser-cutting, how much track does it need to justify a FDM printer instead?

p.s. I see the 2S has jumped £33 on Amazon since I ordered mine 4 days ago. :(

cheers,

Martin.
 
_______________
message ref: 5334
Hi Martin,
Just trying to keep an open mind, vis a vis, laser cut ply versus printed bases.
I have now found some ballast and Copydex to play with, and have been reading through the ballasting topic where you mention the method used on Adavoyle Junction.
Copydex sounds more like a duplicating machine!

Steve
 
_______________
message ref: 5335
Hi Martin,
The pack of reverse action tweezers that you recomended has arrived, and they seem very good value for money.
It will be interesting to see how your tailor made replacement tips for handling loose jaws work out.
Steve
Ps performed a "chair dance" in London Bridge station lower concourse but it seems to invoked the wrath of the rain gods instead!
Steve
 
_______________
message ref: 5337
Hi Martin,
SOme more feedback
I have tried some more resin track bases having adjusted some parameters:-
1668534976758.png


In particular the "clip bottom offset" has been increased to 0.5
1668535067936.png

I also adjusted the resin shrinkage rates down to zero to establish a baseline:-
1668535142682.png


I attach the fixed files (joint_green & plain_pink).
1668535378192.png


The joint segement consists of timbers A30 to A22, with 5 timbers to left of rail joint & 4 timbers to right of rail joint.
The middle of timber A30 is at x= 17.47
The middle of timber A22 is at x= 90.30

90.30- 17.47 gives distance apart of centres = 72.83mm
Timber width 3.33mm, so adding 2 half width timbers (ie one timber) gives distance apart of outside edges of timbers A30 - A22 of 72.83 + 3.33 = 76.16.

The bases were arranged in landscape mode across the build plate (two per slice), and having printed and cured for 4.0 minutes, I measured the outside faces to be between 76.15mm and 76.17mm apart using my digital calipers, naking sure I held the base down on a flat surafce.
Also most of the timbers measured 32.00mm long using my digital calipers, with a few either side +-0.01mm

I did warm the resin in a basin of hot water and gave it a good shake before filling the VAT.

In hindsight I made an error in judgement reducing the side flange width to 0.5mm, and also possibly the flange depth to 0.5mm, but I wanted to experiment and see the effect, as I managed to snap one of the tommy bars off.

Regards Steve
 

Attachments

  • oo-sf_joint_green_zero_fixed.stl
    366.4 KB · Views: 9
  • oo-sf_plain_pink_zero_fixed.stl
    251.2 KB · Views: 9
_______________
message ref: 5338

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
@Steve_Cornford

Hi Steve,

Many thanks for the feedback. You are getting very accurate results there, with no shrinkage allowance?

It's not surprising the clips snapped off -- they are not actually attached to the brick! - there is a small gap between the clips and the timber flanges (according to TurboCAD).

The intention with the clips is that you use the clip size mouse action to extend them into the actual timber a fraction, not just attached to the flanges. See:

https://85a.uk/templot/club/index.p...mbering-brick-from-a-track-plan.295/post-4041

index.php




Some progress with the tweezer tips today -- first iteration of the design resin printed. I made 3 pairs of tips, expecting them to fail to print, or break on first use. But the first pair is still going strong:

plug_track_tweezer_tips1.jpg


plug_track_tweezer_tips2.jpg


plug_track_tweezer_tips3.png

They are working quite well. You can grip the loose jaw pin and pull it off the raft, hold it upright, and position it over the chair by pushing the tweezer tips into the rail web. Then release it and it drops straight into the slot. Untouched by human hand -- or podgy fingers. :)

Or not.

It does work, but I have made it too chunky to actually see the slot! The pins now have a taper on the bottom, so most times you can feel them locate in the slot. But it would be much better if you could actually see it. Something of an oversight there. :(

The tips need to be much more slender, but then I'm not sure they will be strong enough. I may need to experiment with different types of resin.

But it's a promising start.

cheers,

Martin.
 
_______________
message ref: 5339

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
Your replacement tips for handling loose jaws look good.
Steve
@Steve_Cornford

Hi Steve,

I have now discovered (if that's the right word, I'm not Columbus) that you don't need both bolts on the tweezer shanks. The end one alone is sufficient. That makes it possible to rotate the tips to any desired angle. For example on plain track it is probably easier to come at the chairs from the side, whereas in a complex formation you would likely want to drop onto them from above, between the rails. There is enough "give" in the resin that the tips can be rotated in situ without needing to slacken the nuts.

Also having 4 tweezers means you can have a variety of 3 different tips set up for specific tasks. (The other being kept as supplied for general modelling, they are quite handy tools.)

One of my 4 is significantly stiffer to use than the others. I don't know if that is by design for the widest tip, or just poor quality control. Yours?

p.s. I have sent you an email.

cheers,

Martin.
 
_______________
message ref: 5343

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
.
Every day another setting. :)

neptine_bed_thermal.png


I mentioned that the Z dimensions from the Neptune 2S were all over the place. I have finally found what's going on. As the bed heats up, the bed rises a fraction. This seems to be caused by thermal expansion of the levelling studs. These are much longer than on the BIBO and solidly attached to the heater plate. This is presumably why I never noticed the effect on the BIBO.

The effect is to reduce the thickness of the printed component from the design sizes (assuming the levelling was done cold).

On the Neptune it takes about 30 minutes for the effect to become noticeable after starting the print, by which time some of the lower layers of the component are likely to have been already printed.

For the timbering bricks, the exact thickness of the lower layers isn't critical -- the splints, clips and flanges all get buried in the ballast.

But it's important that the total thickness of the timbers is known and consistent, so that timbering bricks printed at different times, or on different printers, do match with each other at the timber tops, and to ensure the plugs do not bottom in the sockets.

To achieve the design timber thickness of 3.2mm (1/8") on the Neptune 2S I have found it necessary to raise the print head by 0.24mm if printing after the bed has fully heated up. So for FDM printing with a heated bed, there are 2 new settings, as above. For Z dimensions more than 2.0mm above the bed, 0.24mm is added to all Z-dimensions (after any shrinkage allowance). These defaults can be changed if necessary to suit your printer -- set both to zero to cancel any effect.

Remember to allow for this when taking measurements to calculate the Z-shrinkage allowance.

This will be in the next program update. It is about time I released it, but there are still a lot of loose ends to be fixed first.

cheers,

Martin.
 
_______________
message ref: 5351
A couple of evenings ago, I was just about to stick my resin printed timber bases down onto a prepared cork bed test plank, when fortunately I made a last minute check with a piece of rail 240mm long (to represent 60 foot length) and spotted that being a numpty I had made a miscalculation :(
One of my joint sections (5+4=9 sleepers) and 2 of my plain sections (7 sleepers) = 23 sleepers!
Why did I think that 2 * 7 = 16? brain fade perhaps?
I have now increased the plain section up to 8 sleepers, so now have the correct number of sleepers per 60' segment.
4 + 8 + 8 + 5 = 25 sleepers per 60' segment.
Having printed the replacement plain sectiions I have now stuck the requisite bases down onto the cork trackbed, and then plugged in a pair of 60' rail lengths with the resin printed S1 plug chairs.
 
_______________
message ref: 5355

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
.
n.b. A couple of times now, switching on the Neptune 2S with the USB cable connected to my computer, has caused the computer to do an instant Windows restart. Which is extremely annoying if you are in the middle of doing something important. :(

Presumably caused by a startup blip on the 5-volt line in the USB cable.

It is safer to switch the Neptune on first, and then plug in the USB cable. Or use the SD card instead.

Something to bear in mind.

cheers,

Martin.
 
_______________
message ref: 5357

Martin Wynne

Admin
Thread starter
Location
West of the Severn UK
Info
.
Enjoy using Templot?
Thanks.

Please do not send requests for help direct to me via email.

Post your questions on the forum where everyone can see them and add
helpful replies.
_______________
message ref: 5358
Top