Month: February 2024

3D Printer Fowl Up

barkmannLeave a comment

After years of reliable service, our Tevo Tarantula Pro (TTP) finally bit the dust. Well, better to say I pushed it over the edge. After the disappointment of the formed concrete roadbed, we switched gears and starting printing casting molds for roadbed “bricks”, and they were printing like clockwork… Until the nozzle clogged. That’s not the end of the world, we’ve had clogs before, but this time it was the death knell for the hot end.

Without going into a lot of detail, it boils down to the leads to the heating element frayed over time, to the point where they became intermittent. This caused the temperature of the hot end to randomly fluctuate drop below the melting point of the plastic filament. Try as I may to replace the clogged nozzle, fiddling with it for more than a week, all I end up doing is snapping off the new replacement nozzle in the hot end.

I manage to extract it, but in my attempt to replace it, all I manage to do is break off the heat break in the hot end. The ensuing attempt to extract the threaded part of it still stuck in the hot end reveals the threads have been damaged from one or the other breaking off in there. Not even a thread chaser will repair the damage I’ve done. I managed to find a replacement hot end kit on Amazon and ordered it.

Totally Irreparable Hot End

Long story short, it appears it was my attempt to install it, leaving the connector inline and not just hard wiring it into the harness is the issue now. It allows way too much electronic noise in the temperature sensing feedback circuitry for stability, causing thermal runaway shutdown. At least that’s my best guess as to what’s going on.

If you have no idea what that means, don’t worry, only tech savvy folks familiar with 3D printing and the associated electronics will. The short answer is the replacement isn’t working the way the control electronics expects it to, meaning the 3D printer is still broken.

A New Hope

Totally frustrated by the inability to 3D print anything, and for weeks now, it’s time to buy a new printer. At the very least look at what’s out there now compared to when I bought the TTP. Not sure if fortune was smiling down, but an ad for a Sunlu S9+ popped up, a Creality CCR10 knockoff. I decided to take a look. I compared it with other printers and could not find anything close to the price. It even includes a filament dryer!

I pulled the trigger and bought it! For little more than I originally paid for the TTP printer I get more print volume. Much more. Where the old TTP was barely able to print 8½”x8½”x9″, the new S9+ can print 12″x12″x15″! It doesn’t sound like much, until you can’t print your design as a single part. The design has to account for that, “breaking” it into pieces, to be assembled together later. How do those pieces get assembled? Snap together? Glue together? Doesn’t matter when you can print a single monolithic part!

But I’m getting ahead of myself. Had to wait a few days for it to be shipped, but when it arrived I had to open the box at the very least. Before I can unbox it I have to collect up all the pieces of the old printer and stash them out of the way until I can get back to fiddling with it. Now there’s space for the new one and boy is it well packed! Man is this thing is sweet! It’s practically ready to print right out of the box!

Not Quite Ready To Print

Some Assembly Required

As you can see from the picture, the printer isn’t quite “turn-key”, but all that remains is to attach the Z axis portion (right) to the base (left). The base portion contains the controller, power supply, heated print bed, and touch panel display. The Z axis portion contains the dual stepper drive motors, the X axis “bridge” and stepper drive motor, and the print head with hot end and cooling fans. To connect the two, only three screws per side are required to attach them. Once those zip ties are cut that is.

Z Axis Attached

The Z axis stands nearly as tall as the wooden closet rod that holds my spools of filament, but thankfully just fits beneath it. Once I rotate it so that the sag over time is now pointing upward that is. All that remains is attaching the support rods on either side to keep the Z axis rigid and aligned. A few more screws and everything’s ready to go!

Well, after I figure out how to tell Octoprint how to talk to the new printer and configure the Cura slicer for it everything’s ready to go. Turns out since the S9+ is a CCR10 knockoff, I can just tell everything that’s what it is and it just works! Like I said, this thing is sweet!

Ready To Print!

Finding Out The Hard Way

If I give this deal any dings, it going to be the filament dryer. It’s one of their first generation models and has a “pigtail” harness connection into the side of the printer base. I should have saved myself the extra money and bought one of their next generation dryers like I already have. The paired version must be commanded on via the printer front panel, which sits in the back right corner facing to the right, and it’s difficult to get to.

The shelf is just wide enough for the printer to sit sideways, and must be placed like that because there’s not enough space with the front panel facing out for the bed to travel back and forth without hitting the wall and window frame. I knew this when I bought it. The only thing I didn’t know was the filament sensor and extruder must be fed from the side that’s facing out, not from behind.

If you look closely you’ll see a set of miniature roller bearings, one above and one below the opening for the filament, meant to guide the filament up or down depending on the height of the X axis bridge. It doesn’t do much to guide the filament fed from the side. Coupled with my decision to try out their miniature sample spool of filament, I soon discovered how light and unstable the filament dryer is, ready to throw it across the room after it falling and having to catch it for the umpteenth time! Good thing it’s tethered to the printer!

Printing Roadbed “Brick” Mold Parts

And We’re Printing

All that aside, we’re printing again! That’s all that matters now. We were on our way to casting roadbed “bricks” until the old TTP died. We’re back in business and printing them again. You can see from the picture there are three parts to the brick mold. There is a single “U” shaped channel in the middle that serves as a “well” for the track to sit in, and two sides that are mirror image ballast profiles that “cradle” the track.

These parts need fastened together into a single unit. We have an assembly jig designed and ready to print. In fact, it was the very next print job when the old TTP failed. It was the first print job once the new S9+ was ready to go. The jig is also three parts, identical, and ½” wide with the proper shape to hold the brick mold parts together for gluing on both ends and in the middle.

These brick molds are 8″ in length and roughly 5½” wide, and there are a number of reasons for these dimensions. The 5½” is easy, it’s the width of a 2×6. The 8″ dimension was mainly the limitation of the print volume of the old TTP, but that’s not the only reason. For 1:24 scale, 8″ is a scale 16′, the distance between the trestle bents for our trestles. It’s also a convenient length for making curved sections.

For example, a 10′ diameter curved section of track is approximately 32″ long, or four 8″ segments. Same for a 14′ diameter curved section of track. The only difference is the 10′ diameter track needs only 12 pieces (30° each) to complete a circle where the 14′ diameter track requires 16 (22.5° each). The 20′ diameter curved sections are just shy of 48″ long or six 8″ segments. Overall a very convenient arrangement.

Roadbed “Brick” Casting Molds

Fine Tuning

Believe it or not, this thing is already tuned as close as anything off the shelf could be. Everything is pretty much spot on! About the only thing I need to adjust is the extrusion multiplier. The default of 100.00 needed increased slightly to 102.30. Also needed to fine tune the X and Y home offsets slightly, only off by ¼” at most in the X and a miniscule amount for Y. Other than that, Z offset is spot on. PID temperature control for both the bed and hot end are spot on.

I did end up ordering a couple different flexible PEI print sheets. The “factory default” sheet is about the stickiest I’ve ever encountered. They even provide a metal putty knife to pry the prints off of it! This thing is wicked! If I ever run into a print that absolutely refuses to stick, I’ll use this thing! Otherwise, the textured PEI sheet does nicely. I have yet to print anything that won’t stick to it. And I have yet to have any print fail to release after cooling off. In fact, if the print doesn’t just slide off, I just need to flex the sheet a bit and it pops right off.

As I said before, this thing is SWEET! I couldn’t ask for better. And the price was right!

Lifting The Deck

barkmannLeave a comment

Let’s start with the obvious question, “Why lift the deck?” Good question. Glad you asked.

The main reason has nothing to do with the deck and everything to do with the dogs and the bridges… Well, what used to be the bridges.

Let’s rewind…

When we originally built the approach trestles for the Howe truss bridge, they were built 20″ tall, roughly forty scale feet tall. After we refit the upper loop into the “triple decker” arrangement, we needed a second bridge. We didn’t have a second bridge.

The Howe truss bridge was built over the course of months, with strong attention to detail, which explains why it took months. We don’t have months to build another, let alone time to “refresh” the old one. Not if we want to run trains anyway.

We’d been wanting to try using some metal channels as a stand in for deck plate girder bridge spans. We saved the metal channels from some discarded Levolor® blinds, the tops with the mechanisms removed, conveniently powder coated brown.

Long story short, we quickly cobbled together a couple of crude wooden frames to slide inside the channels and provide a means of securing the two new bridge spans to the set of bridge trestle approaches. Those crude wooden frames were attached to the trestles with screws to keep them in place.

Makeshift Deck Plate Girder Spans

The Problem

The trestles themselves weren’t fastened to the ground in any shape or form, so when the dogs plowed into the bridges, they just repositioned the whole kit and kaboodle! Eventually they managed to rip one of the bridge spans loose, exposing the sharp end of the screw that was torn away from the trestle.

Time to remove them altogether, along with the 10′ span of track that we’ve grown weary of placing back on top of the bridges. We kept telling ourselves if only the bridges were higher, with enough clearance for the dogs to run under them without hitting them, then they may stand a chance of staying where we put them.

Great idea. Only problem is the span that comes off the deck leading up to the bridges is already a 2% grade or better. The bridges would have to be at least 4″ taller, 24″ vs. 20″, but 26″ would be better. If we raised the deck by 4″ as well, it would solve the problem.

The Solution

We’ll stick with “lifting” the deck, since “raising” the deck sounds too much like we’re “razing” the deck. And we’ll have to compromise and go with lifting it 3½”, the thickness of a 4×4, rather than 4″. Regardless of how much we’re lifting it, it must remain level.

As a test, I thought I could lift the deck at one support point at a time while Ann placed a thick shim of wood between it and the deck. I was fooling myself about how heavy the deck is. Even using a chunk of 2×4 as a lever, the most we could place was a chunk of 1×6.

First Attempt To Lift By Hand

We did manage to get one entire side lifted, but it was all we could do to get those chunks of 1×6 in place, and now the deck is tiling a bit off level. That will have to be good enough for now. It’s going to take a hydraulic jack to lift the deck enough to place those 4x4s. And that will have to wait until the weekend.

Before the jack can be used, we’ll need some means to accommodate its placement. It sure would be nice if it fit beneath the lifting points, but it doesn’t. It’s close, but no cigar. A chunk of 2×2 that needs removed anyway should fit the bill. In most cases just attaching the 2×2 to provide enough of a “ledge” for the tip of the jack to gain purchase is all we need. So one by one each support location is raised and the chunk of 1×6 is replaced by a chunk of 4×4.

Deck Lifted 3½”

The Lift

It’s safe to say that where the deck rests directly atop the concrete post base a chunk of 4×4 is called for and where the deck joist rests in the slot in the base a chunk of 2×6 is required. On the side of the deck closer to the fence the ground is higher than the opposite side which is why part of the deck relies on the slots in the base. Thankfully the depth of the slot in the base is roughly 2″ deep.

There’s a hitch when it comes to lifting the side closest to the fence. The part of the deck that’s supposed to emulate an excavated rock face with a tunnel beneath for the lower loop is nowhere near complete. We’re still working on “what-if” scenarios, trying to figure out what works. The idea is to have a gristmill with a waterwheel fed by a waterfall, eventually, but we’ll save that for another post.

The other “feature” of that part of the deck is that it’s slotted to provide a path for the track to loop back under itself, creating the “triple decker”. That slot is framed on either side by a  separate 2×6 joist, each with its own post base. We purposefully made that section of the deck only 6′ wide and angled away from the line of the fence to provide a wedge shape that grows to nearly 2′ wide at the exit of the track slot.

The final shape may not be a wedge exactly since it also needs to provide for a short tunnel for the lower loop to pass through. It’s difficult to find a design that accommodates all the requirements we’re giving it. The point is there are two jacking points, not just the one. That’s where that 2×2 had to be removed since it would be in the way of jacking those points.

Two Support Points Surround The Slot

The Future

That’s about all we can do for now. The next steps will be designing the trestle approach to the bridges and a means of solidly fastening it to the ground. Certainly hope that 24″ tall is enough to keep those bridges taller than the dogs.