As I continued upgrading my Graber i3, I gradually turned my attention to the printer's mechanics.
Initially, I wasn't planning any major changes. All I needed was to design a proper mount for the inductive bed probe I'd installed during earlier upgrades.
However, as often happens with DIY projects, solving one problem gradually led to discovering others.
In the end, I had to make some changes to the X carriage, redesign the hotend mount, and reassemble parts of the Z-axis mechanism.
The most interesting part was discovering mistakes I'd made during the printer's original assembly about five years earlier.
How It All Started
I carried out the first tests of the inductive probe without a permanent mount. I already had the sensor, but there was no suitable bracket for it.
So the next step was to design a mount that would hold the probe securely in position relative to the hotend.
At first glance, it seemed like a fairly simple task.
But my Graber i3 has some unusual design features. Many of its mechanical components are made of plywood, including parts of the X carriage.
That's where I discovered the first problem.
Because of the way the plywood parts were joined, I couldn't assemble the carriage rigidly enough. There was some play in the structure, which affected the stability of the hotend's position.
Since I needed to design a new part anyway, I decided to address that issue at the same time.
The First Probe Mount Was Inconvenient
Before moving on to a new design, I tried a simpler approach.
Following the example of many DIY 3D printers, I attached the inductive probe directly to the hotend cooling assembly.
This allowed me to position the probe near the nozzle without making major changes to the carriage.
But the drawback quickly became apparent.
Whenever I needed to remove the hotend for cleaning or a nozzle replacement, the probe mount got in the way.
Essentially, adding one component had made another component harder to service.
I wasn't happy with that arrangement.
I wanted a design where the probe had its own mount and wouldn't interfere with hotend maintenance.
The Original Hotend Mount Had a Drawback Too
It's worth explaining how the hotend was originally mounted.
I used an aluminum bracket designed for mounting 16 mm linear shafts.
Its dimensions and construction made it a very good fit for the job.
It held the hotend securely enough, and for a long time I was perfectly satisfied with the arrangement.
But there was one drawback.
If I had to remove the hotend, for example for maintenance or a nozzle replacement, it was practically impossible to reinstall it at exactly the same height.
I had to check the nozzle's position relative to the bed again and adjust the settings if necessary.
On a printer that I regularly modify and work on, that was rather inconvenient.
So when designing the new carriage, I decided to change the hotend mounting method as well.
A New X Carriage — Plywood and 3D Printing
I wasn't planning to abandon the original design entirely.
Instead, I decided to make a hybrid carriage.
One part remained plywood, while I produced the other using 3D printing on the same Graber i3.
I used the geometry of the existing plywood part as the basis for the printed component and added the necessary mounting features.
This allowed me to incorporate two features into a single part:
- a bracket for mounting the hotend;
- a mount for the X-axis endstop switch.
The endstop had previously been mounted on the Z carriage, but after redesigning the assembly, I moved it to the X carriage.
The new part therefore combined several functions.
At the same time, I managed to preserve much of the printer's original mechanical design.
A More Rigid Design and Easier Maintenance
The biggest advantage of the new carriage wasn't even the provision for mounting the inductive probe.
After redesigning the connections between the parts, I was able to assemble the entire unit much more rigidly.
The play that had existed in the original plywood construction was eliminated.
The hotend's position relative to the carriage is now considerably more stable.
The new hotend bracket is also an integral part of the printed component and stays on the carriage when the hotend is removed.
As a result, I can remove the hotend for maintenance and reinstall it at almost exactly the same height.
Of course, it's still a good idea to check the nozzle position after working on the hotend, especially if the nozzle itself has been replaced.
But now, routine removal and reinstallation no longer require me to find the correct position of the entire assembly all over again.
The inductive probe has its own separate mount, screwed to the back of the plywood section of the carriage. It no longer interferes with removing or servicing the hotend.
That's exactly the result I wanted.
I Decided to Reassemble the Z Axis Too
While the printer was partially disassembled, I decided to pay some attention to the Z-axis mechanics as well.
There was also some play here that I wanted to eliminate.
However, I wasn't planning any major structural changes.
I limited myself to minor modifications of the existing parts, particularly tightening up the fits for the linear bearings and guide rods.
This helped reduce unwanted movement between components.
But during reassembly, I discovered something else.
Two T-Nuts Aren't Always Better Than One
When I originally assembled the Graber i3, I installed two T-nuts on each side of the Z axis.
At the time, I thought this would provide a more rigid and stable mechanism.
And for about five years, I never even questioned that decision.
However, during reassembly, I discovered that the additional nuts were creating unnecessary resistance to the rotation of the lead screws.
Instead of the expected improvement, I'd added extra load to the movement mechanism.
It's quite possible that slight misalignment between the nuts was forcing the lead screws to overcome additional friction.
So I removed the extra T-nuts, leaving one on each side.
After that, the movement became smoother and less restricted.
It's important to point out that using two nuts isn't necessarily a mistake in itself.
There are designs where additional nuts are used successfully, for example to reduce axial backlash.
But in my case, the arrangement turned out to be a poor choice.
Sometimes trying to make a mechanism more rigid produces exactly the opposite of the result you expect.
Removing the Play Revealed Another Problem
The most interesting part came after reassembly.
Once I'd tightened up the bearing and shaft fits, eliminated the play, and started readjusting the mechanics, I noticed that one of the Z carriages was binding.
And I hadn't noticed this behavior before.
At first, it seemed rather strange.
The printer had worked before disassembly, but after removing the play, one of the carriages began to stick during movement.
I had to inspect the entire assembly carefully.
And the cause turned out to be surprisingly simple.
A Bolt Mistake from Five Years Ago
The Z carriage carries a pulley for the X-axis belt.
The pulley rotates around a fastener that passes through the corresponding part of the structure.
During the original assembly, I should have used a screw with a flat head.
Instead, I used a regular bolt.
Its protruding section was rubbing against the carriage body.
The surprising thing was that this had barely caused any noticeable problems before.
While the structure still had some play, the parts could shift slightly relative to one another, so the contact didn't cause obvious binding.
But after reassembly, everything changed.
Once the fits became tighter and the carriage more rigid, that unwanted freedom of movement was gone.
As a result, the protruding bolt began interfering with the mechanism's normal movement.
In other words, fixing one problem revealed another that had been there since the printer was first assembled.
The Fix Was Surprisingly Simple
I didn't have a suitable flat-head screw available.
So instead of looking for a replacement fastener, I simply reversed the existing bolt.
Now the protruding section no longer rubbed against the carriage body.
At the same time, the pulley mount remained sufficiently rigid.
After this small change, the binding disappeared.
In the end, I achieved exactly what I'd been aiming for from the beginning: a rigid structure without noticeable play that could still move freely along its guides.
Rigidity Shouldn't Restrict Movement
This whole experience reminded me of something important.
When assembling a 3D printer's mechanics, simply fastening everything as rigidly as possible isn't enough.
You need to find the right balance between rigidity and freedom of movement.
A carriage shouldn't wobble on its guides, but it should still move smoothly, without binding or excessive resistance.
In my case, the original play had actually been hiding an assembly mistake.
As long as the parts could shift slightly, the protruding bolt barely affected operation.
But once the structure became more rigid, the problem immediately became apparent.
Only after eliminating the source of friction did the mechanism begin working properly.
Upgrade Results
Initially, all I intended to do was make a mount for the inductive probe.
But in the end, I managed to solve several problems at once.
The new hybrid X carriage allowed me to make the assembly more rigid and eliminate the previous play.
The separate inductive probe mount no longer interferes with hotend maintenance.
The new bracket allows me to remove and reinstall the hotend with almost no change in its height.
The X-axis endstop switch is now mounted directly on the new printed section of the carriage.
In addition, while reassembling the Z axis, I improved the fits for the linear bearings and guide rods, removed the extra T-nuts, and eliminated the binding caused by the incorrectly installed bolt.
As a result, the mechanics are now both more rigid and freer to move.
What I particularly like is that none of this required rebuilding the printer from scratch.
I was able to preserve most of the original construction, making changes only to individual parts and correcting old assembly mistakes.
For me, that's exactly what gradually upgrading the Graber i3 is about: not building a new printer every time, but steadily improving the existing one and fixing the shortcomings that become apparent along the way.