In the previous article, I explained why I decided to move away from Marlin and switch to Klipper, and how I started using an old Linux-powered TV box instead of a Raspberry Pi.
Switching to a different control system made configuring the printer noticeably easier. I could now change many parameters directly in the configuration files, and I had more convenient calibration tools available.
But the main problem that had led me to experiment with the settings in the first place was still there.
I still couldn't achieve a consistent first layer across the entire print bed.
When Manual Adjustment Isn't Enough
At first glance, leveling a 3D printer's bed seems like a fairly simple procedure.
You need to set the correct distance between the nozzle and the surface, check it at several points, and adjust the bed position if necessary.
In practice, however, it turned out to be more complicated.
I repeatedly tried to adjust the bed manually. I checked the distances, made small adjustments, and started another print.
Sometimes the result looked quite acceptable, but as soon as I tried printing a model that covered a larger area, the problems became more noticeable.
In one part of the bed, the first layer might look fine, while in another, the distance between the nozzle and the surface was already different.
I had to go back and adjust everything again.
Switching to Klipper made the configuration process more convenient, but it didn't eliminate the need to evaluate those distances manually.
That's when I started thinking about installing a bed probe.
Why Does a Printer Need a Bed Probe?
The idea is fairly simple.
Instead of manually estimating the distance between the nozzle and the bed every time, you can let a sensor perform the measurements.
It can determine the position of the surface at selected points and use those values when configuring the printer.
Of course, a probe doesn't fix poor mechanics or make the bed perfectly flat.
But this approach has one important advantage: repeatability.
With manual adjustment, the result depends heavily on how accurately you can determine the gap at each point.
Even small differences can affect the first layer.
A probe allows measurements to be taken more consistently and compared with one another.
That was exactly what I had been missing.
Installing the Probe
I decided to install a probe only after I had been using Klipper for some time.
In other words, it wasn't part of my initial transition from Marlin.
At first, I expected Klipper's more convenient manual adjustment tools to be enough.
But gradually, it became clear that improving the results further would require a different way to measure the bed's position.
So installing a probe became the next stage of the upgrade.
Of course, simply adding the probe didn't mean all the problems were solved.
I needed to position it correctly relative to the nozzle, configure its operation, and check how consistently it performed measurements.
Only then could I use the collected data to adjust the bed.
Four-Point Bed Leveling
One of the most useful results of installing the probe was the ability to check the bed's position at four points.
Previously, I had tried to achieve the desired result manually, judging the distance between the nozzle and the surface.
Now I had measurements that allowed me to evaluate the differences between different areas of the bed.
This noticeably changed my approach to adjustment.
Instead of repeatedly guessing which area needed to be raised or lowered, I could rely on actual measurements.
What mattered most was that those measurements were highly repeatable.
When I ran the same check several times, I could expect comparable results.
That made it much easier to determine whether an adjustment had actually improved the bed's position.
For an old printer that I was gradually bringing back into working condition, this was a significant improvement.
Repeatability Matters More Than You Might Think
Before installing the probe, I mainly evaluated the results by looking at how the first layer was laid down.
Of course, this approach is still useful. Ultimately, print quality is what shows how well a printer has been configured.
But there's a problem.
If the first layer improves or gets worse after an adjustment, it isn't always clear what exactly has changed.
Perhaps it's the bed position. Perhaps it's the nozzle height or some other parameter.
Once I had repeatable measurements, it became easier to separate bed adjustment from other factors.
I could check the surface, make a change, and measure it again.
This made calibration a more controlled process.
And although the probe didn't eliminate every design flaw, it significantly reduced the amount of trial-and-error adjustment.
A Probe Isn't a Universal Solution
There's one important point worth mentioning here.
Installing a probe doesn't mean you can forget about the printer's mechanics.
If the frame isn't rigid enough, the linear guides have play, or individual components aren't working correctly, measuring the bed surface won't fix those problems.
The probe simply helps provide information about the bed's position.
How accurately the printer can use that information depends on its mechanical condition and configuration.
That's why I didn't see installing a probe as the final solution to all the Graber i3's problems.
It was more of another tool that allowed me to keep making progress.
And, as it turned out, a very useful one.
What Changed After Installation?
For me, the main benefit wasn't any particular Klipper feature, but the ability to obtain consistent measurements.
I could now check the bed's position at four points and evaluate the results of adjustments with much more confidence than before.
The calibration process became clearer and more predictable.
That didn't mean the printer had suddenly turned into a modern, high-precision machine.
It was still an old Graber i3 with its own design characteristics and limitations.
But now I had a tool that I had been missing for a long time.
Small changes like this gradually made the restoration process more systematic.
Instead of endlessly trying to get things right by guesswork, I started getting data I could actually work with.
What's Next?
Installing the probe was an important stage in upgrading the Graber i3, but far from the last.
As I continued working with the printer, I kept finding issues that I wanted to address.
One of the next areas was the Z axis.
The printer's design used two motors, and I decided to change how they were controlled to gain more configuration options.
This required changes to the electronics and Klipper configuration.
But that's something I'll cover in the next article.