From out of nowhere, the usually dependable K40 C02 laser started having a hissy fit. An unnerving sound of grinding and gnashing made the machine sound like it’d been possessed by a poltergeist. When your laser cutter sounds like it’s from a horror movie, something ain’t right.
As it turned out, the noise was symptomatic of an underlying issue that impaired the machine’s movement along the X-axis, resulting in failed cut after failed cut.
While I had my suspicions about the cause of such an alarming racket, what follows was the process used to diagnose and resolve the problem.
Hopefully the pain-in-the-arse experience I went through to eventually fix this misbehaving K40 will help some other poor sod who might run into a similar issue.
Hypothesis 1 – Faulty Stepper Motor
The first possibility that came to mind was a potentially faulty stepper motor on the X-axis. The logic being: gantry is trying to move, motor not moving gantry; therefore: faulty stepper motor.
It’s not unreasonable to assume one of the cheap-as-chips stepper motors that actuate the cutting head may have failed. Both the X-axis and Y-axis drive units are Nema 17 stepper motors, like those used on 3D printers. Being no-name Chinese electronics, their reliability is questionable to say the least.
The focus was specifically on the X-axis stepper as this is the motor that drives the gantry which was experiencing limitations of movement.
The X-axis has a dual shaft motor, simply meaning there is a common output shaft that protrudes from two sides of the motors body. On one output shaft there is a toothed pulley, moving the right-hand side of the gantry via a drive belt. The opposing output shaft is connected to a long transfer rod, and subsequently to the left-hand side belt of the gantry.
There was the possibility that the gantry may have been binding at some stage, or possibly a piece of debris had lodged in a bearing path, resulting in the motor drawing excessive current and burning out.
To test if the motor was dead, it needed to be removed from its mounting on the internal frame within the enclosure. While it may sound simple enough to remove one tiny motor, it was not as straightforward as it initially seemed. OR maybe I’ve been spoiled by 3D printers with their easy to access and service stepper motors.
The simple task of removing the motor was made difficult solely because there’s not enough space to get even a small and stumpy screwdriver into the claustrophobic enclosure. The only way to undo the mounting screws that hold the Nema in place is to remove the entire frame from the enclosure.
Four nuts and bolts hold the internal frame in place within the sheet metal enclosure. Once these are removed, the anodised brown aluminium frame can be carefully lifted out. It’s a tight squeeze and can be tricky to lift it out. This is something that shouldn’t be rushed as the ribbon cable for the Y-axis motor could be damaged if the extraction is hurried.
It’s only after the frame is out that the stepper motor can finally be accessed and worked on. Once the screws are removed and the (simply terrible) shaft coupling also undone, the motor comes out.
An additional bit of advice is to loosen off the drive belt from the toothed pulley to get more leverage for the motor. There are adjustable tensioning screws at the back of the frame, one on either side for each belt. Backing these off to release tension makes taking the motor out that little bit easier.
With the motor out, it could be tested with no load. This was done by controlling the motor using software, in this case – Lightburn. A G0/G1 command can be used on the console or alternatively the move control buttons. The software just needs to tell the machine to move along the X-axis. I set my movements for 100mm at 50mm/s and clicked the X-axis directional arrows, then observed the motor’s output shaft.
As the motor was mechanically disconnected, the output shaft turns as if it’s trying to move the gantry. It’s important to test movement in both directions to ensure correct functionality, a strategically placed piece of electrical tape can help determine which direction the shaft is turning.
From this basic test, it was proven the motor had free and uninterrupted movement in both directions. Meaning the motor was not the issue…
Hypothesis 2 – Failed Bearings
A busted bearing was the next rational assumption. It made sense after all, the gantry was visually and audibly struggling against something. Almost as if it were trying to overcome the friction of metal on metal.
It may have been possible that one of the nylon bearings on the right-hand side of Y-axis gantry upright was not rolling as it was seized, gunked up, flat-spotted or any other number of possibilities.
Like with the stepper motor, getting access to inspect these little plastic wheels is tough. They can be inspected in situ albeit with an obstructed view, hence being able to properly diagnose any problem with the bearings is next to impossible with the frame still mounted in the enclosure.
Luckily, as the frame had already been removed for inspection of the stepper motor, access to the bearings was straight forward. However, if I hadn’t already inspected the stepper motor, removing the frame would have been inevitable.
As with the stepper motor, the bearing/s weren’t the issue. While cheap and underwhelming, the little white wheels turned freely with no resistance limiting their freedom of movement.
Hypothesis 3 – Wonky Frame
By this stage, things were getting desperate.
The two most likely fault causes had been investigated and debunked. The X-axis stepper motor was chugging away nicely, and the bearings were doing their best Limp Bizkit impression and just kept rollin’.
Clutching at straws, another possible cause was the frame itself. If this was not perfectly square it could cause the gantry to bind and prevent the gantry from moving, as was being witnessed.
If you’ve read my Ender 3 Pro instruction manual, you’ll know that it’s important to use a builder’s square (or similar measuring tool) to ensure the frame is square to avoid wonky prints. However, as the K40 comes as a fully assembled piece of equipment I had never checked whether the frame was in fact true.
Sure enough, the frame was indeed out of square from the factory. Not by a large margin, only about 2mm along the longer edges, but enough that maybe, just maybe, it would cause the gantry to bind up while rolling along its path.
Squaring everything back up wasn’t too tricky. At each corner there’s a set of three hex head bolts and by loosening these fixings, manipulating the corners a bit, and then tightening it all back together in the correct order, the ‘squareness’ of the frame can be improved.
While still not perfectly true and square, there was at least a 50% improvement using this method of adjustment. To achieve even better results shims and spacers would need to be used. I did not have these items on hand at the time, but will certainly be installing some in the not too distant future.
But even after improving the squareness of the frame, the issue still remained.
What eventually did resolve the problem was luckily a simple and easy solution.
After going through the wringer of checking this and that and after pulling the machine half to pieces, a solution was finally discovered!
While I was on the right track, the cause was somewhat related to my second hypothesis, that being a faulty or failed bearing. As it turned out, the bearing on the left-hand side of the gantry was dry as a bone.
This linear style bearing, with a piece of 12mm rod and 2 cylindrical bearings over the top, had simply run out of lubrication. Each time it tried to move back and forth it was catching on the steel, the motor couldn’t overcome this amount of friction.
On a whim, I gave the rod a squirt of spray lubricant. And wouldn’t you know it, that was all it was. Finally, fixed!
A Lesson Learnt
While taking apart the K40 was a worthwhile learning experience in understanding it’s assembly and design, the most valuable lesson was that sometimes it’s the simplest solution that may in fact be the answer.
As you read earlier, my original hypotheses involved removing significant amounts of the laser’s componentry. In chasing the cause of the problem the machine was dismantled to within an inch of its life, yet the eventual fix took all of 5 seconds to apply and yielded instantaneous results.
While hindsight is 20/20, a good place to start when dealing with any technical issue is to look at the simplest possible solutions. Had I done so, it would have saved much effort and frustration.
At least this K40 is once again cutting away nicely.
Be Sure To Check Out My Other K40 CO2 laser articles:
If you have enjoyed reading this article please share with your friends and followers on social media! while you’re there don’t forget to follow RB42, on Instagram, Facebook and Pinterest for updates and new releases.