Y-axis 0 endstop question

Since the e3D nozzle upgrade I’ve noticed that the probe position for the first (front) washer is nearly besides the washer. It turns out that the bed physically can’t move any further back because it hits the end-stop. My bearing holders are not original because I’m using LM8UU linear ball bearings, but I’ve checked Biolumo’s version of the double bearing holders and the hot-inserts position exactly match the original Lulzbot’s position. I also checked he x-carriage and there doesn’t seem to be any offset from the bearingholders towards Y-axis as well.

I’ve taken a picture of the nozzle/washer situation:

And here’s a close-up of both the original (top) and Biolumo’s double bearings holder, showing the hot-inserts for the end-stop microswitch:

I can’t for the life of me discover any difference that would account for the nozzle displacement? I’ve checked the Y-axis stepper and there is nothing in the way for the endstop to reach the rod mount.

This is the output of a G28, G1 Y0 sequence

Send: G28
Recv: X:0.00 Y:191.00 Z:158.00 E:0.00 Count X: 0 Y:19196 Z:252800

Send: G1 Y0
Recv: ok
Recv: echo:endstops hit:  Y:2.44

Is the endstops hit: Y:2.44 normal?

I’m wondering if Y=0 (= endstop) is so close to the washer rim for others?

The Gladiola/1.04 revision made the rear Y-rod block (the one over the Y motor) about 6mm shorter, right where the limit switch actuates. The firmware released concurrent with the 1.04 builds changed homing from that shortened rear block (Y-Min) to the unchanged front idler (Y-Max), so that both pre-1.04 and 1.04+ builds would have a consistent “home” point. However… You are correct that a pre-1.04 running that firmware will hit the rear endstop before getting to Y=0, and it is just barely within the washer perimeter when probing.

The fix… At least what I did… Print the 1.04-version Y-rod block and replace your pre-1.04 rod block. You’ll need to install heat inserts, but I think you know how to do that as it sounds like you made your own bearing holders. Here is the link to the model:

That new version block is about 6mm shorter, so the bed will not hit the endstop until about Y= -4.

With a little care, you don’t have to remove belts or anything to replace that block. Just remove the screws, carefully support the bed and hold the rods in place while pulling off the old block. Put the new block in place, and replace the screws.

Here is a comparison of the old/new designs for that block. As you can see, the left-hand side is about 6mm shorter on the new design.

Thanks, Scott! Will print and replace the new rod mount asap.

What I don’t get is, that even if I would run pre-1.04 firmware the physical limitation of the Y=0 endstop would have prevented normal washer reach anyway. I wonder if it has always been this bad but just reared it’s ugly head because of the new e3D nozzle shape?

I run E3D nozzles on my Mini. It is an original, first run machine. I have no problems with the tap location on the washers. Changing to E3D nozzles shouldn’t change the tap location since the center bore of the nozzle is in the same location.

Changing the nozzle would not change the physical probe location. The front location has always been defined by the Y-limit actuation as the bed moved fully rearward. I can only surmise that focusing on the new nozzle and the probe sequence perhaps made you “more aware” of relative difference between front/rear probe positions.

The newer firmware (with new origin, bed homed forward at Ymax) changed the physical probe locations for the rear washers (i.e., bed forward) slightly, but did not change the probing points of the front washers because they were always determined by the y-end-stop actuation when the bed goes fully rearward.

Changing the Y-rod support to the newer, shorter version allows the bed to move further rearward. That physically allows getting further inboard on the front washers, and the new firmware has front/rear positions defined to be equidistant from the washer edge with that new block.

Not only that, the pointy shape of the E3D nozzle accentuates the (mis)alignment. But yes, it must be an observational issue.
It’s much better with the new rod mount now!

YES. I believe this is your issue. The new fimware need the slightly modified y-axis hadware to match. Makes a difference when bed probing.

If you print the new Gladiola revised y-axis hardware and install the heat inserts with a soldering iron you should be good to go like i did.

I, too, agree with ScottW’s suggestion: my Lulzbot Mini 1.0.3 had the same problem (wiping too far back), and replacing the 1.0.3 Y motor mount with the 1.0.4 part fixed the problem!

Interestingly, simply replacing the part didn’t, at first, solve the problem. The 1.0.4 Y motor mount has two pair of Nyloc nuts and screws, that screw against the ends of the Y axis rods. When I first installed the 1.0.4 part, I noticed the belt was suddenly loose, so I (against the assembly instructions) tightened those two rod screws until the belt was tight again. Then I printed a test print, and was terribly disappointed to find the nozzle still wiped too far back on the wiper area. So just replacing the part didn’t solve the problem.

Then I re-read the 1.0.4 assembly instructions, and noticed that it very clearly said to tighten those two Y rod screws only enough so that the tops of their heads were flush with the face of the Y motor mount. So I unscrewed the screws to that point, which made the belt very loose, then I tightened the belt, and Voila! The printer now wipes right in the middle of the wiping area!

Here is what I think is going on: the nozzle wiping is now effectively a dimensional error test - which is actually useful. If your Y axis is too long or too short, the nozzle will wipe too far behind or ahead of the wiping area. I imagine this is because, as ScottW pointed out, the 1.0.4 firmware changed the Y axis reference from the back endstop (whose position changed with the 1.0.4 parts) to the Y axis front endstop (whose position was common across all Mini models). If your 1.0.3 Mini had a too-long Y axis, the 1.0.3 would have no problem hitting the wiper area because it’s so close to the back endstop. The same too-long axis on a mini running 1.0.4+ firmware will miss the wiper area, because it’s so far from the front endstop that a Y scale error across that distance accumulates a few millimeters of error.

When I inappropriately tightened the new 1.0.4 Y motor mount rod screws until the belt was as tight as it had been, I was adding a few millimeters to the effective length of the Y axis, which caused Y dimensional errors, which in turn caused the nozzle to miss the wiping area.

When I then loosened the 1.0.4 Y motor mount rod screws to the distance described in the assembly instructions, I was reducing the Y axis length, which required me to tighten the Y belt. …and now my printer works great, hitting the wiping area every time.

So, whatever the reason, replacing the 1.0.3 Y motor mount part with a 1.0.4 version, and installing that part according to the 1.0.4 instructions, fixed the problem for me - thanks, ScottW! :smiley: :smiley: :smiley:

By the way, my notes collecting all the info you need to print and assemble the 1.0.4 Y axis motor mount is on another thread here: https://forum.lulzbot.com/t/y-axis-off-cleaning-hits-pad-holder/6125/3 That post gives the McMaster-Carr urls for the parts, based on the Lulzbot Mini Bill Of Materials, where to get the soldering iron tips for installing heat-set inserts, where the STL file is, etc.

Now that I’ve had a day to think about it, I realize my theory of the 1.0.4 part fixing some Y scale problem can’t be the right explanation. So I have no idea why replacing the 1.0.3 Y motor mount with the 1.0.4 part fixes the nozzle wiping position problem. I’ll just have to be content that it works, with no idea why.