Stop throwing your nuts!!!

Homing the Z axis resulted in throwing the nuts every time when I started using the RoBo 3D. No matter how I adjusted the switch, the carriage seemed to drop onto it but not trigger it.

Finally, I put my finger on the plastic T bar as it approached Z = 0.00, pushing down as the nut lowered it, and the switch triggered, stopping the descent perfectly.

I have since slipped a threaded pipe connector over the long bolt, adding just enough weight to ensure that the switch is pushed down enough to trigger the "homed" signal and stop the descent. The connector is much larger than the bolt so that there is no chance of it catching, and just heavy enough to make the stop screw push down the switch arm instead of letting the switch hold up the carriage.

 

Hey everyone,

I had this problem, too. The problem didn't appear to be left side/right side calibration. The problem for me seemed to be that the activation of the z home switch was really variable / unpredictable.

A useful test I found is to take the z axis near the bed and then move the head up and down by .1 increments and note where the switch opens and closes. I was having problems where the switch was closing unpredictably, and I'd throw nuts all the time.

Now, the structure of the switch is like this:

+-----\
+----o+

Where the top bar is a piece of metal, and the o is the switch location. Now, the way my printer was shipped, the calibration bolt was positioned like this:

|​

|​

+---\
+--o+
Or, directly over the bolt. I noticed in some pictures of calibration, it appeared that the bolt wasn't directly over the switch for other people. I suspected manufacturing variability, so I'm running an experiment:
1. Removed top cover
2. Removed switch assembly
3. Reversed switch assembly
4. Put it back together.
Now, it's more like this:

|​

|​

+------\
+-----o+
That is, the calibration bolt is now over the metal bar, but not the switch. I don't know if that's how it's supposed to be, but I have some early optimism about the fix, as I was throwing bolts 80% of the time before, and it has been reliable on 100% of a few prints (though I ran out of time).
Let me know if I'm insane and shouldn't be doing this.
Your point, tesseract, is a good one though... in order to throw bolts that would have to mean something was resting on something as the hot end came down. I tried to figure this out a few times. I determined it wasn't the head on the bed, but I didn't prove what it was. I fancied, I guess, that it was the calibration bolt resting off to the side of the switch without activating it, but I don't know that for sure. I'll find out more in the next couple of days and post back.
Cheers,
Pete

 

I will check that out.

I agree that I should not have to do that.

However, I really looked carefully, and it seemed that the switch was actually pushing up the right side carriage instead of the right side carriage triggering the switch. And, just as Pete noted, the screw pushes on the switch (on my printer) very close to the fulcrum; I wondered about re-positioning the screw or the switch or both.

 

tesseract: your message suggests that on your printer the adjustment bolt is not directly over the sensor part of the switch. If it were, bending the metal wouldn't do much. I'm sorry the diagrams suck... among other things, this message board removes extra whitespace which makes things hard.

I'll take a picture of what my switch looks like after the reversal if I get a chance.

Here's the picture I saw in the setup guide that made me suspicious:

It pretty clearly looks in that picture like the bolt (right side of red rectangle) is over the METAL part of the switch, not the actual sensor, which is on the left. If you imagine something like this, where the bolt is directly over the actual sensor, that's what mine looked like before I turned the switch around. You can see in that picture the back side of the two bolts you unscrew to remove the switch. It's reversible, so that's what I did. I had to make a little play in one of the wires to get it to work.

Hope that helps with diagnosis. Like I say, it seemed more reliable on the first couple of prints, but it's too early to say.

 

Thanks for the picture from the manual, Pete. Mine definitely does not look like that.

So, it appears that re-positioning the switch to move the stop screw farther along the trigger arm would not only solve my problem with activating the Z axis home signal but also bring my printer in line with current specifications.

 

I let Braydon know about the forum post. If you reverse your switch, let us know if this fixes your problem. It's probably pretty obvious, but the steps were:
1. Snip the wire tie that holds the hot end wire conduit (right under the top of the housing). Careful not to snip any wires!
2. Remove four bolts around the base of the housing.
3. Remove housing.
4. Identify two bolts that hold the z-axis switch in place. You can see the back of them in the picture above.
5. Remove those two bolts.
6. Reverse the switch. Verify that the z-axis adjustment bolt now is over the metal part, further along the arm than it was. If it isn't, this isn't going to help.
7. Replace two z-axis switch bolts.
8. Replace housing, and bolts for housing
9. Re-attach the conduit to the top of the printer housing. I used a twist-tie rather than a wire tie, in case I need to do it again!

 

The stop screw is through the "T" of the carriage on my printer, not on a separate "L" plate as in the picture from the manual that you posted. Reversing the switch will move the activator to the far end of the support arm and put the end of the metal strip under the screw, instead.

You can just make out all that in the picture that I posted, above. It looks as though the design changed, and my Z axis switch was mounted for the other iteration of the stop screw mounting.

 

My z -axis started acting up, and the adjustment screw was practically on top of the microswitch button, and the weight of the carriage would not always activate the Z-axis switch. When this would happen, the printer would throw it's nuts, as it never found the Z-axis home position.

For now, I just bent up the microswitch laver, so the nut directly activates the microswitch button. Not as ideal as repositioning the switch to the right place, but it is working for now.

 

I turned my Z-axis switch around (which was no big deal -- cut loose or unscrew the cable holding the wiring to the top of the cover, unscrew the four bolts from the sides of the cover, lift off the cover, unscrew the two bolts holding the switch, rotate the switch, repeat the first few steps in reverse order), and it was improved, but it still threw the nuts because the switch is stiff enough to lift the carriage when it first makes contact. Once the other side has dropped a bit more, the switch is thrown.

So, I put the weight back on, as pictured, above, and it has been working like a charm, since.

 

With the switch turned 180°, the stopper bolt is making contact roughly in the middle of the lever arm, slightly closer to the actuator.

As best I can tell, everything is level and straight. When the chassis lowers, the bolt makes contact with the switch and that side of the chassis momentarily stops dropping. The nut screws out a bit (the other side continues to drop, but the extruder hits the bed if it is already homed on the X-axis), and then the right side chassis drops, triggering the switch.

With the added weight of the pipe connector, the switch is triggered smoothly.

I agree that there should be enough weight on there, already. I conclude that my switch is unusually stiff.

The extra weight on the Z-axis screw rod on the trigger side does not seem to have any other effect on the operation of the printer so that I am using it with the weight, now. If someone wanted to send to me a replacement switch, I would be happy to swap out my current one to see if it is just a stiff switch.