Fix z-banding w/ no backlash- "Slop Nut" (Level 2 techie - Consumer ) (Page 5 of 7)

im getting my printer dialed in more and more....but as i do i noticed my banding is HORRIBLE!  It probably always was bad but i never saw it because i was messing with other stuff.  I still have the stock 5/16 rod and just wanted a general instruction on this???  So i go with what file (post # or link would be great)?  Do I remove the stock nut on the bottom of the bed?  What needs adjusted about the nut/nuts inside the enclosure?  Do i leave the 4 m3 screws slightly loose that holds it to the bed hence the reason for 2 nuts?  Would 1 locking nut work instead of the 2 nuts?  As asked by gettingbored above do i need to wait for the one with the anti backlash nut?

Thanks guys...my mind just runs 100mph and i like to get a whole project sorted out in my head before i start anything

I have done this upgrade with the M5 rod and the results for Z banding are great.  However I am having a major problem.  When the bed changes direction I am losing .5 to .4 mm of travel and it takes about 4 movements to recover to normal layer change height.  This would seem like a backlash problem but I can't figure out where the issue is.  I am using a flex coupler and the aluminum bed.  The nuts holding the slop nut mechanism to the bed are not overly tight, the rod can move but not as freely as in 2n2r5's video.

One possible problem is the bed itself.  Without the rod connected, should the bed freely slide up and down the rods in the back?  If you were to let go of the bed should it just fall to the bottom?  Mine moves if you move it with your hands but if you just let go of it, it will stay put.  I am wondering if I could have bent the rods during disassembly/assembly of the platform or is there a particular way it must be assembled to get the free movement?

Thanks for any help.

There was a small spacer plate that goes on first which is .5mm. If that didn't get installed then you would have that much hysteresis in the z axis. If you don't feel like ripping it apart to put in the spacer, you can try fixing it in firmware. If you use the m99 command you would just add M99 Z.5 to your start code before it homes the bed.

I have an update to the 5mm aluminum bed slopnut. I added the file below. It has a bit more play. Just make sure that you tighten all nuts and bolts evenly but do not over tighten. I generally tighten the bolts down until it just starts to affect the movement on the threaded rod then back it off a 1/4 turn. Also, make sure that you have a little grease between all the layers. When you get finished the rod should move freely. The rod might not necessarily go back to center when you release it. Mine would go back to centerish after i ran it up and down a little. the important thing is that the flex happens inside the nut and not in the bed.

first 3 lines of the start section of G-code editor in slic3r would read something like this

G21; set mm units
M99 X0 Y0 Z0.5 E0 ; set hysteresis values
G28 ;home all axis

If you have a dial indicator you can figure out exactly what to change this to.

i'm not totally sure I understand what you are measuring there. Is that measuring you pressing down then up over and over? so basically there is no change in height? If so that means you have more than 1mm in hysteresis. Here is my procedure that I follow...

---- Use this procedure to null position -----
Enter "M99 Z0" into G-Code line of manual control tab in Repetier-Host
Home your bed to 0mm using home button
Then Lower your bed 10mm
then raise it 10 mm

---- Use this procedure to measure Z hysteresis ----
Zero Dial Indicator
Note Postion of bed (should be 0 in RH)
Hit Z+ 1mm until you dial indicator starts moving at 1mm intervals
Note new position in repetier-host (calculated position)
Note the reading on the dial indicator (measured position)

---- Math/Set/Verify -----
hysteresis setting = calculated position - measure position
enter m99 z(hysteresis setting) into G-code line of manual control tab in Repetier-Host
enter M98 to read back settings to ensure they match what you entered

Now if you try any z movements you should notice 1mm should equal 1mm of movement for any direction.

I did the procedure you listed but it didn't work.  I still don't get 1mm movement.

One thing I did notice is that when I enter M98 I get:

H=X0.00 Y0.00 Z0.29 E0.0 SHIFTS:X=0.00 Y=0.00 Z=800.0 E=0.00

What does the Shifts: Z=800.0 mean?

I did enter the M99 Z0 and then M99 Z0.29.  Regardless of what M99 value I enter for Z, M98 shows the SHIFTS: z=800.0

Edit: After trying a few different values and then reentering Z0, the Shifts now reads Z=81.00

Yes it is taking the value, I just thought maybe that shifts number might have something to do with it.

I have set the M99 Z value to .15 and that gives me a 1mm movement on the dial.  The second movement is only .96.  After that each move is right around 1mm (give or take .02 or so).

However, if I home the Z axis, zero the meter and then give a 10mm movement it only moves 9.14mm and then when I move it back up 10mm the meter reads -0.06 indicating that it is short .04mm.

I think I may be on to something that will fix it.  I appreciate your help and will keep working on it but don't want to take up your time.  I will post here if I need more help or figure out what fixes it.

Is there a way to make the hysterisis value permanent in the printer or do I have to enter it every time?

On a side note, the .stl for the 5mm aluminum bed slop nut, the holes in the nut holders are only M3 size.  I had to drill them out to make this work.

I have not done the Z steps, I put 4000 in as that seemed to be universally accepted number.

Is there a procedure around for it?

4000 should be the correct value for an M5 rod, as the motor takes 3200 steps per rotation and an M5 has a thread pitch of .8 mm per rotation (3200/.8 = 4000).  You shouldn't adjust this value as threaded rods don't really vary in pitch; I think you'd only need to do that with a threadless rod/nut.

Well, I am about to admit defeat on this.  I have done everything I can think of to remedy this.  I am starting to think that maybe something happened to the motor in all of this as it is the only thing I haven't changed.  I have tried different nut holders (even created my own with three nuts super glued in it), I have printed a solid coupler to replace the flex coupler.  I am at my wits end on this.

Does anyone have any other suggestions as to what I might try?  The hysterisis helps, but doesn't cure the problem and it causes the layers to look terrible.

No matter what I do, when the motor changes direction I lose .2 to .3 mm movement.  And over the next 3-4 layers the loss becomes less until about the fifth layer when the layer changes are the correct amount.

Anyone know what the part number for the z axis motor is?

I'm sorry to hear the slopnut didn't work for you. I should put in the instructions that you should put all the components together outside the machine. Then you can double check to make sure the heights of everything will match up. I am glad you got everything working though. Make sure you post some pictures.