During assembly, discovered my build wasn’t as square as I thought, had to redo some work. Despite efforts to square, I still had different sized gaps on the left and right side. Ended doing some test Crown prints and using Klipper skew correction to get Square prints (in XY), even if I couldn’t get the frame physically ‘True square’ in XY.
Yeah, luckily I used a square while I was building the frame and checking now - frame is squared. As I said all good without belts. I just tighten belt on one side and it is squared, but I am questioning now if I have too much tension. Is this a thing?
I just learned on my most recent build that the best thing to do is to get the X axis as square as possible BEFORE putting any belts in. With no belts attached, leave the X rail mounting bolts (4 bolts) loose so you can wiggle each side back and forth somewhat independently. Then I used a pair of calipers to put both XY trucks in the same position dimensionally from the back extrusion rail. Once I had them both in the same spot (square) I then put stop blocks on either side of each XY truck to keep them from moving. Now I can belt the XY system knowing the X-axis is set square and cannot move. This results in a pretty square axis when I am done and then can make small adjustments with the tensioning.
Personally haven’t figured out how to prevent for all situations (e.g. power loss, reset or filament sensor trigger). Have tried to reduce number of conditions that result in freefall via Klipper Macro that moves to safe height after print completes, or before idle timeout powers down the steppers.
In the event of power failure… not much you can do.
Maybe brakes on the Z axis, if the ENA pin on the motor driver is logic high, the brake is released, otherwise engaged? This might be fast enough to catch the Z axis in a power fail. Will almost certainly be enough on a motor power down, but I don’t know that I’d want to coint on it for a print resume, given that a mechanical motion is necessary to engage.
My v4 has a home routine using sensorless homing to the bottom of the travel, and I add that to the end of my print routines… but that can’t work on the SKR boards. You could add optical stops and use v3 firmware, maybe? At least I don’t get the bed dropping on power down.
Counterweights are tricky, because while you can account for an unloaded bed, once the print starts, the weight is harder to balance. I suppose you could probably get close enough for friction to manage. But it’s kind of messy at best, and subject to all kinds of potential problems.
when looking from the front opening, Z is the front left, Z1 is the center rear, Z2 is the front right. (at least on my machine as defined in the klipper printer.cfg file) You can test these in klipper with
Apparently my problem was me not resetting EEPROM after changing XY positions for z leveling. Anyways, here is my very first print without cooling fan (amazon did not deliver it today and I could not wait)
While in the process of input shaping I noticed there is rippling not associated with acceleration or frame vibration. It is much more prominent then Y axis is moving and less noticeable on X. As the speed of printing increase - this tippling starting to smooth out but spacing between ripples is the same regardless of the speed. I am assuming this has something to do with the motor /belt / gears on the Y axis?
I think your belts might be a bit too tight. I think that stuff is the concentricity of the idlers and pulley’s but it only shows when the belts are super tight. The belts across the back should not make a sound when plucked.
I think I found a speed limit of my printer. This benchy printed at 500mm/s at 6000mm/s^2 acceleration. Still came out vey good, no steps were skipped but printer was jumping in the table
I must just suck at building 3D printers. I couldn’t get mine to run at 50/mms with 3000 acceleration lol. Here you are at 500 and 6000. I think my acceleration is down to 800-1000 and it’s running along happy as can be. So that makes me happy lol