RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I just grabbed the LA15/0.9 version of the firmware and I'm getting the following compiler error:
sketch\menu.cpp: In function 'menu_goto':
sketch\menu.cpp:63:1: internal compiler error: Segmentation fault
}
^
Please submit a full bug report,
with preprocessed source if appropriate.
See < http://gcc.gnu.org/bugs.html> for instructions.
lto-wrapper.exe: fatal error: C:\Program Files (x86)\Arduino\hardware\tools\avr/bin/avr-gcc returned 1 exit status
compilation terminated.
c:/program files (x86)/arduino/hardware/tools/avr/bin/../lib/gcc/avr/5.4.0/../../../../avr/bin/ld.exe: error: lto-wrapper failed
collect2.exe: error: ld returned 1 exit status
exit status 1
Error compiling for board RAMBo.
I grabbed an up to date version of the 0.9 firmware and was able to compile it without error. From the error it looks like it's just a stray bracket in menu.cpp. I've made only the same edits as I did in both firmwares. Editing the config.h for language, then making the couple of edits in the copied/renamed Configure_prusa.h file.
Any ideas?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
weird. I just downloaded the LA15 + 0.9 from my repository and it compiles fine on my Mac. Tried both MK3 and MK3S variants and both were successful.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I've been compiling them on my Windows machine without issue. I'll throw it on my MacBook and see if I get a successful compile. Odd though that the normal one will still compile, but not the new one.
Just to confirm what version of Arduino are you running? I have 1.8.8 on my Windows machine.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Arduino 1.8.9 on my Macs. The new one compiles fine on it.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
BTW, I'm also finding the LA15 optimal k values to be quite low. On latest test,
PETG, Mosquito, Tungsten Carbide nozzle 0.4 mm, BNBSX 54:16 gearing, the optimal k is 0.1
Set up your Marlin linear advance test code to test in the lower range. The generator's defaults will be too high and coarse.
http://marlinfw.org/tools/lin_advance/k-factor.html
I set mine to step by 0.025
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Yup... I already updated this morning before reading your reply to 1.8.9 and it's compiling correctly. May want to update your note about Arduino version on the Github.
I was testing the LA15 versions of the firmware before I jumped on board with your 0.9 firmware. I'm wondering why they altered the scale to be so radically different. Not so much in the values, but the decimal representation. Just seems nice whole numbers are easier for people to understand in general and far less of a chance to misplace a decimal point.
In other news, in complete frustration, I finally got and wired up my OMC 0.9s for the X and Y. First off... they are ridiculously smooth to index by hand compared to my Moons. So I think you were right with your first diagnosis that my Moons were bad. Just insane that 3 new moons shipped directly from them would all be bad.
I believe my Moons were actually fighting to overcome their own magnetic resistance because they were super notchy. So the driver couldn't get a good reading of when endstops (or misidentifying crashes/missed steps in stealth mode) were being triggered. I've already sent a request to get them replaced to Moons (via Amazon). We will see what they say.
In comparison, my Moons 0.9 and 1.8 have about the same level of resistance to being turned by hand. The OMC, however, is buttery smooth with very light force needed to rotate with just fingertips. Significantly less than the Moons by a large margin which I had to grasp pretty tightly to rotate. My reluctance to move away from the Moons was belief they had to have better QC and bad motors should be pretty rare. It would seem to get 3 out of 3 bad motors in a single delivery would imply they have some manufacturing issues unrelated to possible shipping damage. So your statement to someone else from a while back was dead accurate: "A bad Moons is far worse than an average OMC".
When I made the swap I also threw on the Balitensen pulleys the printer couldn't care less. I'm still having issue passing a self-test (getting a Y-axis length error even after running a new XYZ calibration), but I can set my belt tension to pretty much whatever I want now without the worry of homing failure. I was also able to finally print successfully in stealth once again.
I did set both of my lin-corrections for X and Y to 1.130 as you had noted. Is that the correct value, or is there a test print to calculate that value?
The first few test prints looked really good, going to run a few more
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Recall, I also suffered through two bad Moon's in a row.
I have added additional note to test how the moons rotate upon arrival. They really should be similar to the LDO's in their smoothness - not notchy.
Also updated the Arduino version note in my readme
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Yes linearity correction 1.130 to 1.140 is what I got as best for the LDO's.
I suspect there is something else going on that is causing your y length to fail during calibration. Is the total free travel being reduced by unusual stop or belt mount?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
It's free sliding along the entire length of the bed with when no belt is connected and will slide when the machine is tilted... the only thing I think could be affecting Y it is my heater bed cable. When the steppers are disabled and the bed is slid all the way toward the control panel it will spring back about 1cm the first time. It's not much, but I guess it's possible. I do have a piece of nylon filament in there to keep the wires from sagging, but I may need to adjust the angle, and length a bit to free it up. The reported size mismatches I get on the console are from 1-3mm.
It's possible that the heater bed cable and extruder bundle (on the occasional fails I'm getting on X) are causing some tension on the axis movement, but it's really slight. I can't find anything else that could be impeding the movement. Both axes move freely without any sort of binding. I'm still a bit suspicious of the Balitensens... the machining of the teeth is pretty sloppy with low spots. I wanted to try them to see if I could reduce the last few VFAs before I try out my preferred pulleys. The Misumi ones arrived and they are beautiful and look like a precision machined part, but are a bit longer than stock in length. I'll take a few photos later. I also ordered one of their 50 tooth gears to try to swap onto the BNBSX. The 56 they make is far too large has some huge gear lips to retain the belt which would hit the print fan. I'll report back when the 50 tooth arrives.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
It's free sliding along the entire length of the bed with when no belt is connected and will slide when the machine is tilted... the only thing I think could be affecting Y it is my heater bed cable. When the steppers are disabled and the bed is slid all the way toward the control panel it will spring back about 1cm the first time. It's not much, but I guess it's possible. I do have a piece of nylon filament in there to keep the wires from sagging, but I may need to adjust the angle, and length a bit to free it up. The reported size mismatches I get on the console are from 1-3mm.
It's possible that the heater bed cable and extruder bundle (on the occasional fails I'm getting on X) are causing some tension on the axis movement, but it's really slight. I can't find anything else that could be impeding the movement. Both axes move freely without any sort of binding. I'm still a bit suspicious of the Balitensens... the machining of the teeth is pretty sloppy with low spots. I wanted to try them to see if I could reduce the last few VFAs before I try out my preferred pulleys. The Misumi ones arrived and they are beautiful and look like a precision machined part, but are a bit longer than stock in length. I'll take a few photos later. I also ordered one of their 50 tooth gears to try to swap onto the BNBSX. The 56 they make is far too large has some huge gear lips to retain the belt which would hit the print fan. I'll report back when the 50 tooth arrives.
I opted for something a bit different on my setup and have been running the X and Y on 0.9 for a solid 3 weeks now with no failures.
I decided to go all Gates on my pulleys, But I opted to go with the "serious business approach on my Y and X idler. I have 2 additional bearings in each making it silky smooth. When I first set up the bear, I was having the same issues with X homing as I was with the stock frame. The skew was absolutely terrible on the stock frame which was causing everything to bind. I was running bear X ends at the time too and still found no resolution. After the bear was finished, I ended up fine tuning all the measurements , reprinting the X ends with all new belt, idlers and such. after that it was misumi rods and bearings (biggest difference) and added a spare vesconite bearing to the bottom rod of the x axis. For the first week I was still having a bit of issue homing, but this was a "break in" phase I guess, because not long after I began the quest of trying to home in stealth mode.... First time it worked! Super happy with it so far! I am still tuning in retractions with exotic filaments like PC/CF but my main worry is that I would have issues once I got the enclosure at about 60C.... Not a issue whatsoever 🙂
All credit goes to our hardworking community and of course... Guy-K2
Guy-K2 is a god among men!!! 🙂
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Does the frequency of VFAs change with printer speed? I've contemplated taking the 0.9 degree jump and ran a cube on my stock motors to see how bad it really was - and was playing around with the front panel speed control; I noticed that when I changed it to 150% they became more pronounced and about 2x as frequent on X, but the opposite on Y.
Pic: You can see a band about midway up where the speed was changed.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Does the frequency of VFAs change with printer speed?
Yes, those are VFA's. The doubling you see is non-linearity happening at two points in the e-cycle. When you change velocity you also alter vibration characteristics of the carriage. That changes which components of VFA can visibly alter velocity. They still remain vertical and aligned. With good 0.9 degree motors, the VFA's greatly decrease. It's a reversible change to the machine.
If you want to test as cheaply as possible, do it with OMC motors (provided you can solder the wiring adapter). The Moons' are great, but you must test them upon arrival by turning their shaft with your fingers. The detentes should feel much lighter and finer than that of the original 1.8 LDO motors. If a Moons' is notchy, don't accept it.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Thanks for confirming; I'd already planned to pick up an OMC at some point (since it'd been noted it performed better with linearity correction enabled) so I'll be back when that happens 🙂
May start with just one for tinkering and a second as time/slush fund allows.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Having trouble compiling on Windows:
eeprom.h:221:1: error: non-constant condition for static assertion
static_assert(((sizeof(M500_conf) + 20) < EEPROM_LAST_ITEM), "M500_conf address space conflicts with previous items.");
^
eeprom.h:221:1: error: the value of 'EEPROM_Sheets_base' is not usable in a constant expression
sketch\eeprom.h:191:23: note: 'EEPROM_Sheets_base' was not declared 'constexpr'
static Sheets * const EEPROM_Sheets_base = (Sheets*)(EEPROM_NOZZLE_DIAMETER - EEPROM_SHEETS_SIZEOF);
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
We know the firmware compile successfully when the Arduino environment is correct.
Focus on finding what is different in Arduino setup.
Is it version 1.8.9?
Delete and reinstall Arduino IDE.
Repeat initial IDE setups.
Bring down a fresh copy of the firmware (but pretty unlikely you got a corrupted version)
Reattempt compilation.
If fails to compile, try downloading and compiling Prusa's firmware. Can you compile that?
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I was in the MK3 branch by mistake. I switched to the LA15 and 0.9 and it works now.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
I tracked down the issue, which was unrelated to the firmware, causing my y-axis length and y-axis loose pulley messages during a self-test.
The Y axis belt setup with stops I was using had an incorrect overall travel length of ~213mm (the y-belt holder was too long). So it was failing to get either a measurement that was long enough to meet the minimum 215 required, And also having some odd bounce issues when the belt was just a bit too loose that could cause a measurement difference of anywhere from 2-5mm difference between the measurements.
I remixed a couple of different length parts from the Serious Business MK3 Y belt holder to reduce it from 52mm to a 50mm and 48mm versions. I ended up needing the 48mm to get a 215.890 measurement back to back. Serious Business doesn't allow remixes, so I am happy to share the files in case anyone wants or needs them.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Guy - I'm curious - what is your explanation for this pattern? Blips are 0.32 mm apart, or two full motor steps. VFA's from the micro step table would be 1/2 the spacing. So I'm thinking it's down to coil or driver anomalies.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
Tim, I don't know. It's even odder that it looks to appear every other extrusion layer.
1. Do they stay the same spacing when you vary print speed? It it is tied to the e-cycle, they should remain same spacing but vary in intensity as the speed changes alter vibration frequency but not spatial synchronization.
2. Do the stay same spacing if you print with object 45 degrees rotated? It they did, then I would be thinking about extruder rather than x/y.
Neither of the above should be an every other layer phenomenon.
RE: Stepper Motor Upgrades to Eliminate VFA's (Vertical Fine Artifacts)
They are X-Axis only, speed invariant, definite phase reversal alignment. Diagonals have the same modulation - a little tough to see in this image. But shows that it is X, and Y is clean. Short of setting a new motor, I'm just going with an odd winding in my X motor. This image is 25 mm squares. The first image was a 100x50 mm patch, so are different speeds. And in any patch - with accel - there are multiple speeds showing, yet pattern is unchanged.
As for every other - we're are seeing a directional artifact, aka snow plowing: nozzle is putting down more plastic one way, slight offset in perpendicularity.
Don't have the patience to do a 4-wire milliohm measurement of the windings, and my 5 digit meter is not helping much (stops at 0.1 ohm resolution and I am too tired to remember how to get to 0.01 ohm res... lol (manual says it is possible, but totally evading me). Great - got out the manual - it says minimum reading is 0.01 ohm, but later says best displayed resolution is 0.1 ohm. And this is Fluke ... lmao. But haunting me is the fact I know I have seen a 0.00 ohms display and I've used it before.