Tired of unsuccessfully calibrating this thing - please help

[miglo]

Hi guys,

I have spent countless, and I mean COUNTLESS hours trying to calibrate this thing. I've followed the manual to the letter for days. Today I took things apart, retightened the screws, made sure the towers are as straight as possible (with a metal carpenter's angle ruler), and made sure the bed was leveled as much as humanly possible. Retried calibrating the old manual method.

Tried calibrating using the "easy manual zprob calibration" from artexmg.

NOTHING is working

I still get this:

[img]https://lh3.googleusercontent.com/-jHrc ... 222838.jpg[/img]
[img]https://lh3.googleusercontent.com/-Lhc6 ... 222845.jpg[/img]
[img]https://lh3.googleusercontent.com/-TBZq ... 222856.jpg[/img]

Between the ZX and ZY it lifts off the bed by more than 1mm it seems so nothing adheres. Between the XY the head is scrapping the bed. Everything is as close to near same height using the calibration. Horiz radius is 130.

My setup is a 2wk old Rostock Max V2, TL arms, TL fly-n-strude.

I tightened all the arms, varied the tension on the fly-n-strude, change the bowden length. I don't see any black powder. I made sure all the cheapskates are just tight enough to roll freely, and that all 4 wheels are rolling equally. Belts were all retightened and appear to be of equal tension all around. I don't know what else to do.

[stonewater]

your horizontal radius is off..... its not sticking between the towers.... towers look good. find it in eeprom, should default at 130.00 make it smaller in .5 increments till it is the same at the bed edge and at the center.

Tom C

[miglo]

My horiz radius is at 130. I've had it at between 131-129.

[bot]

Try 130.2, and get a set of feeler gauges instead of using paper. Get it to within <0.01mm at each tower and center, and that sized test print circle should print nicer. I've noticed that excess play in the ujoints/arms can exacerbate this problem, and even at my "best" calibration, there is a variation at the extremities between towers.

[miglo]

One of the major problems with this, is if I lower one tower, the other one rises. And I do this back and fourth thing for hours. So once I think 1 tower is good, I go back and check the other one, its messed up again. So I fix that, now the other one is messed up.

And I had it at 130.25 just earlier. But after having tried to fix the calibration, I ended up at 130 again because 130.25 was smashing the head to the bed at the center.

I have the TL arms, figuring that would help out with the excess play.

[AndyB]

I've struggled with tuning my machine just as you have. I don't feel in hindsight that it was a waste of time, because I now understand how the geometry of the delta affects the print of the first layers.


"And I had it at 130.25 just earlier. But after having tried to fix the calibration, I ended up at 130 again because 130.25 was smashing the head to the bed at the center."

If the nozzle is scraping the bed at the center, by definition you have a convex (inverted bowl shape) to the path of your extruder across the bed. If it is lower in the center than it is at the Z tower it's concave (bowl shape). This is adjusted with the horizontal radius value. If the center is higher than the tower measurements (after setting the tower height), then decrease the horizontal radius (raise the nozzle).

I wanted to tell you what I do, but I have to go off to my non-hobby (job). Keep at it, and you'll get something close.

(I'll come back and write more later)

[stonewater]

ok so here is how I got my machine fixed after chasing the Z around the towers.

1- set Z zero in the center

2 set z zero at each tower, after each adjustment to the individual tower run the Z to home and back down to that individual tower, and check the individual tower. do not go back to the center Z and do not move on to the next tower until it is calibrated. then repeat until all 3 towers are done. do not go back and check them again.

3- after the 3 towers are done go to center of bed Z- do NOT re adjust this with the normal Z zero calibration-now its horizontal radius time. if it is higher than the towers increase the radius, if it is lower (scraping) decrease the radius. at this point in the calibration Z zero is the center of an arc that needs to be bent until it is flat....

now run your Z zero calibration and recheck your towers.


this took me a while to figure out.

Tom C

[heathenx]

This particular issue is the reason why I sold my Rostock. In the beginning it took me quite a while to get my towers dialed in. In the end I could do it as quickly as 5-10 minutes...BUT I used a dial indicator which I thought was much faster. Still I ran my head down an extra .001" on the very edges with my horizrad. setting.

I often wonder how many people struggle with this. If I was a SeeMe employee I wouldn't rest until calibrating towers was a cake walk for any new user. Constantly tweaking your printer when you don't want to takes the fun out of 3D printing.

[miglo]

2 set z zero at each tower, after each adjustment to the individual tower run the Z to home and back down to that individual tower, and check the individual tower. do not go back to the center Z and do not move on to the next tower until it is calibrated. then repeat until all 3 towers are done. do not go back and check them again.

Tom C

Is this my mistake? Going to check them again I mean. I would dial in the z tower...then y tower...then x tower...then i check z again, and its off by +- .03mm. I fix the z tower again...then y would be off by that much...and round and round we go.

[miglo]

So I spent another 2 hrs this morning trying to dial it in again. I made sure the bed height is with +.1mm at each screw. Checked the tower lean again using the trusty L-ruler.

Trying what you guys have suggested and using axe-something's method with the endstop offsets, I manage to get all 3 towers with .01-.03mm of each other. Sounds great right? Horiz Rad is 130.25. Print out the calibration circles...its improved by maybe 5-10%. 2 inches below the center, at the X-Y is still no man's print zone as the head is slammed into the print bed. 2 inches above to the ZX and ZY, nothing sticks.

I have to rotate my 2-3" pieces just to get it to print in my now 4" diameter print area. Honestly, I'm about to throw this thing out the window. This is truly, NOT FUN.

It is frustrating beyond belief. I think I've spent more time trying to calibrate this thing than actually printing. Being a glutton for punishment, I've bought the smoothieboard and stuff to try the auto-calibration. I really want to like this printer and make it work. I would settle for a 6" print area at this point. My digital angle reader arrives today too, so I'll check to see if maybe my L-ruler is off. If this doesn't work, I'm probably just going to get rid of this and get a lulzbot taz 5 or something and never look back at deltas.

[KAS]

I went through 3 different carpenter squares that were not square enough. You can test yours by drawing a fine line( razor line is better), flip the square over 180 and draw another line next to it. With digital calipers measure the bottom and the top, any difference should give you an idea.

The instructions in the manual only have you squaring the towers to the bed in one direction. It doesn't mention the left/right lean that can happen by twisting the top section or just by resting the towers on that one screw on the bottom.
Either way, the digital angle finder worked wonders for the same issues I was having. If the issues you are seeing are mechanical, the auto leveling will only go so far to correct small issues.

[bot]

I see that you are using a different brand of extrusion than what was supplied with the printer...are they extra tall extrusions? Perhaps that are leaning in and causing your troubles.

[Jimustanguitar]

I've chased this for 2 years, and it's been inspected and looked at by SeeMe employees multiple times. JJ got me fixed up once by shimming the front of my bed with a big lock-nut and I was able to print the test piece you showed pictures of up above.

I stupidly switched to a newer version of the heated bed, and let the magic out... I've spent hours with shims and a dial indicator and can't get it dialed in again now. I'm back to where I have to start prints 4 or 5 times and move them around the bed to figure out where each print will actually work.

[Jimustanguitar]

your horizontal radius is off..... its not sticking between the towers.... towers look good.

Horizontal radius effects the entire printable area though. It makes it all convex or concave, and not just the spots between the towers.

[miglo]

I've chased this for 2 years, and it's been inspected and looked at by SeeMe employees multiple times. JJ got me fixed up once by shimming the front of my bed with a big lock-nut and I was able to print the test piece you showed pictures of up above.

I stupidly switched to a newer version of the heated bed, and let the magic out... I've spent hours with shims and a dial indicator and can't get it dialed in again now. I'm back to where I have to start prints 4 or 5 times and move them around the bed to figure out where each print will actually work.

Hearing this is...depressing

[miglo]

I see that you are using a different brand of extrusion than what was supplied with the printer...are they extra tall extrusions? Perhaps that are leaning in and causing your troubles.

I'm using 626pilot's e3d v6 mount. I had to make the holes smaller for a more snug fit, but otherwise untoched. I printed out another one from another author and will try that one later.

[KAS]

No doubt it's frustrating and it's one of the more common issues once people get past the novelty of printing in the center. Have you ever been able to print the full circle before the mods/upgrades?

I've never hung the hotend below the effector, so I don't know if that would exacerbate the issue when the effector isn't parallel throughout the bed.

[stonewater]

your horizontal radius is off..... its not sticking between the towers.... towers look good.

Horizontal radius effects the entire printable area though. It makes it all convex or concave, and not just the spots between the towers.

I had this exact same problem as he was having, because of where the towers are calibrated, then you would think doing the x y z towers would solve horizontal radius, but it does not. there is some thing different about horizontal radius.


@ Miglo,

so it sound like your Z home is adding when you Z zero, after you set the new Z zero, watch the z position on the LCD and it should go up somewhere around 360 to 365... I dont know what it will be with the hot end you have it will be much lower, perhaps 320 or so....

the only times I have ever had my Z slam into the bed was when my Z was doubled by a software glitch.

Tom C

[miglo]

your horizontal radius is off..... its not sticking between the towers.... towers look good.

Horizontal radius effects the entire printable area though. It makes it all convex or concave, and not just the spots between the towers.

I had this exact same problem as he was having, because of where the towers are calibrated, then you would think doing the x y z towers would solve horizontal radius, but it does not. there is some thing different about horizontal radius.


@ Miglo,

so it sound like your Z home is adding when you Z zero, after you set the new Z zero, watch the z position on the LCD and it should go up somewhere around 360 to 365... I dont know what it will be with the hot end you have it will be much lower, perhaps 320 or so....

the only times I have ever had my Z slam into the bed was when my Z was doubled by a software glitch.

Tom C

The more I read, the more this just seems like an ongoing struggle for a lot of people. My X-Y zone the hotend scrapes the bed. My ZX and ZY zones, the hotend lifts up .5-1mm.

I ended up shimming ZX and ZY with 2-3 sheets of paper just to be able to print...and even then its within that 6" diameter, a far far cry from the 12" print area. This is a really sad solution.

[KAS]

How are you measuring or determining where to place shims?

[stonewater]

you should not have to shim, you should not have this problem, there is an answer to it. there are thousands of rostocks happily printing away all over the world. we just need to figure out what is going on.

1- have you done a clean install of the config.H? if not write down your pid settings, and if you had to change your stepper directions write those down also, and do a clean install of the software .91 release

the reason I say this is there are settings in it that affect the stepper motor steps, and they may not be correct, need to find out which stepper motors you have, and make sure that the config.h matches your steppers.

2- take that hot end off and put it on the correct way, it is probably amplifying the tiny errors that may be there because it is so far from the center point of the effector. or better yet put the stock hot end back on for now.

3- put a dab of locktite on the screws that hit the endstop switches so they do not rotate from vibration.

4- the feeler gauge is a great idea..

let us know how it goes....

Tom C

[drfuyutsuki]

One thing we did was replace the stock end stop screws with socket cap screws. It makes it easier to make minute changes to the end stops.

[miglo]

I swapped out of 626pilots below-the - effector mount to another one that had it sit on top, with the distance closer to stock. This helped a little bit. I'm eagerly awaiting the digital angle gauge that should arrive today. I'm hoping it's just crazy tower lean.

As to where to shim, I just put it mid way between the zx and zy. That helped with the adhesion for those two zones.

On another note, adjusting the enstops via the offset in eeprom for some reason no longer works. No matter what values I put in, it's completely ignored. Strange and frustrating. I may do what stonewater suggested and just reflash the firmware.

BTW I have the kysan stepper motors, so I'll need to adjust the voltage.

[TFMike]

Is this my mistake? Going to check them again I mean. I would dial in the z tower...then y tower...then x tower...then i check z again, and its off by +- .03mm. I fix the z tower again...then y would be off by that much...and round and round we go.

Yes, goto page 234 here and you should be sorted out in an hour or so, I was scared to do it but did it just now and it worked out fine: http://download.seemecnc.com/rostockmax ... dition.pdf" onclick="window.open(this.href);return false;

[stonewater]

On another note, adjusting the enstops via the offset in eeprom for some reason no longer works. No matter what values I put in, it's completely ignored. Strange and frustrating. I may do what stonewater suggested and just reflash the firmware.

I have never had to adjust endstop offset in eeprom.

Tom C