Rostock Max bed very convex - radius doesn't help

[itilguru]

This has got me stumped, and I hope someone can help me, please.

Here’s my mixed hardware. I bought the Rambo and servos first thinking I could build my own printer, but then bought the Rostock Max kit without electronics in April 2013. I've only just now finished building it!

Rostock Max kit with second generation pulleys (have the black bumpy texture like in geneb’s latest manual) and Onyx bed, but before they started shipping with EZstruder. I did buy an EZstruder before building this, though. Not relevant, since I can’t print yet anyway.
Repetier-Host 0.90C
Rambo 1.0d board
Because of this board, I modified default settings in configuration.h from 16 microsteps to 8 microsteps. The following settings are impacted:
MOTHERBOARD 301
MICRO_STEPS 8
MICROSTEP_MODES (8,8,8,8,8)
EXT0_STEPS_PER_MM 46.2 //Note that I haven’t gotten to the extruder part yet due to convexity issues – keep reading

Some other interesting settings were left at default:
DRIVE_SYSTEM 3
DELTA_DRIVE_TYPE 0
BELT_PITCH 2
PULLEY_TEETH 20
STEPS_PER_ROTATION 200
MOTOR_CURRENT (195,195,195,195,0) // because I only have one extruder, I guess. Rambo 10.d can handle 1.6A each.
DELTA_DIAGONAL_ROD 269.0
PRINTER_RADIUS 195.5

X_AXIS_STEPS_PER_MM 80 (same for all three axes, which matches EEPROM)

Due to some wiring issues that I can’t track down, I have two axis reversed:
INVERT_X_DIR true
INVERT_Y_DIR true
INVERT_Z_DIR false

My height settings are perfect at X0 Y0 Z0
Y_MAX_LENGTH 367.1
X_MAX_LENGTH 367.1
Z_MAX_LENGTH 367.1
This height matches real world measurements pretty closely. It's the same in EEPROM and in the Printer Settings.

I have been able to calibrate the hot end and Onyx bed. I went through the latest geneb manual step by step to calibrate, but I did end up calibrating Z0 X0 Y0 first for some reason.

I ran into lots of problems with bad connections, but bought a new wire stripper and crimper and fixed every last one of them. I work on computer infrastructure for a living and have worked with computers for 30 years. Still, I’m new to building things with my hands, and new to 3d printing. This thing has humbled me several times already!

I’ve leveled the axes with a t-square. I’ve measured the height of each axis to within 1 mm. I adjusted the endstop screws so all three edge points in geneb’s manual are the same value, it’s just not zero!

At the edges, when the hot end touches the paper, Z reads 20.1 mm. By edges, I mean X0 Y90, X77.94 Y-45, and X-77.94 Y-45.

To be clear, I’ll restate the issue:
To gently pin the paper, I can run G28, then G0 Z0 X0 Y0 F3500. Perfect every time.
I can run G28, G0 Z20.1 X0 Y90 F3500. Perfect every time.
I can run G28, G0 Z20.1 X77.94 Y-45 F3500. Perfect every time.
I can run G28, G0 Z20.1 X-77.94 Y-45 F3500. Perfect every time.
But if I try to go lower than Z20.1 at the edges, it crashes into the bed. Clearly it’s too high at the center. Is that convexity or concavity?

I tried changing PRINTER_RADIUS. When I change it to 180, the Z at the edges reads even higher, about 24mm, while X0 Y0 Z0 is still perfect.
When I change the PRINTER_RADIUS to be higher, the arms push past the limits and it’s really ugly. I won’t do that again. Remember, by measuring from the posts to the center, it’s pretty close to 195.5 mm.

I need to know how to get the edges and the center to be the same Z height. They are FAR off today.

If I were to set the height to be another 20mm larger, then the Z0 X0 Y0 would be 20mm inside the print bed, which is bad. Also, that’s not what I measure from the hot end tip to bed center when at G28 rest. I measure pretty close to my current setting.

What can I try next?

Tom

[JohnStack]

Same response to the other new post... (going top down, not oldest first, sorry):

http://forum.seemecnc.com/viewtopic.php?f=37&t=2365

Search for more about this - it's an ongoing problem...

[itilguru]

You've missed that I'm not off by a few mm, I'm off by 20mm! I've never seen anyone post off by that much.

[JohnStack]

[edited] I didnt miss a thing. Your eeprom is not seeing your settings or reset your eeprom to the newest base settings and start over. don't picnic. excuse my brevity, from my mobile.

Lots of posts about this in the forum.

Repetier-Host 0.90C
Rambo 1.0d board
Because of this board, I modified default settings in configuration.h from 16 microsteps to 8 microsteps. The following settings are impacted:
MOTHERBOARD 301
MICRO_STEPS 8
MICROSTEP_MODES (8,8,8,8,8)
EXT0_STEPS_PER_MM 46.2 //Note that I haven’t gotten to the extruder part yet due to convexity issues – keep reading

Some other interesting settings were left at default:
DRIVE_SYSTEM 3
DELTA_DRIVE_TYPE 0
BELT_PITCH 2
PULLEY_TEETH 20
STEPS_PER_ROTATION 200
MOTOR_CURRENT (195,195,195,195,0) // because I only have one extruder, I guess. Rambo 10.d can handle 1.6A each.
DELTA_DIAGONAL_ROD 269.0
PRINTER_RADIUS 195.5

X_AXIS_STEPS_PER_MM 80 (same for all three axes, which matches EEPROM)

Due to some wiring issues that I can’t track down, I have two axis reversed:
INVERT_X_DIR true
INVERT_Y_DIR true
INVERT_Z_DIR false

My height settings are perfect at X0 Y0 Z0
Y_MAX_LENGTH 367.1
X_MAX_LENGTH 367.1
Z_MAX_LENGTH 367.1
This height matches real world measurements pretty closely. It's the same in EEPROM and in the Printer Settings.

I have been able to calibrate the hot end and Onyx bed. I went through the latest geneb manual step by step to calibrate, but I did end up calibrating Z0 X0 Y0 first for some reason.

I ran into lots of problems with bad connections, but bought a new wire stripper and crimper and fixed every last one of them. I work on computer infrastructure for a living and have worked with computers for 30 years. Still, I’m new to building things with my hands, and new to 3d printing. This thing has humbled me several times already!

I’ve leveled the axes with a t-square. I’ve measured the height of each axis to within 1 mm. I adjusted the endstop screws so all three edge points in geneb’s manual are the same value, it’s just not zero!

At the edges, when the hot end touches the paper, Z reads 20.1 mm. By edges, I mean X0 Y90, X77.94 Y-45, and X-77.94 Y-45.

To be clear, I’ll restate the issue:
To gently pin the paper, I can run G28, then G0 Z0 X0 Y0 F3500. Perfect every time.
I can run G28, G0 Z20.1 X0 Y90 F3500. Perfect every time.
I can run G28, G0 Z20.1 X77.94 Y-45 F3500. Perfect every time.
I can run G28, G0 Z20.1 X-77.94 Y-45 F3500. Perfect every time.
But if I try to go lower than Z20.1 at the edges, it crashes into the bed. Clearly it’s too high at the center. Is that convexity or concavity?

You have to get the edges correct before worrying about bed settings.

Another read through this:
First, you can pin at the center, then you can or cannot pin at the edges? Your code says Z20 for the edges and you're lowering it from there and then it is perfect or perfect that it goes to Z20? Z0 at the edges should pin the paper. If not, your overall eight is not correct (being stored in EEProm or recognized by the board - my earlier thesis). What's perfect?

1. I'm not sure about your stepper settings - please search and make sure that they're correct. I thought they were 16 on the newer ones. 100% not sure here but lots of posts.
2. I'm not sure you are setting the base ZYZ heights correctly.

Hope that helps. Your statements about "perfect" are confusing...

[m4r1n5]

Mine did not have THAT big of a Radius issue. But I did figure out that through Arduino's "Configuration.h" you need to 1) set EEPROM to 0 2)load 3)set the Printer radius 4) load 5) REset the EEPROM back to 1 6) Load. THEN start your calibration from the beginning, through Repetier. From MY short experience with this, MY PRINTER (not sure if ALL printers have the same outcome) hotend extruder moves DOWN .3mm for every 1mm I enlarge the "#define PRINTER_RADIUS".

Hopefully this will help some.

[Cleveralias]

X_AXIS_STEPS_PER_MM 80 (same for all three axes, which matches EEPROM)

Could be wrong but I believe if you have Rambo 1.0 with 1/8 stepping this value should be halved from the default value, like you've done for your extruder. If you're sure you have a 1.0 board and the newer, 20 tooth pulleys, I'd try 40 steps/mm.

[itilguru]

Sorry I wasn't clear. Here's what happens:
G28 G0 Z0 F3500 goes to the center of the bed and gently pins the paper when I have XAXIS_STEPS_PER_MM defined as 80. If I set that to 40, the hot end only goes halfway down.

G28 G0 Z0 X0 Y90 F3500 would crash into the bed with 80mm set as stated above. What does work perfectly is using G0 Z20.1 X0 Y90 F3500. I've never seen anyone say their edges were that far off. Most people say they're off by a few mm.

I strongly suspect my motors and pulleys are not right. I bought the motors and pulleys early, then the Rostock Max kit later. I'm not sure the pulleys are really 20 teeth and belt pitch is really 2. I have old pulleys but new belts. I use 8 micro steps because my Rambo is 1.0d.

Changing PRINTER_RADIUS has minimal impact.

[Cleveralias]

Okay, gotcha. The older pulleys are 15 teeth rather than 20. Both are GT2. I would first try change that in the firmware and see if it gets you anywhere. Edit: maybe I misread. If these aren't pulleys from SMC I would count the teeth and search out a reference to help you confirm the pitch from the belt. Or just try 15 and see what happens. What do you have to lose.

If that doesn't work (and even if it does, just check to see before changing more variables) I would take a look at the digipot settings for the steppers. 1.0 board put out more current for lower numbers so you may need to adjust the numbers down. Not sure what's good for your board but I think it may be about 120.

[JohnStack]

I strongly suspect my motors and pulleys are not right. I bought the motors and pulleys early, then the Rostock Max kit later. I'm not sure the pulleys are really 20 teeth and belt pitch is really 2. I have old pulleys but new belts. I use 8 micro steps because my Rambo is 1.0d.

Changing PRINTER_RADIUS has minimal impact.

Seriously? How about emailing John at SeeMeCNC and validating what they sent you. This information would have been good to have in the beginning. Otherwise, we're guessing for you.

[itilguru]

Sorry about that JohnStack, the info was buried in my mound of text in the first post. Too much data!

Good news though. Based on the info here I was able to find the original info and discovered my original order included 0.9 degree steppers. At first I only changed STEPS_PER_ROTATION to 400. With that plus my other settings I only got the hot end halfway down at Z0. So I loaded a new copy from scratch, modified only those things I needed, and tried again.

Success! I'm about 0.5mm off at the edges but I know how to solve that. I'll post back tomorrow after I redo the calibration again.

Many thanks to everyone here, and again sorry for the confusion. These things are tricky but great fun.

[itilguru]

Well, that totally worked. Now I think I have the delta arm blues or possibly an extruder problem, so I need help with that.

Summary of my problem:
I bought the electronics kit and servos to build my own printer. That came with RAMB0 1.0d and 0.9 degree stepper motors.
Later on, I bought the Rostock MAX kit without electronics. My kit has the Onyx bed and the black knurled plastic belt guide bearing covers and the white acetal carriage bearing covers. The extruder mounts on the side. It came with a Steve's extruder, which I not using (yet).
I bought an EZstruder when they came out, and I'm using it.

I did the hot end and heated bed calibration, turned on EEPROM, and made those settings. I followed geneb's latest manual and did the calibration chapter using his values for settings.

I can now print! My original problem has been solved, and now I have a nice looking, but spotty calibration cube. I need help understanding if I have the delta arm blues or not enough extrusion speed.

Here are my relevant settings for reference. Anything not mentioned here is default for Reptier-Host version 0.90C:
BELT_PITCH 2
PULLEY_TEETH 20
STEPS_PER_ROTATION 400 // due to my 0.9 steppers - this was the magic bullet
MICRO_STEPS 8 // due to my RAMB0 1.0d controller

EXT0_STEPS_PER_MM 46.2 // I cut this in half from 92.4 because of the microstepping, but since I have 0.9 steppers, maybe I should have left it doubled??? I do have the EZstruder that I bought after the kit.

INVERT_X_DIR true // I had problems early on with bad crimps, but I re-did them. I guess these are just backwards somehow, but they look right. It works this way so I'm not complaining.
INVERT_Y_DIR true // X and Y are inverted, Z is fine.

X_MAX_LENGTH 367.5 // This matches both printer settings value and EEPROM for all three axes.
Y_MAX_LENGTH 367.5
Z_MAX_LENGTH 367.5

MICROSTEP_MODES {8,8,8,8,8} // because I have a RAMB0 1.0d controller
MOTOR_CURRENT {195,195,195,195,0} // I got advice earlier to modify this to 160, so I'll try that today, too.

DELTA_DIAGONAL_ROD 269.0 // this is really close to real-world measurements too

PRINTER_RADIUS 199.5 // this setting flattens my center out and solved my original problem along with using the correct steps per rotation value.

ALLOW_QUADSTEPPING false // I set this to false because of the RAMB0 1.0d controller


Many thanks to all.

Here's my first calibration cube. Measured size is 24.7mm, should be 5mm. I think if I solve the sticky arms issue that may resolve itself, so I'll leave it for last.

First print, probably with delta arm blues though

[JohnStack]

Great!

You're printing almost 5x the size it should be? 24mm instead of 5 mm?

I would keep things simple and avoid a multivariate analysis - I don't think you're to the blues part yet...

The top appears to be starved for filament - so start futzing with your slicing software...

Can't tell anything about the sides...

Once you've gotten a stable print that's filled well, not stringy or too much plastic, then maybe go back to tweaking your eeprom/hardware.

I've gotten extremely clean prints without adding hardware. You can too.

[Cleveralias]

EXT0_STEPS_PER_MM 46.2 // I cut this in half from 92.4 because of the microstepping, but since I have 0.9 steppers, maybe I should have left it doubled??? I do have the EZstruder that I bought after the kit.

Manually extrude 100mm in Repetier and measure how much filament is drawn into the extruder. If 80 steps/mm is correct on the axes then ~92.4 should be correct for the extruder.

[JohnStack]

Cleveralias...good suggestion.... forgot that the whole entire eeprom on this sucker is whack! LOL

Yes, do that. LOL

(Sorry, it's saturday, needed a laugh before the end of the week.)

[itilguru]

Ok, that totally worked. I changed the extruder value back to 92.4 and now I'm printing beautiful cubes and already moving on to new prints!

This thing is really loud. I love it, but now I'm setting up a webcam and moving it to my garage.

[lordbinky]

There's a thread for that

http://forum.seemecnc.com/viewtopic.php?f=54&t=1235