Everything is out of scale

[PatriotSpade]

New Rostock Max V2. Prints great except for the fact that all axis are out of scale.

40mm calibration block
X - 41.65
Y - 41.65
Z - 46.00

As you can see, it's larger on all axis, but much larger on the Z than the others.

I can't seem to figure out the issue. Scale is set to 1.0 on both MatterControl and Repetier Host with the same results.

Does anybody have a clue what would cause this? It seems like there's some firmware setting that is off.

[Eaglezsoar]

See this link for a possible fix: http://forum.seemecnc.com/viewtopic.php?f=77&t=9261" onclick="window.open(this.href);return false;

[PatriotSpade]

See this link for a possible fix: http://forum.seemecnc.com/viewtopic.php?f=77&t=9261" onclick="window.open(this.href);return false;

I have verified that I do have the ball-cup arms and my diagonal is set to 290.8 mm.

[Polygonhell]

If z scaling is incorrect, you have to have the steps per mm set incorrectly for the axis.

[PatriotSpade]

Here is my EEPROM settings. I am just realizing that the Steps per mm (92.65) is the same as the Extr. 1 steps per mm (92.65). I don't remember changing that value, but I did change the Extr. 1 steps per mm per the user guide, section 2.6. Should the Steps per mm setting be different than the Extr. 1 steps per mm? Also, I don't see a Z-Axis steps per mm, so I assume that's the setting you believe is wrong?

[img]http://i.imgur.com/nSLOqXa.png[/img]

[guanu]

steps per mm should be 80

[geneb]

The steps per mm figure shown is the value I'd expect to see for the extruder and is indeed set incorrectly.

I DO know that the printer hasn't been calibrated. The Horizontal Radius value is VERY wrong for a ball/socket & injection molded carriage configuration.

What software are you using?

g.

[PatriotSpade]

steps per mm should be 80

I changed Steps per mm to 80.000 and attempted a quick calibration cube (20 x 20 x 10).

Results are:
X & Y: 18.50mm (Expected 20mm)
Z: 10.50mm (Expected 10mm)

These values are MUCH better than what I was getting. I think this may have fixed my major issue. Obviously, still not perfect. So how would I go about achieving perfection? Or is this as close to perfection as is expected?

The steps per mm figure shown is the value I'd expect to see for the extruder and is indeed set incorrectly.

I DO know that the printer hasn't been calibrated. The Horizontal Radius value is VERY wrong for a ball/socket & injection molded carriage configuration.

What software are you using?

g.

I did the calibration using MatterControl and the "Tower Cal" script provided by SeeMeCNC in the user guide, section 2.5. The EEPROM settings shown were taken using Repetier-Host as I'm more familiar with it.

[Polygonhell]

You need to make sure the Horizontal radius is correctly calibrated, and as Gene mentioned that's not the right value for a Max with newer arms.
You should pretty easily get within a couple of tenths of a mm when the calibration is correct.

[PatriotSpade]

You need to make sure the Horizontal radius is correctly calibrated, and as Gene mentioned that's not the right value for a Max with newer arms.
You should pretty easily get within a couple of tenths of a mm when the calibration is correct.

Are there any additional instructions other than what was included in the user guide? I started with Horizontal radius of 144 and the X, Y, Z towers were all off by a decent amount.

[mhackney]

You can use this to calculate steps/mm http://prusaprinters.org/calculator/" onclick="window.open(this.href);return false;

I believe the V2s have 20 tooth pulleys. When I plug in the numbers I determine that your steps/mm should be 80. Then, just to make sure that's right, look at this:

you had it set to 92.65 so your scaling factor in Z would be:

92.65/80 * 40mm (the size the cube should be) = 46.3mm - which happens to be the height you got in the our print.

Z height is affected with the steps per mm. Realize that on a delta there is one step/mm parameter for movement and that applies to all 3 towers. Which, unfortunately, are called X, Y and Z by most here. This is confusing because our X, Y and Z towers ARE NOT the Cartesian X, Y Z directions. The Kossel crowd name the towers alpha, beta and gamma - a lot less confusing!

So that takes care of your height problem. Now to the X-Y scaling problem X-Y scaling is adjust by the diagonal arm lengths. We can calculate what your arm length is from your data.

actual diagonal length = 290.8 (current length in EEPROM) * 41.65mm / 40.00 mm = 302.80

Note that longer arm length decreases the X-Y scaling of the print.

[PatriotSpade]

You can use this to calculate steps/mm http://prusaprinters.org/calculator/" onclick="window.open(this.href);return false;

I believe the V2s have 20 tooth pulleys. When I plug in the numbers I determine that your steps/mm should be 80. Then, just to make sure that's right, look at this:

you had it set to 92.65 so your scaling factor in Z would be:

92.65/80 * 40mm (the size the cube should be) = 46.3mm - which happens to be the height you got in the our print.

Z height is affected with the steps per mm. Realize that on a delta there is one step/mm parameter for movement and that applies to all 3 towers. Which, unfortunately, are called X, Y and Z by most here. This is confusing because our X, Y and Z towers ARE NOT the Cartesian X, Y Z directions. The Kossel crowd name the towers alpha, beta and gamma - a lot less confusing!

So that takes care of your height problem. Now to the X-Y scaling problem X-Y scaling is adjust by the diagonal arm lengths. We can calculate what your arm length is from your data.

actual diagonal length = 290.8 (current length i EEPROM) * 41.65mm / 40.00 mm = 302.80

Note that longer arm length decreases the X-Y scaling of the print.

Now THIS sounds promising. A lot of things to take into consideration and lots of little changes to make. I will report back promptly.

PS: Thank you very much for explaining the effect that these parameters actually cause. I hail from the Cartesian side of things and tend to confuse myself between the two.

[PatriotSpade]

SUCCESS!

Thank you everyone for your assistance! I'm leaving a list of variables I changed here for future reference:

• I forgot I had changed the PULLEY_TEETH variable in Configuration.h of the firmware to 15. I changed that back to 20.
• I changed my steps per mm EEPROM variable to 80.
• I changed my horizontal radius back to 144.
• Ran tower calibration again

After changing these variables, I printed the 20mm calibration cube (20 x 20 x 10) again and got the following results:
X & Y: 20.05
Z: 10.10

Moral of the story: Don't go chasing rabbits to the point that you have multiple variables changed just trying to "hit the lottery." Take your time and change one variable at a time with extensive testing/calibration between changes.

Thanks again everyone for the help! Now... time to finally print those fan shrouds (5th attempt!)

[mhackney]

Great! As for your comment "Moral of the story: Don't go chasing rabbits to the point that you have multiple variables changed just trying to "hit the lottery." Take your time and change one variable at a time with extensive testing/calibration between changes."

#8 here: http://forum.seemecnc.com/viewtopic.php?f=36&t=7361" onclick="window.open(this.href);return false;

[nitewatchman]

SUCCESS!



Moral of the story: Don't go chasing rabbits to the point that you have multiple variables changed just trying to "hit the lottery." Take your time and change one variable at a time with extensive testing/calibration between changes.

Thanks again everyone for the help! Now... time to finally print those fan shrouds (5th attempt!)

Learned this over and over again "adjusting" multiple carburetor setups on "fast" cars. One variable at a time will simplify your life greatly.

[Eaglezsoar]

I am glad that you got success!
Happy Printing!

[drandolph]

I'm having a similar issue and I think it's because I bought my machine at MakerfareNYC from SeeMeCNC but just after that they changed out the arms to the ball ones. So in the build instructions they just tell you to set the Diagonal rod length to 290.8 which threw me off for awhile but with this post I changed it back to 269 and it's closer but still way off.

When I had the diagonal rod length set to 290.8 My prints were smaller than expected at a ratio of about .91
When I set it to 269 it's a ratio of 1.10

I check my configuration.h and my teeth are at 20
My steps are at 80
My horizontal radius is at 146.6 (I thought the horizontal radius was just to deal with the cupping of the head as the arms move which is how i ended up adjusting it up to 146.6)
I'm using the new carriage assembly but the arms are the injection molded squares with pegs.
My one exception to a standard config is that I'm running a Volcano E3Dv6 hotend which I didn't think would affect much other than z-height and some PID settings which I have configured and are printing good layers and temps.

So my questions are.

Did I miss a section on calibration somewhere?
Does having a different hotend change the scale on a delta printer? (Not the way on any of my gantry style machines)
Are there any other setting differences with the new ball arms vs the older square style arms?

[Eaglezsoar]

I'm having a similar issue and I think it's because I bought my machine at MakerfareNYC from SeeMeCNC but just after that they changed out the arms to the ball ones. So in the build instructions they just tell you to set the Diagonal rod length to 290.8 which threw me off for awhile but with this post I changed it back to 269 and it's closer but still way off.

When I had the diagonal rod length set to 290.8 My prints were smaller than expected at a ratio of about .91
When I set it to 269 it's a ratio of 1.10

I check my configuration.h and my teeth are at 20
My steps are at 80
My horizontal radius is at 146.6 (I thought the horizontal radius was just to deal with the cupping of the head as the arms move which is how i ended up adjusting it up to 146.6)
I'm using the new carriage assembly but the arms are the injection molded squares with pegs.
My one exception to a standard config is that I'm running a Volcano E3Dv6 hotend which I didn't think would affect much other than z-height and some PID settings which I have configured and are printing good layers and temps.

So my questions are.

Did I miss a section on calibration somewhere?
Does having a different hotend change the scale on a delta printer? (Not the way on any of my gantry style machines)
Are there any other setting differences with the new ball arms vs the older square style arms?

No, there are no other arm settings other than the length which you already have correct.
What are the steps that you have for the filament extruder and have you calibrated it?