Welcome to the SeeMeCNC Community

RollieRowland wrote:Version 2.0.0 will be up within a couple days.

This version will be the automatic calibration update! I am currently testing it, making sure that the program won't break your printer, although it is working!

This update will require a Z-probe and Repetier firmware, so I will give everyone time to set this up! Since this requires a Z-Probe, you will have to re-flash your firmware.

Firmware Changes:
Configuration.h:
//////////////////////////////////////////////
#define FEATURE_Z_PROBE true
#define Z_PROBE_PIN 10
#define Z_PROBE_PULLUP true
#define Z_PROBE_ON_HIGH true
#define Z_PROBE_X_OFFSET 0
#define Z_PROBE_Y_OFFSET 0
#define Z_PROBE_WAIT_BEFORE_TEST false
#define Z_PROBE_SPEED 2
#define Z_PROBE_XY_SPEED 150
#define Z_PROBE_HEIGHT 39.91
#define Z_PROBE_START_SCRIPT ""
#define Z_PROBE_FINISHED_SCRIPT ""
/////////////////////////////////////////////////////////////

The Z_PROBE_PIN is set the the Z minimum endstop on the Rambo board. You can use any pin, just dig through your pins.h and find the pin you want to map the Z Probe to.

The program needs to have your controller drivers installed for your printer, as it connects to your COM ports. You also need to know your plate diameter, and Baud Rate.

I will update everyone on the progress, but I hope to have this ready in a day or two.

This is great news! It's amazing how fast you're progressing with this.

So to clarify,
1) the Z-probe employed here will be the endstop switch?
2) isn't Repetier firmware the standard you are led to install when you first follow the build instructions...?
3) Plate diameter: this means the build-able diameter the nozzle can reach, not the actual plate diameter, right?
4) Baud Rate: found in configuration.h, right?

I'd really like to try this calibrator, but I have an Orion and no clue how to modify the G code. Can anyone help with that??

HumanLiberty wrote:

RollieRowland wrote:Version 2.0.0 will be up within a couple days.

This version will be the automatic calibration update! I am currently testing it, making sure that the program won't break your printer, although it is working!

This update will require a Z-probe and Repetier firmware, so I will give everyone time to set this up! Since this requires a Z-Probe, you will have to re-flash your firmware.

Firmware Changes:
Configuration.h:
//////////////////////////////////////////////
#define FEATURE_Z_PROBE true
#define Z_PROBE_PIN 10
#define Z_PROBE_PULLUP true
#define Z_PROBE_ON_HIGH true
#define Z_PROBE_X_OFFSET 0
#define Z_PROBE_Y_OFFSET 0
#define Z_PROBE_WAIT_BEFORE_TEST false
#define Z_PROBE_SPEED 2
#define Z_PROBE_XY_SPEED 150
#define Z_PROBE_HEIGHT 39.91
#define Z_PROBE_START_SCRIPT ""
#define Z_PROBE_FINISHED_SCRIPT ""
/////////////////////////////////////////////////////////////

The Z_PROBE_PIN is set the the Z minimum endstop on the Rambo board. You can use any pin, just dig through your pins.h and find the pin you want to map the Z Probe to.

The program needs to have your controller drivers installed for your printer, as it connects to your COM ports. You also need to know your plate diameter, and Baud Rate.

I will update everyone on the progress, but I hope to have this ready in a day or two.

This is great news! It's amazing how fast you're progressing with this.

So to clarify,
1) the Z-probe employed here will be the endstop switch?
2) isn't Repetier firmware the standard you are led to install when you first follow the build instructions...?
3) Plate diameter: this means the build-able diameter the nozzle can reach, not the actual plate diameter, right?
4) Baud Rate: found in configuration.h, right?

1. Yep
2. Also yes, however some decide to install Marlin. This will actually work with any delta running the Repetier firmware, I am just testing with the Rostock Max V2.
3. You will want to use the max build diameter, will need to change that in the program to clarify.
4. You can either find it there, or in your EEProm.

I am hoping to have this tweaked soon, it is calibrating, although it is messing up the horizontal radius and diagonal rod. I am assuming it is the difference in input. As the Z-Probe returns the height that it moves before clicking the switch plus the Z-Probe Height. If I can fix this in the next couple hours I will put up a version for everyone to play with, of course with a disclaimer...

Also, schlrobe, what I did for the auto cal program(to figure out gcode for multiple printers) was this:
-For the values that equal 130, replace them with 0.5 * your build diameter
-For the values that equal 112, replace them with 0.41 * your build diameter
-For the values that equal 63, replace them with 0.23 * your build diameter

Make sure you keep the negated values negative.

Sweet! That worked like a charm...

now I just need to sort out the probe attachment....

Version 2.0.0

Ok, the auto calibration program just calibrated my bed with no issues. This program is (kind of) in an Alpha state, so there are bound to be issues. I.e. keeping your kill switch ready.

For now, the program is set to only calibrate within a tolerance of +/- 0.05 mm, however, my printer finished with a tolerance of +/- 0.02mm.

Several things to take note of:
-You do not need to reset your EEProm. I have tested from defaults and after entering in random but close-ish values.
-The program may seem frozen but it is most likely not, I put the threads into sleep mode when communicating with the printer to prevent code from being lost/skipped. This will prevent user input, so if something goes wrong, then use the power switch on your printer.
-Pleeassee test your Z-Probe before using this program (G31 - hold switch, send/let off switch, send - state should change), the state should read L when you are not pressing the switch, and H when you are pressing it.
-Make sure your Z-Probe height is measured and set correctly. Also, if for some strange reason your Z-Probe height is above 70, do not use this program. PM me and I will increase the g-code height.
-This program is compiled for Windows, it was tested on Windows 8.1 and 10, but should work back to XP as well.
-Disconnect from your printer before closing, else you will receive lag.

Requirements:
-Z-Probe
-Repetier Firmware 0915
-Windows OS
-Delta 3D Printer (duh)
-USB cord connected to printer (why did I type that? you know this)

Even though this works well on my printer, that does not mean there may not be issues with yours. On the first couple passes keep your hand near the kill switch. That is a warning. I will not be liable for any issues this may cause.


Anyway, here is Version 2.0.0 - https://drive.google.com/open?id=0B06WC ... U1JdE9YNDA (I tried to upload to GitHub, but there were issues based on how it is compiled.)
If the link is dead, or the program doesn't operate properly PM or reply.

Awesome! Might as well start the MAC OS fork now, since you know it's coming!

ramai wrote:Awesome! Might as well start the MAC OS fork now, since you know it's coming!

I'll get OSX downloading! I do plan on eventually adding support for Linux and Mac though. I haven't used either very much, so I do not know how difficult that may be...

thanks a lot - that sounds very ggod.
can youplease upload some pics of how you assembled the probe?

thanks

duvdev wrote:thanks a lot - that sounds very ggod.
can youplease upload some pics of how you assembled the probe?

thanks

Yes I can! It is a rig but it works for now.

I gave this a whirl on the smoothieboard. but some of the values don't appear to translate correctly to smoothieboard in regards to the EEPROM/repetier values.


Specifically these listed below; compared to the Alpha Rotation ABC changes recommended, ( I was using the java script from the first page link, not the .exe listed above.)

delta_tower1_offset 0.0
delta_tower2_offset 0.0
delta_tower3_offset 0.0
delta_tower1_angle 0.0
delta_tower2_angle 0.0
delta_tower3_angle 0.0

Thanks.
What about the high difference between the probe and the hot end?

KAS wrote:I gave this a whirl on the smoothieboard. but some of the values don't appear to translate correctly to smoothieboard in regards to the EEPROM/repetier values.


Specifically these listed below; compared to the Alpha Rotation ABC changes recommended, ( I was using the java script from the first page link, not the .exe listed above.)

delta_tower1_offset 0.0
delta_tower2_offset 0.0
delta_tower3_offset 0.0
delta_tower1_angle 0.0
delta_tower2_angle 0.0
delta_tower3_angle 0.0

I do not want to assume, but I am thinking that the delta tower angle 1-3 would be the equivalent of alpha rotation with the base values being stored in the firmware. So you would have to put(also assuming it is set that 1=x, 2=y, 3=z for towers) zeroes in for the alpha rotation abc, as opposed to 210, 330, and 90. I am not sure though, all you can do is try!

duvdev wrote: Thanks.
What about the high difference between the probe and the hot end?

The Z-Probe height? What I did was I measured from the tip of the hot end to the tip of the switch. Mine is around 31.5mm. It doesn't make a difference with calibrating the bed, although if it is set right then your bed should be ready to print on.

Rollie, here's a feature suggestion for you.

In order to discover the exact trigger height of the z-probe, why not write a bit of code that does this:

1. Goes to Z-Zero point on bed.
2. Moves to Z+100
3. Instruct user to attach Z-Probe
4. Move down until switch closes
5. Report height to user in order to update firmware

Sound reasonable?

g.

geneb wrote:Rollie, here's a feature suggestion for you.

In order to discover the exact trigger height of the z-probe, why not write a bit of code that does this:

1. Goes to Z-Zero point on bed.
2. Moves to Z+100
3. Instruct user to attach Z-Probe
4. Move down until switch closes
5. Report height to user in order to update firmware

Sound reasonable?

g.

Fantastic idea! What about having the user manually set the Z-height first? Then the program will start by setting the Z-Probe height to 0, once that is set, it will find the distance moved before the switch is clicked. Subtract that value from 100, then set the difference as the probe height. Should work and will still be automatic!

Either way works. It's your baby.

g.

geneb wrote:Either way works. It's your baby.

g.

Definitely need to implement this though!

Have you seen if anyone has used their hot end itself as a switch? That would be fantastic.

RollieRowland wrote:

geneb wrote:Either way works. It's your baby.

g.

Definitely need to implement this though!

Have you seen if anyone has used their hot end itself as a switch? That would be fantastic.

Yep, I had this crazy idea to airspray down a piece of aluminium foil and use it as a min endstop. ground was connected to the foil and signal was clipped to the hot end. It worked great when I held the foil in the air on the test runs. Something about it being smoothed out over the glass plate ended in destroying a nozzle. not sure what it was, but it didn't work for me.

Here's an idea - alligator clip on the wrench part of the nozzle and another wire screwed/soldered to a feeler gauge. Give 'em a, "Yes, my little bit of metal is under the nozzle" button and you're golden.

g.

Here's another idea - if your license is compatible with/matches the one MatterControl uses...integrate! (or better yet, make it a plugin for MC)

g.

RollieRowland wrote:

KAS wrote:I gave this a whirl on the smoothieboard. but some of the values don't appear to translate correctly to smoothieboard in regards to the EEPROM/repetier values.


Specifically these listed below; compared to the Alpha Rotation ABC changes recommended, ( I was using the java script from the first page link, not the .exe listed above.)

delta_tower1_offset 0.0
delta_tower2_offset 0.0
delta_tower3_offset 0.0
delta_tower1_angle 0.0
delta_tower2_angle 0.0
delta_tower3_angle 0.0

I do not want to assume, but I am thinking that the delta tower angle 1-3 would be the equivalent of alpha rotation with the base values being stored in the firmware. So you would have to put(also assuming it is set that 1=x, 2=y, 3=z for towers) zeroes in for the alpha rotation abc, as opposed to 210, 330, and 90. I am not sure though, all you can do is try!
.

The Alpha Rotation - ABC changes after running the calibration are:
A 210.003
B 330.005
C 89.998

That's probably not enough to make a difference, but I'll try to figure out the conversion to smoothie.

KAS wrote: Yep, I had this crazy idea to airspray down a piece of aluminium foil and use it as a min endstop. ground was connected to the foil and signal was clipped to the hot end. It worked great when I held the foil in the air on the test runs. Something about it being smoothed out over the glass plate ended in destroying a nozzle. not sure what it was, but it didn't work for me.

I think working the nozzle itself would be the best way for accuracy, IF you can get it to work properly/consistently.

geneb wrote: Here's an idea - alligator clip on the wrench part of the nozzle and another wire screwed/soldered to a feeler gauge. Give 'em a, "Yes, my little bit of metal is under the nozzle" button and you're golden.
g.

Hmmm, I am going to try this. I want to try at least 3 or 4 methods and see which works the best! My current switch moves too much (as expected).

geneb wrote: Here's another idea - if your license is compatible with/matches the one MatterControl uses...integrate! (or better yet, make it a plugin for MC)
g.

Not sure exactly how integrating would work but I can try doing a plugin!

KAS wrote: The Alpha Rotation - ABC changes after running the calibration are:
A 210.003
B 330.005
C 89.998

That's probably not enough to make a difference, but I'll try to figure out the conversion to smoothie.

Oh those values are nothing to be concerned about! Still an assumption but I would use 0.003, 0.005, and -0.002, if you do feel it's necessary.

I've been using FSRs now for 6+ months on my Rostock, TAZ, Kossel and big Kossel and they work great and are reproducible. No offset required - touch and go. They are not that expensive either and very easy to mount and once there, no muss or fuss. You might check them out. I have an FSR mount STL in my thread that works great.

How about using an FSR to play "chase the hot end" in the calibration process? You just move it to be under the nozzle each time, rinse, repeat. AFAIK, it needs some kind of little adapter board to change the resistance to an on/off signal.

g.

I actually did exactly that! Same principle as a strip (.001") of metal with an alligator clip to complete an electrical circuit with the nozzle. However, the issue I discovered is that the FSR is fairly thick and you need to compensate for that. But, when it is under the glass plate, you are touching off directly against the build surface so that does not matter.

Well then.. Those would be incredible! I'm going to order a couple, then rig up a board which will convert the resistance into a switch, like geneb was saying!