OpenDACT(Delta Automatic Calibration Tool) - For Repetier

[RollieRowland]

I haven't had a chance to test on Windows 7, although I did try a couple things that may help: https://drive.google.com/file/d/0B06WCA ... sp=sharing

Also, this version is not ripped from my debugging folder like the previous ones. If this doesn't work then I would recommend installing .Net Framework 4.5.2. Which is a requirement for the app

[mhackney]

Installing now, note that the installer automatically detected and installed .Net FW 4.5.2 on my Windows 7 box.

However, I still run into the same problem when running it.

[duvdev]

Hi
About the fat kit. I will order it now. It there any spacer or any plastic part that I need to print?
Do I have to configure the board or just to connect it to the rambo?

Thanks

[mhackney]

duvdev, PLEASE follow the link I posted in this thread a day ago. There is lots of information about mounting FSRs on the web - the google delta group has the majority of it since that is where Johann posted his initial FSR design. But, several of us post here too so you can simply search and learn everything you need.

I strongly suggest that you do some reading - I even posted links to the details on the board that describes how to configure and use it - click it!

[Pyropainter]

Good Morning

I have just finished using your manual version for calibrating my Rostock Max v2 and am very happy with the results, even though I have not yet tried to print anything. Thank you for investing your time in something that is sure help many others. I have been thru this thread and not found an answer to the following concern and questions. If I have overlooked it. Please send me in the right direction. While performing the calibration I remembered reading that one way to improve printed part dimension inaccuracies is to change the diagonal rod length and noticed that changing it is required when using your manual calibration program. I do have a couple of questions if you do not mind. First, does your program some how adjust the other parameters to compensate for the changes made to the diagonal rod parameter that will correct the inaccuracies? Second, If not how do I correct them. I assume that the only option would be scaling the stl in a slicer.

Thanks again.

[RollieRowland]

Good Morning

I have just finished using your manual version for calibrating my Rostock Max v2 and am very happy with the results, even though I have not yet tried to print anything. Thank you for investing your time in something that is sure help many others. I have been thru this thread and not found an answer to the following concern and questions. If I have overlooked it. Please send me in the right direction. While performing the calibration I remembered reading that one way to improve printed part dimension inaccuracies is to change the diagonal rod length and noticed that changing it is required when using your manual calibration program. I do have a couple of questions if you do not mind. First, does your program some how adjust the other parameters to compensate for the changes made to the diagonal rod parameter that will correct the inaccuracies? Second, If not how do I correct them. I assume that the only option would be scaling the stl in a slicer.

Thanks again.

At the moment you may get inaccuracies in the XY dimensions IF your printers frame is not square. I do not have a fix for this currently, however, I am looking into using inductive proximity sensors to correctly find the delta radii offsets (move towards tower until the sensor triggers). This would still not entirely fix the issue, that relies entirely on your frame. If the diagonal rod is changed more than 0.3 +/-, I would consider measuring your rods or rebuilding your frame. Otherwise this will not cause a significant change in the XY dimensions.

Also, sorry mhackney, I will try to find a fix soon but I haven't had time.

[mhackney]

No worries, I'll keep my eye out for it. If there is anything I can do to help, let me know. As you probably may know, I'm a software guy by profession and know most languages and platforms well enough to be dangerous.

[crocky]

FSR wiki has useful information too, http://reprap.org/wiki/FSR on how to hookup...

[duvdev]

So just to make it clear the app will be able to work with z probe ? Because I don't have a of fsr now

[Jrjones]

So far I'm up and running after using Rollie's v1.03 of the HTML calibration calculator, and my largest deviation is 0.12mm (of 169 measured points, 8 points are 0.10mm or greater.)

I haven't noticed any discrepancies in X and Y or Z measurements, I scale my parts by 1.6% to account for ABS shrinkage in the X and Y directions.
I'm happy I can relax a bit about watching the first layer to make sure it adheres properly.

I may run it again with my second dial indicator to see if that improves it at all (my easy to read digital indicator is +/- 0.0254mm, my old analog one is +/- 0.0025 I believe).

But for now I am leaving it as is because I can successfully print larger objects than 60mm!

[RollieRowland]

So just to make it clear the app will be able to work with z probe ? Because I don't have a of fsr now

Yes, this should work with both.

So far I'm up and running after using Rollie's v1.03 of the HTML calibration calculator, and my largest deviation is 0.12mm (of 169 measured points, 8 points are 0.10mm or greater.)

I haven't noticed any discrepancies in X and Y or Z measurements, I scale my parts by 1.6% to account for ABS shrinkage in the X and Y directions.
I'm happy I can relax a bit about watching the first layer to make sure it adheres properly.

I may run it again with my second dial indicator to see if that improves it at all (my easy to read digital indicator is +/- 0.0254mm, my old analog one is +/- 0.0025 I believe).

But for now I am leaving it as is because I can successfully print larger objects than 60mm!

On your last calibration, what was your measurements of the 6 outer points? Also, how many iterations did you run?

Glad it has helped!

[DGBK]

I wanted to update that after getting a longer bowden tube, I have had pretty good success using v1.03 with manual calibration (adjusting the z height with the lcd and using paper). My calibration is about as good as I will get it manually. I've been getting much better prints since implementing your leveling aid. I have some FSRs on order, so I will be able to get a more accurate calibration and hopefully take advantage of the automatic version of the program.

Thanks a lot for your work.

[RollieRowland]

I wanted to update that after getting a longer bowden tube, I have had pretty good success using v1.03 with manual calibration (adjusting the z height with the lcd and using paper). My calibration is about as good as I will get it manually. I've been getting much better prints since implementing your leveling aid. I have some FSRs on order, so I will be able to get a more accurate calibration and hopefully take advantage of the automatic version of the program.

Thanks a lot for your work.

Yeah, I just installed my FSRs around 4am yesterday. I do need to drop the z-probing speed to improve accuracy (added to the list for 2.0.1). With the current speed - which I think is 60mm/s - it only gives an accuracy of +/-0.075 in three iterations. I am hoping to get a lot done on this in the next day or two.

You're welcome!

[crocky]

Still waiting for mine to arrive, they are in Melbourne according to tracking... So maybe tomorrow, or tuesday.. I am designing new FSR pads to go into the holes where the bolts go, line up should be a breeze.

[mhackney]

I just dug out a Windows 8 laptop that I'd been using on my Droplit printer to try this calibration code on. As soon as this print is complete, I'll give it a shot.

[mhackney]

No glory, this was a windows 7 laptop and it failed exactly the same as windows 7 VM on my Mac. I don't have windows 8.

[mhackney]

Ok, it turns out my son has a laptop with Win 8 that he got for college last year. So I can now give it a shot. However, I have FSRs so the homing speed may be an issue.

[eigenbot]

This was super helpful in leveling my first layer for large surfaces - thanks!

I've read through most of this string, and it seems like you're going more towards an automated system than the g code direct entering, but since I'm running with pretty basic tools I have a couple comments about running it manually with the .html file in your zip. Sorry if I missed something earlier in the thread!

I'm using the new carriages, and I think increasing the horizontal radius to 140 makes the lest position of x0 y-130 outside of the print boundaries. I ended up doing mine manually because I could get the paper test to be more consistent than trying to rig up the callipers so I just changed my G code for that line. I ended up changing to a radius of 126 instead of 130 because my z-axis bottomed out at z=0 for anything larger at the opposite z tower - but the algorithm still really improved my large first layer printing.

Also, the G code maths? If the radius is 130, the y corners should be at +\- 65 rather than 63.

Again, thanks for such a handy tool and I super appreciated the diagrams about how each dimension affects the towers!

[RollieRowland]

Ok, it turns out my son has a laptop with Win 8 that he got for college last year. So I can now give it a shot. However, I have FSRs so the homing speed may be an issue.

I am getting close to having 2.0.1 finished, the probing speed will be adjustable but will be initially set at a lower speed so all z-min endstops work properly/accurately.

This was super helpful in leveling my first layer for large surfaces - thanks!

I've read through most of this string, and it seems like you're going more towards an automated system than the g code direct entering, but since I'm running with pretty basic tools I have a couple comments about running it manually with the .html file in your zip. Sorry if I missed something earlier in the thread!

I'm using the new carriages, and I think increasing the horizontal radius to 140 makes the lest position of x0 y-130 outside of the print boundaries. I ended up doing mine manually because I could get the paper test to be more consistent than trying to rig up the callipers so I just changed my G code for that line. I ended up changing to a radius of 126 instead of 130 because my z-axis bottomed out at z=0 for anything larger at the opposite z tower - but the algorithm still really improved my large first layer printing.

Also, the G code maths? If the radius is 130, the y corners should be at +\- 65 rather than 63.

Again, thanks for such a handy tool and I super appreciated the diagrams about how each dimension affects the towers!

Thank you for pointing out the error in the gcode that I was using, I made the correction in the original post. I'm glad that it's helped you so far!

[geneb]

Were you able to add a "let's figure out how tall the z probe is" feature?

g.

[RollieRowland]

Were you able to add a "let's figure out how tall the z probe is" feature?

g.

Yes, I am still implementing it though. One thing to take note of is that it does require the z-height to be set first.

[geneb]

Ok, cool. The Z height set is the simple part and something most people should have a handle on how to do already. I'm just thinking it would be cool to have a "redneck" z-probe that you can just slap on and go.

g.

[bvandiepenbos]

Were you able to add a "let's figure out how tall the z probe is" feature?

g.

Yes, I am still implementing it though. One thing to take note of is that it does require the z-height to be set first.

Set z height MANUALLY to run "auto" calibration ??????
Now that is just silly.

[geneb]

It makes perfect sense. You have to know what the machine Z height is if you're going to try to discover how tall the z-probe mechanism is.

g.

[RollieRowland]

Were you able to add a "let's figure out how tall the z probe is" feature?

g.

Yes, I am still implementing it though. One thing to take note of is that it does require the z-height to be set first.

Set z height MANUALLY to run "auto" calibration ??????
Now that is just silly.

Yeah, I can see why you think this way. The program will only truly be an automatic calibration with FSRs installed. Otherwise it is a semi-automatic calibration. This wouldn't be necessary if the z-minimum probe was the same exact height as the nozzle.