Calibration

Support for UP Software. To report a bug, post with a title [BUG REPORT]. To request a feature, post with a title [FEATURE REQUEST]
roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Calibration

Post by roller » Sat Dec 08, 2012 2:48 am

I printed the calibration model to see if my printer was all in check before doing some pieces for a friend who needs dimensionally accurate parts. I found on printing the model X1 and X2 were a couple of mm different - all other dimension were fine. I entered the values I measured with digital calipers and reprinted - the print is now consistent across diagonals but oversized i.e. 125 x 125 not 124 x 124. So I put in these new values and reprinted - same result, 125 x 125.

So is this normal - is the perimeter about 0.5mm oversized - or am I using this calibrate feature incorrectly.

I'm using Up Software version 1.3 (Mac)

marto
Posts: 87
Joined: Wed Nov 14, 2012 2:01 pm

Re: Calibration

Post by marto » Sun Dec 09, 2012 8:59 am

It would be nice to know if this worked? As its treating the calibration print just like a normal print I would assume if the calibration was working it would change size.

Steve

LonV
Posts: 220
Joined: Thu Nov 15, 2012 11:53 pm
Location: Phoenix, Arizona, USA

Re: Calibration

Post by LonV » Tue Dec 11, 2012 4:13 am

roller, did you "Reset" before inputting the values? If not, it's adding the new values to the old values.

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Tue Dec 11, 2012 4:23 am

I didn't hit reset because I thought I would want to add to the old values i.e. I want to accumulate the first correct (to correct skew) and second correction to correct size.

I find it a bit strange that the first calibration produces an oversize print and need to be corrected again... and stranger putting in the second calibration which supposedly "adds" to the first makes no difference at all.

marto
Posts: 87
Joined: Wed Nov 14, 2012 2:01 pm

Re: Calibration

Post by marto » Tue Dec 11, 2012 9:23 am

Ok so I assume that we should reset. Then print then calibration. Input new values. Apply. Then reprint to test makes sense.

I assume the calibration procedure assumes that you have reset your printer before printing the calibration part so its always at the stock factory settings.

Otherwise it would be a process of scaling up and down repeatedly until you got it right. And if you wanted to update the firmware you would then need to repeat exactly the same sequence of values in order to restore your configuration. This might explain why your correction is having no effect.

BTW I have no idea if this is what its doing this is just me looking for a logical explanation to the situation.

Steve

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Tue Dec 11, 2012 10:00 am

My printer is brand new so presumably I don't need to reset first. What I did was:
1.Printed calibration model (print skewed)
2. Entered measured X1 and X2 into calibration values
3. Printed again - print oversized (print is square and symmetrical but overall is more out than before)
4. Entered Calibration values
5. Printed again - exactly the same as 3.
6. Entered in values again
7. Printed again ... same as 3 & 5.

So it doesn't seem they add up at all. Entering new values of 125 and 125 (oversize by 1mm in X1 and X2) does nothing.

marto
Posts: 87
Joined: Wed Nov 14, 2012 2:01 pm

Re: Calibration

Post by marto » Tue Dec 11, 2012 10:06 am

Yep it sounds like what I expected.

From 3. Onwards would always result in the same part. If it works how I said.

So the printer assumes you reset it. Then when you print you put in the values and it adjusts the print. From 3 onwards you haven't reset so the software assumes you are just retrying to enter in the same calibration. Its going to make the same correction.

I know your machine is brand new. But who is to say they don't apply an approximate initial calibration to all minis?

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Wed Dec 12, 2012 12:44 am

I would think they don't because it would mean they need to test print from each - I suspect that's a mammoth waste of time. I find it hard to believe I need to reset to add a variation but I'll try it and see. Seems nonsensical to me.

marto
Posts: 87
Joined: Wed Nov 14, 2012 2:01 pm

Re: Calibration

Post by marto » Wed Dec 12, 2012 1:53 am

I agree there is no way they would test every printer. But if you tested say 5-10 from the initial batch and the values were different to the theoretical movement (steps * belt pitch or whatever) I would think you would apply the average correction. But as you said they may not bother, or you might just get the stock calibration from the printer they were developing the ROM on.

I dont think you are adding any variation by resetting. You are in fact removing it.

You are resetting the calibration to stock so that it prints the model with its known factory settings. You then measure it and it applies this correction the the factory settings.

I think the term "added" to the previous values in the manual is incorrect. I think it is in fact never added to the previous settings but is always applied to the factory settings, which is why you need to reset before you print. What it is trying to say is that it if the print is done with the previous calibration the measurements will not be correct as there will already be some adjustment to the factory calibration.

Why does doing it like this make sense?
Well lets say I do a reset. Then do calibration (1), its not quite right so I do another (2) and another (3). Lets say I recalibrate again and find that (4) is slightly worse than (3). If it was added to the previous values the only way to get back to (3) would be to reset and enter calibration values (1) then (2) then (3). Which in theory would get me back to the same point.

If it is not added together we would simply enter (3) and be back to where I was.

Lets just say my calibration gets cleared for whatever reason. Maybe I upgraded my ROM or had to swap out my CPU. Similarly I would only need to enter the most recent/best calibration, rather than a sequence.

Also what would happen if I accidentally double clicked the set button? Hmm now I need to start my sequence all over again.

What doesn't make sense is the fact that you need to go into the menu to reset it before hand. Ideally the calibration mode should reset your printer to stock, print the model and then allow you to enter the current sizes and adjust for this. As most people won't ever bother to do this calibration on their Up! I suspect such an integrated calibration mode is not high on their priorities list.

I will have to test these theories tonight.
But another potential explanation is that the calibration mode simply doesn't work at all. In which case entering any value would make no difference.

Steve

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Wed Dec 12, 2012 5:00 am

I'd really like someone official to chime in and explain how it is supposed to work.

The thing is after you put in the correction you shouldnt need to do it again as it's simple maths to adjust it so it should be right. The fact is it isn't and every time I reset I am going to have to do 2 prints just to verify. I suspect it is just an overprint/overfill issue and not a calibration issue ... maybe it's even just a software version issue.

MrMaker
Posts: 5
Joined: Sat Jan 05, 2013 10:56 pm
Location: UK

Re: Calibration

Post by MrMaker » Sat Jan 05, 2013 11:22 pm

I too am having problems with the vertical calibration feature.

I have had my printer for nearly a year but always printed small, flat parts so never bothered with the vertical calibration as I did not notice any significant leaning. However I have had problems printing some taller parts recently, the vertical bits tended to lean a little to the rear of the printer. I decided to follow the calibration process in the new manual. I printed the calibration model, took the measurements, entered the data into the calibration window (after pressing reset) and then reprinted the calibration model to check - expecting it to be near vertical - No! This process has taken a lot of time and made not a jot of difference, the errors are still there - the measurements are identical to the measurements taken before the calibration. Does the software send the calibration to the printer or use it locally to compensate for the errors when generating the gcode (or whatever Delta use to control the printer)? I have spent all day and all evening trying to resolve this but got nowhere, this is what I have tried so far...

- Calibration using v1.16, v1.17 and v1.18 but all versions behave the same. Always clearing previous calibration results to zero the readings in the calibration window title before printing the calibration model.
- Completely uninstalled all versions of UP, deleted all UP related entries in the Windows registry, reinstalled v1.18 and entered the calibration measurements again (after reset)

The measurements I took from the calibration print are always consistent (I have printed this damn thing 6 times today); X1= 125.9 X2 = 124.5 Z = 1.6 H = 40.4 and the software calibration results shown in the title of the calibration window after storing the values are XY 0.64Deg / XZ 2.27Deg.

My printer is an UP! Plus with Firmware v3.00 and ROM v6.041. I am using Windows 7 and have tried versions 1.16, 1.17 and 1.18 of the UP software but none seem to complete the calibration properly.

I have scoured the forum for guidance but other than this thread there seems to be little mention of the vertical calibration or any problems (other than misunderstanding the need to press reset to clear previous calibrations) so I guess I must be doing something wrong but I can't think what?

Thanks in advance,

smarthall
Posts: 2
Joined: Thu Dec 20, 2012 1:57 am

Re: Calibration

Post by smarthall » Mon Jan 07, 2013 11:12 am

Did you try , clearing the calibration data before you print the test print? I think you need to do that so the calibration print is done with the right values.

Sent from my GT-N8000 using Tapatalk 2

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Mon Jan 07, 2013 10:47 pm

I did. There seemed to be no way to get my calibration model to come out correctly. I gave up on having large parts fit together accurately and print them on my reprap instead.

MrMaker
Posts: 5
Joined: Sat Jan 05, 2013 10:56 pm
Location: UK

Re: Calibration

Post by MrMaker » Tue Jan 08, 2013 12:14 am

smarthall wrote:Did you try , clearing the calibration data before you print the test print? I think you need to do that so the calibration print is done with the right values.

Sent from my GT-N8000 using Tapatalk 2
Yes, I reset the calibration each time before printing the calibration model.

I eventually found the cure to this problem :-

1. Uninstall all versions of UP software from the PC
2. Delete any folders left behind by the uninstall
3. Using REGEDIT delete all Windows registry entries that relate to UP software
4. Reinstall UP software - I installed the latest version 1.18
5. Printed the calibrate96 model
6. Entered the measured calibration measurements
7. OK,ed the perform vertical calibration message and reprinted the calibrate96 model
8. Miracle... it now is perfectly vertical and equally sized

I have no idea what was going wrong here, each time before the calibration results in the calibrate window were zero before I started, I pressed the reset for good measure and after I entered the measurements in the window and pressed OK the title bar showed a value for XY and XZ in deg. But the printer did not seem to take any notice! I think there might be an obscure bug in the calibration feature?

After all this, the printer is down yet again - the extruder motor is overheating I think, it stopped working all together last night and today it simply melts the filament in the gear after about 30 minutes operation. I think it is time to buy an Ultimaker and stick this printer on Ebay!

User avatar
wackojacko
Posts: 667
Joined: Fri Nov 12, 2010 5:46 am
Contact:

Re: Calibration

Post by wackojacko » Tue Jan 08, 2013 1:30 am

You probably need to check if the drive gear is clean and doesn't have any plastic residue on it. That's the main cause of air printing.
Bruce
http://www.3DPrintingSystems.com

Need help with your printer, check out our Youtube channel.
https://www.youtube.com/user/3dprinting ... /playlists

User avatar
DrewPetitclerc
Posts: 581
Joined: Sat Feb 12, 2011 7:39 am
Location: Silicon Valley California
Contact:

Re: Calibration

Post by DrewPetitclerc » Tue Jan 08, 2013 5:31 pm

After all this, the printer is down yet again - the extruder motor is overheating I think, it stopped working all together last night and today it simply melts the filament in the gear after about 30 minutes operation. I think it is time to buy an Ultimaker and stick this printer on Ebay!
I think your new problem is probably due to using off OEM filament and the higher temp the motor is reaching causing you feed problems, here are the best fixes for that.
http://pp3dp.com/forum/viewtopic.php?f= ... 9270#p5473
http://pp3dp.com/forum/viewtopic.php?f= ... 9270#p5474
Regards
Drew
Drew Petitclerc
Petitclerc Designs
Owner/Senior Principal Designer, prototype, tooling and test equipment design and 3D printing
http://flash-graphics.deviantart.com/
http://www.thingiverse.com/DrewPetitclerc

MrMaker
Posts: 5
Joined: Sat Jan 05, 2013 10:56 pm
Location: UK

Re: Calibration

Post by MrMaker » Wed Jan 09, 2013 12:53 am

DrewPetitclerc wrote:
I think your new problem is probably due to using off OEM filament and the higher temp the motor is reaching causing you feed problems, here are the best fixes for that.
http://pp3dp.com/forum/viewtopic.php?f= ... 9270#p5473
http://pp3dp.com/forum/viewtopic.php?f= ... 9270#p5474
Regards
Drew
Thanks Drew,

I have ready your upgrade posts when I first got this printer last year. Excellent work, those are really nicely designed parts. I had planned to follow your guidance but have had limited time and was set back when my main board blew up last year! I have already sorted out the leveling of the print bed and made a new platform out of MIC6 aluminum - the original bed was warped all over the place. I am pivoting on a block of high temp rubber disks and the platform is perfect. I use glass with Kapton tape like you and always take the glass off to remove the prints. I have a small CNC mill at work and plan to try to make your extruder tensioner parts and the improved extruder support, in the meantime I have modified my extruder today using the simple 'pen clip' tensioner, I have also remounted the heater block heat sink to the stepper motor using fibre washers to try and minimise the heat transferred to the motor - the bloody thing generates enough of it own heat! I think my problems began when I got a new main board, the newer version and the extruder motor runs very much hotter than with the original board. It gets so hot that it softens the plastic in the gear and stops extruding, when it stopped last I dismantled the extruder without withdrawing the filament to see what had happened and the filament had been 'waisted' between the gear and roller. Tonight I reassembled the machine with the 'pen clip' extruder (I had to make do closely similar pen clip) and managed to print several large prints of 50 minutes to 1 hour duration without any problems at all - hopefully I have cured the problem.

These printers are very good but there are several flaws in the design that Delta do not seem terribly interested in fixing - sad because it could be the best printer on the market if they learned from the experiences of the users and implemented some of the improvements. It wouldn't really cost them much in the scheme of things I don't think, I am surprised at the cheap extruder bracket - that really is quite poor and the print head 'bobs' all over the place - just making it from thicker material would have helped!

Have you had any problems with the vertical calibration feature - I had to go through a very long winded process to get that one to work (see my posts above). What version of the UP software do you currently use?

I very much appreciate your comments and admire the work you have done with this printer.

Regards,

Lee

User avatar
wackojacko
Posts: 667
Joined: Fri Nov 12, 2010 5:46 am
Contact:

Re: Calibration

Post by wackojacko » Wed Jan 09, 2013 3:01 am

The UP Plus from about 3-4 months ago now have heavier duty extruder mounts.
Bruce
http://www.3DPrintingSystems.com

Need help with your printer, check out our Youtube channel.
https://www.youtube.com/user/3dprinting ... /playlists

Tiertime-Edward
Posts: 206
Joined: Tue Sep 07, 2010 9:15 am

Re: Calibration

Post by Tiertime-Edward » Thu Jan 10, 2013 3:04 am

Hi Roller,

Your printer vertical is perfect. X1 and X2 are both 125, that's means your calibration model is 100% vertical. When doing the vertical calibration, it is to make sure X1 and X2 are same.

roller
Posts: 1622
Joined: Sat Oct 06, 2012 9:09 pm
Location: Brisbane, Australia

Re: Calibration

Post by roller » Thu Jan 10, 2013 8:25 am

Edward, thanks for clarifying that. That helps clear everything up.

Post Reply