Welcome to the SeeMeCNC Community

Thanks for the prompt reply, I've got the PID part sorted as it's well documented, like calibrating the hotend temp, but I can't find where it gets it lookup from, I've found the pin it comes in on from pins.h but after that I'm floundering.

http://forum.seemecnc.com/viewtopic.php ... bed#p10917

Mentions it but I'm still searching for the magic #define.

Setting up Steve's extruder.

Any useful hints and tips other than bin it. I've got 1.75mm PLA (measured at 1.77mm before and after extruder), it feed out nicely without the PTFE tube (4mm OD so I suspect 2mm ID) 620mm long. The PLA seems to snag up on the inside of the PTFE tube. I've tried tightening the screw, slackening the screws on the extruder to get it to move through. Getting a bit fed up now so probably time to have a break...

Starting to move ahead now, created my first part, a mount for dial guage. I've taken the arms off and 'fettled' them. Just before putting it to bed for the night decided to play with the dial guage. I really should have listened to my wife and not had another 5 minutes.

I checked the usual calibration points all referenced to centre of the bed as datum zero.

Z tower (X0, Y120) : -0.175mm (X0, Y-120) : -0.155mm
Y tower (X-60, Y104) : -0.1mm (X60, Y-104) : -0.382mm
X tower (X-60, Y-104) : -0.035mm (X60, Y104) : -0.565mm

Not quite what I was expecting, also, I'm not sure where to start looking for the issue. Any help appreciated.

First build and design was a dial indicator mount for a cheap and cheerful Silverline dial indicator. NB: in my design I used an M8 to mount the indicator which meant I had to drill out the mounting hole in the indicator to fit (it was 1/4").
Tried to load up the .3dm but it's not an allowed file so if anyone would like a copy, please PM me.
Having got the dial guage up and running time to check those zero's with the guage.

Good reading if sometimes a touch confusing with every ones unique terminology.

http://forum.seemecnc.com/posting.php?m ... =54&t=4035

Having read though this and many related posts, I found the information confusing till I had a go and spent the day adjusting the tower rotations within the EEPROM. For starters I used the DELTA_RADIUS_A/B/C to define the centre of the build are relative to the towers. I must admit I wouldn't have thought this would make much difference but having calibrated the printer using the recommended manual settings ( zero at each tower base, set stop screws, change DELTA_RADIUS till X, Y, Z tower positions and C = 0). I was fairly happy with the build and prints, then the dial indicator turned up and I could take measurements allowing my CDO (like OCD but letters in correct order). When I tried the opposite sides of the build plate I found quite large errors (0.68mm with opposite X tower). So time to recalibrate.

Set centre, simple enough, extrude a dot at x=0, y=0 then measure to the base of each tower. My measurements were:-

X => A 187.5mm
Y => B 186mm Add them together and divide by 3 => Centre = 186.3333mm
Z => C 187mm

Therfore, in my case modified:-

#define DELTA_RADIUS_CORRECTION_A 0.66666
#define DELTA_RADIUS_CORRECTION_B -0.83333
#define DELTA_RADIUS_CORRECTION_C 0.16666

Put another dot in the centre from the extruder and confirmed all towers equidistant (as accurate as my steel rule and eye could tell). This really threw out my X,Y,Z tower zeros, so went round each tower using

Script (1)
G28
G0 Z10 X0 Y120 F3500 ;Z Tower home
;G0 Z10 X0 Y-120 F3500 ;Inverse Z tower
G0 Z5.6 F50
G0 Z7
G0 Z5.6

Script (2)
G28
G0 Z10 X104 Y-60 F3500 ;Y Tower home
;G0 Z10 X-104 Y60 F3500 ;Inverse Y tower
G0 Z5.6 F50
G0 Z7
G0 Z5.6

Script (3)
G28
G0 Z10 X-104 Y-60 F3500 ;X Tower home
;G0 Z10 X104 Y60 F3500 ;Inverse Y tower
G0 Z5.6 F50
G0 Z7
G0 Z5.6

Script (4)
G28
G0 Z10 X0 Y0 F3500 ; Centre
G0 Z5.6 F50
G0 Z7
G0 Z5.6

The actual heights were unimportant as the dial indicator was reset to what ever I wanted to use as my datum (mostly Script (4)). The last 3 lines of the G-Code are probably superfluous as the whole point of a 3D printer is repeatable and accurate.

So having made the changes and reset the set screws and added 0.05 to my DELTA_RADIUS it was time to start on the tower rotations.

Starting with the following case.

#define DELTA_ALPHA_A 210 // X tower
#define DELTA_ALPHA_B 330 // Y tower
#define DELTA_ALPHA_C 90 // Z tower

Each tower was changed by 0.5 a degree clockwise and counter clockwise. So as not to bore those who are looking for the solution the workings are in the attached spreadsheet.
All the results point to the "simple" fact that changing the DELTA_ALPHA_A/B/C does not effect X, Y, Z and Centre homes but:-

Increasing DELTA_ALPHA_A : increases inverse Z, decreases inverse Y and no change inverse X
Increasing DELTA_ALPHA_B : decreases inverse Z, no change inverse Y and increases inverse X
Increasing DELTA_ALPHA_C : no change inverse Z, increases inverse Y and decreases inverse X

The inverse is true so decreasing will have the opposite effect. So how does this help.

Taking my case, X, Y, Z and C are zero'ish, Inverse Z = +0.21mm, Inverse Y = -0.05, Inverse X = +0.68mm. There must be an error in my mechanical build but having poured over the structure checking and re-checking as of yet I can't find anything so I'll try and tune it out with the software.

As Inverse X is my worst case, I can either:-

Increase DELTA_ALPHA_C to reduce inverse X (at the expense of increasing inverse Y) or
Decrease DELTA_RADIUS_B to reduce inverse X ( at the expense of increasing inverse Z)

or alternativly do a little of both.

Currently I don't have enough info to plot the way the variables change ie linear (y=mx+c) or exponentially so I'm going to compromise and try different numbers. Also, for the purists who are looking for all zero's I don't think it's possible with these tools, I think my best case will be zero at the towers and centre with 0.2mm at the inverse of the towers.

Maybe #define DELTA_DIAGONAL_CORRECTION_A/B/C might offer another method to reduce the differences, if anyone can let me know how this function works it would be appreciated.

Carried on with what I learnt yesterday, just out of interest I tried the bed a different temperatures just to quantify the differences, from room temp. (19 degrees C), the centre raised 0.06mm and the edges dropped 0.05mm uniformly round the edge at 65 degrees C.

Recapping, X, Y, Z and Centre towers are all zero'ish (max 0.02mm difference), Inverse X is the worst at +0.685mm, Inverse Z at +0.231mm and Inverse Y at -0.032mm. Adding them all together and dividing by gives me an average of 0.294mm so this will be my target for all 3 inverse X, Y, Z. (Edit: managed to get them all down to +0.26mm).

As the tower sections and centre aren't as affected, to save time I only probed the centre and Inverse positions. See previous spreadsheet for steps for tuning it in.

Once all the faffing around was completed, hot end refitted and Z0 dialled in, then the first big print left to run overnight.
Unfortunately in my excitement to remove it I popped one of the edges from the base so, today is going to be how to repair it.

What's up with your glass? It takes up the whole Onyx and it looks to be a bit thicker than the glass here stateside, did you have that cut custom?