FLSUN Delta (not mini) Setup

I had my FLSUN Delta working quite well for a beginner up to the point I warped the heated bed.
I replaced the thin aluminum bed with a plate of plate aluminum machined to 1/4” and same diameter as original, still heated.
Now I am trying to reset Marlin and Repetier to the new Delta Height of 306.9mm which was determined by using the -100, -1 and -0.1 Z control in Repetier.
I updated Marlin with this figure at #define DELTA_HEIGHT to the 306.9 number, saved it the uploaded it to Arduino successfully.
When I run G33 P1, the nozzle stops about 10mm above the bed and the printer LCD says I have an error, I reset the printer and home all with Repetier and the LCD tells me that the Z height is 295.
I have uploaded the Marlin code to the printer a number of times always with the same result.
I am unable to figure out where my mistake is.
I am using the FLSUN setup instructions from Github but using the figures measured on my larger printer.
I am using a Mac not Windows computer.
Can someone point me in a direction to solving this?

When you reflashed Marlin with the new height, you also refreshed the EEPROM? New Marlin values don't 'stick' unless you explicitly do that step. (M501 to load new Marlin defaults from firmware image, M500 to save to EEPROM)

You're sure your probe is working? M119 shows correct state when open/triggered? It's not triggering early or just from effector movement?

Quote
NAP1947
I had my FLSUN Delta working quite well for a beginner up to the point I warped the heated bed.
I replaced the thin aluminum bed with a plate of plate aluminum machined to 1/4” and same diameter as original, still heated.
Now I am trying to reset Marlin and Repetier to the new Delta Height of 306.9mm which was determined by using the -100, -1 and -0.1 Z control in Repetier.
I updated Marlin with this figure at #define DELTA_HEIGHT to the 306.9 number, saved it the uploaded it to Arduino successfully.
When I run G33 P1, the nozzle stops about 10mm above the bed and the printer LCD says I have an error, I reset the printer and home all with Repetier and the LCD tells me that the Z height is 295.
I have uploaded the Marlin code to the printer a number of times always with the same result.
I am unable to figure out where my mistake is.
I am using the FLSUN setup instructions from Github but using the figures measured on my larger printer.
I am using a Mac not Windows computer.
Can someone point me in a direction to solving this?

I have a variant (I modified) of the FLSUN delta. The default Marlin config files for 1.1.8 caused exactly this behavior. This didn't happen when using Pronterface, but did when I sent gcode to the printer. What I discovered is that the "nozzle basher" bed detection setting in the Marlin code caused the head to start printing about 10mm too high. I do not remember what I modified (I should have written it down), but it was something to do with the auto-leveling offset setting in the config file. I changed that to zero and everything worked correctly. Did you perchance just move to a newer version of Marlin?

DLC

---

Kits: Folgertech Kossel 2020 upgraded E3Dv6, Anet A8 upgraded E3Dv6, Tevo Tarantula enhanced parts and dual-head, TronXY X5SA Pro(E3DHemera).
Scratch: Large bed Cartesian, exchangeable heads, Linear slide Delta, Maker-Beam XL Micro Delta, 220x220CoreXY.

Gecko, thank you for your suggestion, yes I did send M502 and M500 (502 is what I had read, is that wrong?)

DLC60, I will need to digest your response and will do so shortly, currently I am having hard drive failure issues and am just getting a backup Mac working.
As soon as everything is connected I will work through your directions and let you know the results.

Cheers

Getting more confused.

I have re-installed Arduino, Marling and Repetier and setup them up based on the information that came with the FLSun Kossell and a GitHub article.

Once I have uploaded Marlin to the printer board and run M502 and M500, it will successfully run a G33 P1 calibration once.

If I try G33 P1 or any other G33 code the printer drops about 50mm (eyeball) and stops the returns to home for the X and Y axis but not the Z, the LCD screen tells me that X=0 Y=0 and Z=42.53.
S/D---% = 00.00
The bottom LCD line says "Endstops Z"

Delta Calibration screen on the LCD tells me that "HEIGHT = 42.53.

I have been through every line of the Condition.h file and cannot find any 42.53 number anywhere.

I have seen an article on Marlin that says I need to reboot the printer board, using the manual reset button, after uploading and giving M502 and M500 codes but there is no such button on my board which is a MakerBase MKS GEN L V1.0 board.

I am at a total loss as to what to try next.

Can anyone help or perhaps I have a problem with the board.

Check your endstops work correctly with M119 command.

Thanks Dust, M119 shows my endstops are working correctly, triggered at the top of the rails, open 1mm down.
The extruded is triggered down to 1mm from the base plate and open when it touches.

How about posting your configuration.h for examination. upload it to pastebin and show us the link.

Here is my link too Condition.h on PasteBin.

[pastebin.com]

Lot to go through, I really appreciate the help.

This morning the printer is responding more normally.
It confirms M119 at both top and bottom of the Z travel.
G33 P1, P2, P5 all complete correctly.
G33 P10 works up to the fourth set of test steps counterclockwise from the X upright at point 2 from the outer edge of the base plate it fails.
Running M119 show the Z Min failed to close when the nozzle lifted. This is not every time but intermittent.
I am assuming this means the Z min limit switch or its wire are causing part of the problem.
Any other thoughts before I get new limit switches?

Why do you have the X/Y/Z Min endstops anyway?

If you set them to match that the nozzle touches the bed in the middle, it won't be able to print towards the edges, and if you set them to match when the nozle is at the outer limit, it will be able to 'bottom out' in the middle of the print bed.

I have the X, Y, Z endstop switches positioned at the top end of the X, Y, Z, posts as per the kit instructions.
The Z is the only “Min Endstop” set in order to read when the nozzle touches the print bed.

Where is the endstop switch mounted?
Could we have a photo of the parts 'in location'?

I have attached a picture of the Z ENDSTOP limit switch on the effector housing. The other 3 are the same but mounted on the FLSun Kossel uprights.

I spent quite a bit of time last night checking the endstops using the M119 command and moving the Z in 1mm increments at both top and bottom of movement.

The switches are not repeatable, after Homing the printer all top switches show "triggered". Moving the Z down 1 mm changes all to "open" readings, however, moving the Z up it can take between 4 and 6mm to trigger them and they do not all trigger at the same measurement. Same for the Z max switch.

I assume from research this is caused by the metal arm used to move the actual switch button, some recommend removing the arms. I am temped to do this as the 3 upper switch can still be activated by the mount on the column, the defector Z max would need a slight tweak to make constant contact.

The is my first 3D printer and was the cheapest, recommended one I could find to learn on, I don't want to upgrade parts of this printer until I am sure I want to continue this hobby, if I do that I would most likely opt for a better quality printer.

thanks for everyones help.

That doesn't look all that much like a photo...

My mistake, real picture here. Is there a way to delete that first attachment?

I have moved the switch in the photo from that position to one where the setting screw is directly over the switch button rather than moving the lever arm.
It now registers 312.8mm down to open and up from there .2mm to close again.
This is repeatable multiple times with out variance.
Now I am working on the upper limits and have them also within .2mm from home to open on all 3 axis.
Not sure how much more accurate this printer will get.
Redoing the G33 commands now to see if it passes.

I've been googling your printer...
It's made of 20x20 aluminium profiles with plastic parts in the corners?

Maybe this one?
[www.gearbest.com]

It weighs 7Kg and has a build area of 1800 diameter, 300mm height.

My Overlord weighs 12Kg...
Most of that extra weight is in the aluminium extrusions...

I think the stiffness of the frame is going to be the limiting factor in the print quality/print speed issue, so that's probably where you'll have to focus next.
Maybe put in triangular braces in the top corners, and bolting another 20x20 onto the uprights.
(You drill holes through the additional profile, at every 100mm, and install T-slot nuts in the original profiles)

Gadget man, good progress today, not passing G33 tests yet but that is me at this point.
I found your stiffness comment interesting as I am considering making a piece of 1/4” Lexan I have into Y shaped form I can mount across the aluminum top profiles to keep the PLA spool to add weight and keep the foot print small. If I make it fairly large and bolt it on it should add some rigidity. Should I also consider cross brace from the top to uprights?

Hopefully last issue.
Printer is passing all G33 tests and ending up with .1 to .15 mm errors.
Effector Offset calculates out to 0.0.
However, doing a test print of a ThingiVerse test pattern the nozzle touches the base plate in the Z upright area and is stringy at the X and Y upright areas.
I have run the auto level (G33) tests numerous times with consistent results and the towers are all at 90 degrees to the base plate.
The base plate moves ever so slightly on the spring on the X and Y upright area in order to get the base plate at 90 degrees.
Should I remove the springs altogether and use shims to set the base plate upright relationship?
What else could be affecting this printer?

Have you verified that the arms are the exact same length?
And that the vertical riders have the exact same measurements?

"The base plate moves ever so slightly " If the base plate is not fixed solid, the calibration will be off.
0.15 error is a good start; I get to 0.015 on a Flsun kit printer.

But what did you do to fix the error of stopping above the plate with a probing error?

Good morning, worked late (for me) and made some more progress.
I built the arm lengths using a jig and each one measures 222mm centre to centre.
I need to measure from the top of the verticle riders to the centre of the arms, will post.
For the base plate I reversed the knurled knobs in the base plate brackets so the knurled part is on top of the bracket, I had to lightly file the hole in the bracket so the knurled part would turn. I fixed 3mm nuts and bolts to the base plate and threaded the bolts into the knurled part. I then used the knurled parts to level the base plate to the uprights. There is now no measurable movement in the base plate, either verticle or horizontal. (See attached picture)
I ran the G33 command before I went to bed and the Standard Deviation is now 0.047, repeatable.
After a couple coffees I will measure the verticle riders and calculate the Effector Offset then try the test print again.
I will post the results.

I have have found the main part of my issues. Before I tried G33 this morning I used M119 to check my limit switches, the effector auto calculation switch was very inconsistent, I tried to adjust and it stopped working altogether. I did some research and ordered a BL auto touch optical switch from Amazon, it should be here Wednesday. I also found q lot of lateral motion in the hot end pivot mount, I had an injection moulded one complete with cooling fan which I have installed, the BL switch is the exact size needed with this holder.
I made the stabilizing plate and fastened it to the top of the printer and also put the PLA reel on that, seems to feel different but that is subjective.
I am anxious to get these changes finalized to see what I have created, I appreciate everyone’s help to date.

Update, I could not make the BLTouch work and could not find help by searching the net or from the manufacturer, I sent it back.
What I have tested is the lamp and reader from a rotational speed sensor left over from a robot experiment.
It is a very simple plug and play on the original micro switch board connector, 3 leads instead of 2.
Run M119 with nothing between the reader and light and it shows triggered.
Put something between the reader and the light and M119 shows open, same operation as the micro switch.
These are available from various suppliers for between 2 and 4 dollars US, the ones I have use the number B83609.
Would work for all four switch positions.
I am in the process of setting up the device and will update when it is running.
Cheers

To close this thread I tried the speed sensor light and failed to get the prints I expected.
I have removed the auto levelling feature and will setup using the manual calibration method.
Hopefully that will work.